U.S. FOOD AND DRUG ADMINISTRATION

     CENTER FOR DEVICES AND RADIOLOGICAL HEALTH

         MEDICAL DEVICES ADVISORY COMMITTEE

 

                     +  +  +  +

 

     GASTROENTEROLOGY AND UROLOGY DEVICES PANEL

 

                     +  +  +  +

 

                     WEDNESDAY,

                    JUNE 8, 2005

 

 

            The Panel in the Walker/Whetstone Room of

the Holiday Inn Gaithersburg, 2 Montgomery Village

Avenue, Gaithersburg, Maryland at 9:00 a.m., MARK A.

TALAMINI, M.D., Chair, presiding.

 

PRESENT:

 

MARK A. TALAMINI, M.D.       Chair

ABDELMONEM AFIFI, Ph.D.      Voting Member

GEORGE R. ARANOFF, M.D.      Voting Member

CHRISTOPHER BLAGG, M.D.      Consultant

WILLIAM DUFFELL, JR., Ph.D.  Industry Representative

CHRISTOPHER HOY, M.D.        Consultant

RICHARD GIBSON, M.D., M.P.H. Consultant

ROBERT GILLESPIE, M.D.       Consultant

SUSAN J. KALOTA, M.D.        Voting Member

ROBERT LOCKRIDGE, JR., M.D.  Consultant

CHRISTINE MOORE              Consumer Representative

JOHN MORAN, M.D.             Consultant

JOHN SADLER, M.D.            Consultant

GERALD SCHULMAN, M.D., FASN  Consultant

MATT WEINGER, M.D.           Consultant

 

NANCY C. BROGDON, Director, Division of

      Reproductive, Abdominal, and Radiological

      Devices

JEFFREY W. COOPER, DVM, Executive Secretary

 


FDA PRESENTERS:

 

SUSAN S. ALTAIE, Ph.D.

Scientific Policy Advisor, OIVD/CDRH

 

SUSAN GARDNER, Ph.D.

Director, OSB/CDRH

 

CAROLYN Y. NEULAND, Ph.D.

Branch Chief, GRDB/DRARD/ODE/CDRH

 

JOSHUA C. NIPPER, M.E.

Biomedical Engineer, GRDB/DRARD/ODE/CDRH

 

MICHAEL MENDELSON, D.D.S., M.S.

Biomedical Engineer, Director Health Promotion

      Officer, Human Factors Science and Engineering

      Branch

 

CLAUDIA C. RUIZ‑ZACHAREK, M.D.

Medical Officer/Nephrologist, GRDB/DRARD/ODE/CDRH

 

 

PUBLIC SPEAKERS:

 

ROD KENLEY

Aksys, Ltd.


                      I‑N‑D‑E‑X

 

AGENDA ITEM:                                    PAGE

 

CALL TO ORDER:                                     4

 

Introductions                                      4

 

FDA's Critical Path Initiatives, presentation     12

Susan S. Altaie, P.D., Scientific Policy

      Adviser, Office of In Vitro Diagnostic

      Device Evaluation and Safety

 

Post Market Study Design                          19

Susan Gardner, Ph.D., Office of Surveillance

      and Biometrics

 

OPEN PUBLIC HEARING:                              32

 

OPEN COMMITTEE DISCUSSION:                        33

 

1.  Regulatory Briefing ‑                         34

    Carolyn Neuland, Ph.D., Branch Chief of

          the Gastro‑Renal Devices Branch

 

2.  Overview of Conventional Hemodialysis         47

    Systems ‑ Josh Nipper, M.E.

 

3.  Human Factors ‑ Mike Mendelson, D.D.S.,       61

          M.S., Biomed Eng

 

4.  FDA Presentation ‑ Claudia Ruiz. M‑D.,        80

    Nephrologist

 

5.  Questions for the Panel                      103

 

OPEN PUBLIC HEARING:                             309

 

7.  Final Comments                               322


               P‑R‑O‑C‑E‑E‑D‑I‑N‑G‑S

                                       (9:04 a.m.)

                   CALL TO ORDER

            CHAIRPERSON TALAMINI:  I would like to call to order this meeting of the Gastroenterology and Urology Devices Panel.

            My name is Dr. Mark Talamini.  I am the chairperson of the Gastroenterology and Urology Devices Panel.  I am currently professor of surgery at the Johns Hopkins University School of Medicine and Director of Minimally Invasive Surgery there.

            If you haven't already done so, I would request that everyone in attendance at this meeting sign in on the attendance sheet that is available on the table outside the door.

            At this time I would like each panel member at the table to introduce him or herself, state your specialty, position title, institution affiliation, and status on the panel.  And we'll start over here on the right.

                   INTRODUCTIONS

            MS. BROGDON:  Good morning.  I'm Nancy Brogdon.  I'm not a member of the panel.  I'm the director of FDA's Division of Reproductive, Abdominal, and Radiological Devices.

            DR. HOY:  Chris Hoy, a clinical nephrologist from upstate New York affiliated with the Rubin Dialysis Center.  And we have a nocturnal program for seven years.  I'm a consultant.

            DR. LOCKRIDGE:  Bob Lockridge.  I'm a clinical nephrologist from Lynchburg, Virginia.  I started the first nightly program in Lynchburg in '97.  I've trained over 57 patients and have over 159 hours of experience in nocturnal patient care years.

            DR. MORAN:  I'm John Moran.  I'm a clinical nephrologist.  I'm chief scientific officer at Satellite Health Care in northern California and consulting professor at Stanford University School of Medicine.  I'm also a panel member.

            DR. BLAGG:  Chris Blagg.  I'm emeritus executive director of the Northwest Kidney Center in Seattle and professor of medicine emeritus, University of Washington.  And I've been involved with home dialysis and nocturnal dialysis for more years than I like to think.  I'm a consultant.

            DR. GILLESPIE:  I am Robert Gillespie.  I'm a pediatric nephrologist from Corpus Christi, Texas.  I'm affiliated with Driscoll Children's Hospital as well as an assistant professor of pediatrics at Texas A&M University.

            DR. WEINGER:  Matt Weinger.  I'm a professor of anesthesiology, biomedical informatics, and medical education at Vanderbilt University.  And my area of expertise is in human factors engineering and patient safety.  And I'm a consultant.

            DR. GIBSON:  Richard Gibson, adult nephrology and psychiatry, Tulane University, consultant.

            DR. COOPER:  My name is Jeff Cooper.  I'm the executive secretary of the panel and a veterinary medical officer at FDA.

            DR. KALOTA:  I'm Susan Kalota, a private practice urologist in Tucson, Arizona, panel member.

            DR. ARANOFF:  I'm George Aranoff.  I'm chief of the Nephrology Section and professor at the University of Louisville.  And I'm a consultant.

            DR. AFIFI:  I'm Abdelmonem Afifi.  I'm professor of biostatistics and former Dean of the School of Public Health at UCLA.  I'm a biostatistician.  I'm a panel member.

            DR. SADLER:  I'm John Sadler.  I'm a clinical nephrologist from Baltimore, former head of nephrology at the University of Maryland.  I've been involved in dialysis for more than 40 years now and formerly had the privilege of chairing this panel.

            DR. SCHULMAN:  I'm Gerald Schulman.  I'm a professor of medicine at Vanderbilt University.  I'm the director of hemodialysis for the hospital.  And I'm a consultant.

            DR. DUFFELL:  I'm Bill Duffell.  I'm an industry rep for this panel.  My graduate studies were in sciences and pharmacology.

            MS. MOORE:  I'm Christine Moore, retired executive assistant and dean of students at the Baltimore City Community College, consumer rep.

            CHAIRPERSON TALAMINI:  Thank you very much.

            I'll now turn the meeting over to the executive secretary, Dr. Jeff Cooper, who would like to make some introductory remarks.

            DR. COOPER:  Good morning.  I will now read the record of the conflict of interest statement.  "The following announcement addresses conflict of interest issues associated with this meeting and is made a part of the record to preclude even the appearance of an impropriety.

            "To determine if any conflict existed, the agency reviewed the submitted agenda for this meeting and all financial interests reported by the committee participants.

            "The conflict of interest statutes prohibit special government employees from participating in matters that could affect their or their employers' financial interests.  However, the agency has determined that participation of certain members and consultants, the need for whose services outweighs the potential conflict of interest involved, is in the best interest of the government.

            "Waivers have been granted for Drs. Christopher Blagg and John Moran for their interests in firms that could be impacted by the panel's deliberation.  Copies of these waivers may be obtained from the agency's Freedom of Information Office, room 12A15 of the Parklawn Building.

            "We would like to note for the record that the agency took into consideration certain matters regarding Drs. John Sadler and Gerald Schulman.  Each of these panelists reported current and/or past interests in the firms at issue but in matters not related to today's agenda.  The agency has determined, therefore, that they may participate fully in today's deliberations.

            "We would also like to note for the record that the agency took into consideration other matters regarding Drs. Robert Lockridge, Blagg, and Sadler.  The panelists reported current and/or past interests in firms at issue in matters related to today's discussion.  Since the agenda item for this session involves only particular matters of general applicability, the agency has determined that these panelists may participate fully in the discussion.

            "In the event that the discussions involve any other products or firms not already on the agenda for which an FDA participant has a financial interest, the participant should excuse him or her self from such involvement.  And the exclusion will be noted for the record.

            "With respect to all other participants, we ask in the interest of fairness that all persons making statements or presentations disclose any current or previous financial involvement with any firm whose products they may wish to comment upon.

            "The FDA seeks communication with industry and the clinical community in a number of different ways.  First, FDA welcomes and encourages pre‑meetings with sponsors prior to all IDE and PMA submissions.  This affords the sponsor an opportunity to discuss issues that could impact the review process.

            "Second, the FDA communicates through the use of guidance documents.  Toward this end, FDA develops two types of guidance documents for manufacturers to follow in submitting a pre‑market application.  One type is simply a summary of the information that has historically been requested on devices that are well‑understood in order to determine substantial equivalence.  The second type of guidance document is one that develops as we learn about new technology.  FDA welcomes and encourages the panel and industry to provide comments concerning our guidance documents.

            "I would also like to remind you that the tentative date for the last 2005 Gastroenterology and Urology Devices Panel meetings is scheduled for Friday, October 21st, 2005.  You may wish to pencil in these dates on your calendars, but please recognize that this date is tentative at this time.

            "On another note, information on purchasing transcripts or videos of today's meeting can be found on the registration table, which is outside of the meeting room."

            Before I turn the meeting back to Dr. Talamini, please ensure that all cell phones are either turned off or placed in the silent ring mode so that they do not interrupt the business of the meeting.

            Dr. Talamini will now continue.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Cooper.

            Dr. Susan Altaie of the Office of In Vitro Diagnostic Devices will now give a presentation to the panel regarding the agency's critical path initiatives.  Dr. Altaie?

            DR. ALTAIE:  Thank you and good morning.

   FDA'S CRITICAL PATH INITIATIVES, PRESENTATION

            DR. ALTAIE:  I am the focal point for Center for Devices under the critical path umbrella.  And in my daytime job, I am Scientific Policy Adviser for the Office of In Vitro Diagnostics.

            Today I will talk to you about what the critical path is, why is this an FDA initiative, and what are the critical path tools that we speak about, what projects we are pursuing at CDRH, and how you can get involved as an interested individual.  And then we can answer questions if you have any.

            Critical path is a serious attempt to modernize the medical products' critical path and market product development to make it more predictable and less costly.

            If you look at the critical path for device developments, you are looking at basic research prototyping, preclinical, and clinical development, and then hopefully FDA application and large production scales.

            Well, critical path will cover everything in that life cycle of development except for the basic research.  So one might think why FDA gets involved in such a path.  It's an industry job to develop.

            Well, because FDA believes in tremendous benefit of bringing innovative products to the public faster.  We also have a unique perspective and understanding of product development.  And we see the failures, the successes, and the missed opportunities.  And because critical path will help us develop guidances and standards that foster innovation and improve chances of success in the device development.

            We like to work together with companies and patient groups and academic researchers and other stakeholders to modernize and develop and disseminate tools that will help remove the science hurdles that affect product development.

            I spoke about tools on the critical path.  These are methods and techniques that work in three dimensions of safety, efficacy, and industrialization of the devices.

            Under safety tools, those are the tools that will predict if potential products will be harmful.  And the efficacy tools will determine if a potential product will have medical benefit and industrialization tools.  Those will deal with how to manufacture a product at a commercial scale with consistently high quality.

            So these tools that I describe in three dimensions, these are examples of them.  Biomarkers can be one of them by Bayesian statistics, animal models of biomarkers, clinical trial design tools and computer simulations, quality assessment, protocols, and post‑market reporting.  And we are also open to any suggestions any of you have to add to this list of tools to follow up and ease up the device development.

            At the CDRH, we deal with an array of devices, anywhere from Band‑Aids to stethoscopes to CAT scans and heart valves and infusion pumps and glucose monitoring devices.  Everything that you can think of medically are regulated under CDRH.  And so there is a tremendous opportunity for developing tools to get these devices faster to the market.

            However, I want to note that we at CDRH deal with devices that are totally different than the drugs dealt with in CDER.  Even though we parallel in paradigm, the paradigms are different.  We deal with complex components of the devices, and we deal with biocompatibility.

            We have to have equipment that is durable and they long‑last.  And then we also deal with devices that have rapid cycle and turnover to the better model.  We deal with malfunctions of instrumentation and user errors.  And we study them by actually bench testing them.

            Our regulations are different.  We deal with quality systems in ISO 9000, which is parallel but different than the GMPs that CDER deals with.

            These are also areas of interest under the three dimensions that I spoke with you about.  The safety tools we are dealing with biocompatibility databases, effects of products on disease or injured tissue.  When you look at the effectiveness tools, you are talking about surrogate endpoints of cardiovascular device trials, and you're talking about computers simulating modeling of the implanted devices.

            Under the mass production, we are looking at the practice guidelines for follow‑up of implanted devices.  We are also looking at validated training tools for devices with known learning curves by the users.

            These are a few examples of critical path projects that we're pursuing in CDRH.  Most of these are not funded.  And we're trying to leverage funding from outside stakeholders, whether it's industry or academia or government, other government agencies.

            We are looking at developing a serum panel to assess sensitivity and specificity of the hepatitis assays.  We are looking at computer models of human physiology to test and predict failures of peripheral vascular stents before going into animals or human.

            We are looking at a clear regulatory path for intrapartum fetal diagnostic devices.  We are looking for consensus from a safety community.  And under the surrogate markers, we are looking for pathways of statistical validations of these markers.

            We are looking for development to develop practice guidelines for appropriate monitoring of permanently implemented devices.  And we also are looking for knowing the extent of the neurotoxicity testing that is required in the tissue contacting, neural tissue contacting, materials.

            So, as I said, these are the projects we have picked up with not having funding, but all of these projects are in some steps of development and progress.  If you are interested, you can participate in two ways.  You could give your views and comments through the open public docket and tell us about areas that benefit from research and development of critical path evaluative tools.

            You also can help us compile this national critical path opportunities list in the areas that we have missed probably.  And there are ways to contact us.  You could go on the CDRH Web page under "News and Events," and you can go to the links to the critical path white paper.  And then there are links for the docket and the critical path group that you could contact if you have any ideas.

            I would like to leave you with this thought that at CDRH, we work to ensure the health of the public throughout the total product life cycle of the products.  And we believe it is everyone's business.  So I request your involvement in critical path.

            Any questions?

            (No response.)

            DR. ALTAIE:  All right.  Thank you.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Altaie.

            I need to remind all of the panel members and speakers to talk directly into the microphone when you talk for the transcriptionist's benefit.  If they don't hear you, it's like you haven't said it.  Are you the transcriptionist over there?  Is that person in the room?  So if you aren't getting something, raise your hand.  And I'll make sure that we have people speak clearly.

            Next, Dr. Susan Gardner of the Office of Surveillance and Biometrics will give a presentation to the panel regarding the role of OSB in the review of post‑market study designs.  Dr. Gardner?

            DR. GARDNER:  Thank you and good morning.

             POST MARKET STUDY DESIGN

            DR. GARDNER:  I am going to spend a few minutes this morning telling you about an important programmatic change in CDRH and how that might impact you as panel members.

            First of all, the basis of the change is the move of the conditions of approval program from the Office of Device Evaluation, ODE, to the Office of Surveillance and Biometrics, which is the office which I'm the director of.

            Briefly, what we do in OSB is we are involved in pre‑market review by virtue of the fact that we have the statisticians in OSB who are almost always involved in PMAs, and we also have the epidemiologists in the office, who increasingly are going to be involved in the PMA review process.

            We are responsible for adverse event signal detection.  Again, we have the post‑market monitoring tools in the office, including our medical device reporting program, our Medsun program, and other surveillance tools.

            We are responsible for risk characterization.  By that I mean the analysis of the post‑market data and coordination of the center response for health care professionals when we see a post‑market problem and we have a solution to get to.  We are also responsible for interpreting the medical device regulation.

            For condition of approval studies, the regulation tells us that in 21 CFR 814, post‑approval requirements can include continuing evaluation and periodic reporting on the safety, effectiveness, and reliability of the device for its intended use.  Again, this is the basis for a condition of approval studies program.

            So the impetus for change came from an internal study that we did a couple of years ago.  We looked at all of the PMAs that were approved from 1998 to the year 2000.  There were 127 PMAs.  Forty‑five of those were approved with condition of approval study orders.

            So we went back.  Actually, the initial effort was to go back and look at the quality of the condition of approval studies, but we ran into a problem because we actually couldn't locate a lot of these studies.  It was a real red flag for us.

            We found out that we had very limited procedures for tracking the progress and results of these studies.  There was no automated tracking system or even manual tracking system in many cases.

            We also found, as you might expect, that many of the lead reviewers for these PMAs had moved on to other jobs, either in the agency in a different division or outside the agency.

            Finally, the folks in the pre‑market, particularly towards the end of an approval process, are really heads down in improving the product, doing the labeling, and all of the other things they need to do to get the product out the door.  And they really lack the resources to focus on the condition of approval study.

            So we have a strategy for change.  And obviously the goal of the change is to obtain what is really critical post‑market information as the device moves onto the market and to continue to assure the safety and effectiveness of the device.  And the important words, of course, are "in real world use."

            The device is moving from the tightly controlled clinical trials into community use.  And so it's really a critical time to know what is going on if we have post‑market questions.

            We want to be able at that point to better characterize the risk‑benefit profile of the device and, of course, to add to our scientific database so we can make better decisions.

            So what did we do?  Starting in January of this year, we transferred the condition of approval program to OSB, although I will say this was done after a two‑year pilot, where we tested some of the procedures that I am going to tell you about.

            We have developed and instituted and is now up and running an automated tracking system.  So we're going to be able to acknowledge the receipt of study reports and do follow‑up of reports if they're not received.

            We have also added an epidemiologist to the PMA review team when it appears that there is probably going to be a condition of approval study.  The task of the epidemiologist is, first of all, to think post‑market during the pre‑approval process.

            So during the approval process, their task with developing a post‑market monitoring plan, it may or may not include a condition of approval study, but, as I say, sort of when we think there is going to be one, that is when we will have the epidemiologist on the team from the beginning.

            The epidemiologist will take the lead in developing.  And, again, some key words here are "well‑formulated post‑market questions," important questions.  They are going to have the lead in the design of the condition of approval study protocol.

            And they are going to be responsible for evaluating the study progress and the results of the study once the product is approved.  They will continue to work with the PMA team throughout this process.  It's just a matter of adding their expertise to the team and changing the lead as the product goes to market.

            Why do we think these changes will improve the program?  Well, first of all, the development of important post‑market questions and a good study design should be motivation both for FDA and for industry to get these studies done.

            We're certainly willing to discuss the design of the studies throughout the conduct of the studies.  And if something changes while the product is on the market, we are willing to make those adjustments.  Again, it's time, money, and resources.  And it's important that we answer the important questions.

            Second of all, we're committed to acknowledging these studies when they come in to CDRH and giving feedback on the studies.  And, again, this is added motivation, both for the agency and for industry to get these studies done.

            We will have a Web site, where we will post the status of the studies.  So for folks who are doing them well, that will show up.  And if folks aren't doing them well, that will also be on our Web site.  And we do have the ability to mandate a post‑market study or order a post‑market study if these studies are not being done.

            So what is the impact on the advisory panel?  Post‑market questions come up often naturally during the discussion of products as you are looking at them for approval.  And sometimes we are going to have them come up intentionally or pose post‑market questions during the process.  And we are going to look to you.  We want your advice, your suggestions on possible approaches and any help that you can give us in thinking about what the post‑market strategy might be.

            During the post‑market period, we're committed to having either FDA or industry come back to the panel and give you an update on the progress and results of the condition of approval studies for the approved devices.

            Thanks.  Do you have any questions?  Yes?

            DR. DUFFELL:  Will OSB be managing registries as well as post‑market surveillance?

            DR. GARDNER:  Yes.  The condition of approval studies could have many possible designs.  And a registry is certainly one design that could be part of the study.  It would be managed somewhere else, but we would be part of looking at the data and doing the analysis.

            Yes?

            DR. MORAN:  Would it be fair to say that this might expedite the approval process if there are issues that you feel you can deal with after approval, after marketing?

            DR. GARDNER:  Well, certainly the pre‑market requirements for safety and effectiveness don't change.  But I think, again, thinking hard about what our post‑approval questions are, thinking about what a good post‑approval study design could be and how we could answer those questions are just going to make us all feel more comfortable with the product.  And knowing, again, that these studies are going to be done I think, again, will add to our comfort level.

            CHAIRPERSON TALAMINI:  Dr. Gardner, do you know exactly ‑‑ this is Talamini here ‑‑ what percentage of post‑market studies were completed out of that number?

            DR. GARDNER:  Well, I don't.  We had a number that we were using because it was the number that we could find and the ones that were not accounted for.

            What we did not do was to go back to industry and call them up and say, you know, "We don't have your post‑market study.  Did you do it or did you not?"  It's not a place we would have liked to have been, but just because for resource issues, we made the decision to move forward, rather than to go back.

            The products, again, were approved between 1998 and 2000.  So if the studies weren't done, these were products that we had been looking at for a number of years.  So it wasn't clear, again, sort of a resource issue, that going backwards would have been as beneficial as going forward in this case.

            CHAIRPERSON TALAMINI:  Any other questions for Dr. Gardner?  Yes?

            DR. AFIFI:  I appreciate the increased role for an epidemiologist in post‑market studies because of their expertise in data collection and so on.  My question is regarding the analysis.  How will that be coordinated between the epidemiologist and the statisticians?

            DR. GARDNER:  Well, mostly the analysis should be done by the companies.  What they send us is a report, similar to what is sort of done now.  They're conducting the study.  They should collect the data.  They should do the analysis.

            The data will come in to us.  We will review it.  If we have any questions, we get to go back to the company and ask them to provide additional information.  But the epi person will take the lead in doing the analysis, but they will share this with people who were members of the PMA approval team.  So that they will also have input into the condition of approval study.

            DR. SADLER:  You said these things will be posted on the Web site.  And obviously a good posting will be the carrot for the company ‑‑

            DR. GARDNER:  Yes.

            DR. SADLER:  ‑‑ to encourage them to get the studies done, for which I am very happy to hear.  You also said that there might be penalties and other studies if they were not completed.  Could you go a little further on the penalties?  Who would design subsequent post‑marketing studies?

            DR. GARDNER:  Yes.  We have a regulation.  Well, actually, the law says that we are allowed to impose post‑market studies for various reasons.  And essentially the easy way to say it is if we have a major post‑market question, then we think that there might be a danger to the people using the device.

            So if we have a well‑formulated post‑market question for the condition of approval studies that has not been answered, that will translate very well into this requirement that we have to do this post‑market study.  It's called section 522.

            So when we do this, we go to the company.  And we tell them what our question is.  It is their job to go back and design a study and come back to us and say, "This is what we are proposing."

            Now, if we have gotten to this point with the company, obviously we're running into a lot of problems.  But at that point if they are still not doing it, we can impose civil money penalties and the product can be labeled as misbranded.

            DR. SADLER:  Okay.  Would you expect to consult with the advisory panel members or with some of them in reviewing these subsequent studies?

            DR. GARDNER:  I think it is entirely possible.  I mean, we have an appeals process.  And, again, if we are sort of to the point where we are ordering a secondary study because the condition of approval study hasn't been done, there's a message here about some real conflict between industry and FDA.

            So we would go up through the appeal process.  That would also include perhaps going to our dispute resolution panel if the company asked to do that.  So we would go through all of those mechanisms as we sort of wind our way there.

            Again, we're hoping that input from the panel if the product is approved, working hard on designing the studies well, whatever, can avoid those kinds of issues.

            CHAIRPERSON TALAMINI:  Any other questions for Dr. Gardner?

            (No response.)

            CHAIRPERSON TALAMINI:  If not, thank you very much.

            DR. GARDNER:  Thank you.

                OPEN PUBLIC HEARING

            CHAIRPERSON TALAMINI:  We will now proceed with the first of the two half‑hour open public hearing sessions of this meeting.  The second half‑hour open public hearing session will follow the panel discussion this afternoon.

            At these times, public attendees are given an opportunity to address the panel, to present data or views relevant to the panel's activities.

            I would like to remind public observers at this meeting that while this portion of the meeting is open to public observation, public attendees may not participate except at the specific request of the chair.

            I would ask when persons address the panel now or later, they come forward to the microphone and speak clearly as the transcriptionist is dependent on this means for providing an accurate transcription.

            If you have a hard copy of your talk available, please provide it to the executive secretary for use by the transcriptionist to help provide an accurate record of the proceedings.

            No individual has given advance notice of wishing to address the panel this morning.  If there is anyone now wishing to address the panel, please identify yourself at this time.  Do we have anybody in the audience?

            (No response.)

            CHAIRPERSON TALAMINI:  It does not appear that we do.  Okay.

             OPEN COMMITTEE DISCUSSION

            CHAIRPERSON TALAMINI:  Since there are no public comments this morning, we will proceed to the open committee discussion.  We will start with the FDA's presentation on nocturnal home hemodialysis.  The first speaker for the FDA is Dr. Carolyn Neuland.

              1.  REGULATORY BRIEFING

            DR. NEULAND:  Good morning.  Thank you, Dr. Talamini.

            As Dr. Talamini has said, I am the Branch Chief of the Gastroenterology and Renal Devices Branch.  And this is the branch within FDA that is responsible for reviewing medical devices for hemodialysis products, which is the main topic of today's session.

            I would also like to take this opportunity to welcome each of you to the session today and thank you again for taking time out of your busy schedules to share with us your clinical experience and your scientific knowledge in the area of home nocturnal dialysis.

            Before we get into the in‑depth discussion, I would like to cover a few introductory‑type issues.  The first thing I would like to do is introduce the members of the Gastroenterology and Renal Devices Branch who are present today.  These are individuals who review the medical devices related to hemodialysis.

            I will then discuss with you very briefly an update on the advisory panel that occurred in January 2003 and let you know the outcome of that deliberation that you had at that time.

            I will then go into the regulation of hemodialysis devices very briefly.  I know you covered this last night in your training, but I will just touch upon how we review hemodialysis products in general.  And then I will discuss the guidance documents for hemodialysis that currently have been written.

            Finally, I will give you a working definition today of nocturnal home hemodialysis that you can work from and then bring forth the main meeting objectives for today's meeting.

            Okay.  I am going to announce the members of the Gastroenterology and Renal Devices Branch that are present.  If they would please identify themselves when I mention your name?  We have present Linda Carr, who was our consumer safety technician, who prepared all of our slides today.  I guess Linda hasn't shown up yet.

            Dr. Jeffrey Cooper, who is our panel executive secretary, sitting at the head table; Linda Dart, who is a biochemist in our group, who reviews dialysis products; Gema Gonzalez, a biomedical engineer, who is very active in the review of our dialysis product; Dr. Irada Isayeva, who is a polymer chemist in our group; Dr. Kristina Lauritsen, who is a biologist; Barbara McCool, our nurse consultant, very active in dialysis; Joshua Nipper, our biomedical engineer.  And you will be hearing from Joshua Nipper shortly, who is going to give you an overview of conventional dialysis equipment.

            Kathleen Olvey, a biologist in our group; Dr. Claudia Ruiz‑Zacharek.  Dr. Ruiz is going to give the clinical presentation today on our issues related to nocturnal home hemodialysis.

            Rebecca Stephenson, a chemical engineer in our group.  She is the youngest person in our group and the newest to join FDA.  Kellie Straughn, our clerk‑typist; and then Mr. Richard Williams, a mechanical engineer in our group.  I'm sure many of you have talked to these individuals over the history of reviewing these products.  Okay.

            One of the requirements for the advisory panel process is to ensure that the advisory panel members receive adequate follow‑up and feedback on the outcome and the status of previous products that have been brought before this advisory panel.  Therefore, I will take the next couple of minutes to update you on the status of the PMA that was brought before the advisory panel in 2003, the last session of this group.

            This device is what's called the Enteryx procedure kit.  It is marketed currently by Boston Scientific Corporation.  This device is a solution which is injected into the lower esophageal sphincter for the treatment of gastroesophageal reflux disease in patients who currently are responsive to pharmacological therapy.

            This product, this PMA product, came before the advisory panel in January 17th, 2003.  And the recommendation at that time was for approval with conditions.  The conditions were related to modifications to the physician labeling; modifications to the patient labeling; and for a post‑market study, as you just heard.  That was one of the primary recommendations.  And the study was to occur for three years post the last injection of the material.

            Well, I am happy to say we did approve the PMA following that recommendation.  It was approved on April 22nd, 2003.  A post‑approval study was a requirement of the conditions of approval.  It was a three‑year post‑approval study, which required the enrollment of up to 300 patients, some of which were patients that had already been in the initial study and were going to be followed out to three years.  Others were new patients that were going to be enrolled.  The 300 patients were required.

            We also were interested in looking at reinjection of the material in that study.  And I am also happy to say that study is actively ongoing.  We receive annual reports from them.  The study is not completed yet because it was only approved in 2003, but the study is actively ongoing.

            Now, I just wanted to mention briefly that there have been a few MDR reports that have come out on this device which have shown adverse events related to the transmural injection of the Enteryx material from improper implantation techniques.  This is typically physician related in how they are injecting the material.

            This transmural injection has resulted in the placement of the Enteryx material into the aorta, the media steinum, or into the plural space in a few patients.  In one example, the material may have through the aorta through a branch of the right renal artery.

            In addition, one patient who had inadvertent transmural injection into the wall of the aorta developed an aorta or anterial fistula, which that patient then, unfortunately, succumbed to bleeding and passed away.

            As a result of these types of adverse events, again, I wanted to emphasize this is physician‑related.  It is a technique that is occurring.  It was determined that we needed to have some additional labeling changes to the PMA.  And the company has worked extensively with us on this.  And we have rewritten the labeling and instructions for use, stressing the proper technique for injection of the material.  We will continue to monitor this closely.  And I think that with the additional training, this problem should go away.

            Okay.  Now I would like to turn our attention to the main topic of today, which is nocturnal home hemodialysis.  How does FDA regulate hemodialysis products?  Well, most all of the products for hemodialysis are regulated through a 510(k) process.  These products are class II medical devices.  The classification process is risk‑based.  Class II products generally are considered of a moderate level of risk.

            There are some requirements for class II devices for general controls and special controls to ensure the safety and effectiveness of these products.  General controls include such things as registration and listing the submission of a pre‑market notification, quality system regulations, and things of that nature.  And special controls are such items as guidance documents development following performance standards and other types of safety issues such as those.

            As I said, these devices, the hemodialysis products, are regulated under a 510(k) process called a pre‑market notification.  And these devices are determined when they are marketed to be substantially equivalent to a device that is already currently on the market.  We call this a substantial equivalence clearance process.

            I do want to note that currently there are no devices that are currently labeled and approved for the use of nocturnal home hemodialysis in the United States.

            I said this was a clearance process.  Substantial equivalence is determined.  And when a device is placed on a market through this 510(k) clearance process, it is determined to be as safe and as effective as the predicate device.

            Performance data is usually required for these submissions.  Typically it is bench studies.  However, if there are new modalities, we frequently require clinical studies.  And we have required clinical studies for devices in hemodialysis that are going to be put into the home setting.

            Hemodialysis products are listed in the Code of Federal Regulations under a number of classifications.  The primary ones for hemodialysis equipment are under 876.5820 and 876.5860.  The difference between these two classifications has to do with whether an ultrafiltration controller is added to the system, and the 5860 is for high flux dialyzers.  Both of those types of devices are classified under those classifications.

            We have a number of other classifications that deal with dialysis products.  I am bringing this out so that you realize we don't bulk everything into one classification.  We have tried to spread out these different products into different regulations.

            We have a separate classification for sorbent regenerated dialysis delivery systems for hemodialysis.  We also have a separate regulation for water purification systems.  Again, we have products for peritoneal dialysis that are separate from hemodialysis.

            And we have blood access devices and accessories for the dialysis process.  I just would note that these currently are class III, but they are still reviewed under the 510(k) process.  We are looking into the reclassification of these products currently.

            We have written a number of guidance documents for hemodialysis products.  They have included guidances for conventional high‑permeability dialyzers, guidance documents for the hemodialysis delivery systems.  We have a guidance on the reuse of hemodialyzers.  And we also have a fairly detailed guidance document on water purification systems, which is a significant component that raises a lot of safety issues in the dialysis process.

            Now, it's important for me to bring these issues out to you because today you are going to be challenged with giving us advice which may lead to the development of a guidance document for nocturnal home hemodialysis.  So I would just like to make a couple of comments about guidance documents.

            Guidance documents are actually listed in the Code of Federal Regulations.  And it stated that the guidance documents are documents prepared for FDA staff, applicants, and sponsors, as well as the public that describe the agency's interpretation of or policy on a regulatory issue.

            Now, guidance documents represent FDA's current thinking on a specific issue or a device.  Guidance documents do not establish legally enforceable rights or responsibilities.  They do not legally bind the FDA or the public.

            An applicant may choose to use an approach other than the one set forth in a guidance document.  However, the alternative approach must comply with the relevant statutes and regulations that have been written.

            So just keep that all in mind as you go through your deliberations today and your discussions and give your advice because I do hope that we result in a guidance document for nocturnal home hemodialysis.

            Okay.  Well, this leads us into the discussion of today's main topic:  nocturnal home hemodialysis.  And I would like to give you what I consider a working definition for this topic today.  You may add or take away and discuss this further, but this is the definition that we have come up with.

            Nocturnal home hemodialysis is a type of hemodialysis performed in the home by the patient while the patient is asleep.  Typically this may be done at night.  It is done over a six to ten‑hour period using slower flow rates for the blood and dialysate or generally slower flow rates and is frequently performed five to seven days per week.

            Today we would like you all to engage in a discussion of this topic of nocturnal dialysis, and I have here the three main objectives I would see as hopefully coming out of this meeting.

            The first is to discuss and provide recommendations on the clinical and scientific issues associated with hemodialysis device design, the labeling, and the training for nocturnal home hemodialysis.

            Second, we would like you to discuss and provide recommendations on the clinical trial design to study nocturnal home hemodialysis in the home setting.

            And, finally, we would like to obtain scientific feedback, which can be used to help in device evaluation decisions on these products in the future and may lead to the future development of a guidance document for nocturnal home hemodialysis.

            Again, I thank you for coming and for providing us this feedback.  And any suggestions, recommendations, or advice you can give would be greatly appreciated.

            I would now like to introduce the next speaker, Mr. Joshua Nipper, who will be discussing with you the current conventional hemodialysis equipment that is on the market.

            CHAIRPERSON TALAMINI:  Thanks, Dr. Neuland.  I would just remind panel members, jot any questions down because after all of the FDA presentations, we will have the opportunity to ask the FDA specific questions about their presentations.

            Dr. Nipper?

            MR. NIPPER:  Thank you, Dr. Neuland.  And good morning, ladies and gentlemen.

2.  OVERVIEW OF CONVENTIONAL HEMODIALYSIS SYSTEMS

            MR. NIPPER:  My name is Joshua Nipper.  I am a biomedical engineer with the Gastroenterology and Renal Devices Branch.  And I am here today to talk to you about conventional hemodialysis delivery systems as they have been currently cleared.

            Just a brief overview of my talk.  Again, I'm going to be talking about conventional systems.  I'm going to go over some of the models and parameters that dialysis device is responsible for.  I'm going to cover some of the alarms that a standard system would have.  I'm also going to go over some of the accessory devices that would be used during a hemodialysis treatment, including water treatment systems, the hemodialysis blood tubing, remote monitoring systems, and blood access devices.

            Just a very brief disclaimer.  I'd like to point out that any examples I give in this presentation aren't an endorsement or criticism of any specific type of technology, device, or company.  It is merely to give you an idea of what has been cleared.

            I would also like to reiterate a point that Dr. Neuland made in that no devices are currently cleared for nocturnal home hemodialysis.

            As Dr. Neuland mentioned, our hemodialysis delivery systems are cleared under two different classifications:  5820 for low‑permeability systems and 5860 for high‑permeability systems.  I would just like to point out the only real difference between these two classifications for dialysis delivery systems is the fact that the high‑permeability systems have an ultrafiltration controller to regulate the amount of fluid that is removed from the patient.

            We also do have a guidance document available for manufacturers that gives them an idea of what type of information we are looking for in a 510(k) submission.  This covers the traditional bench testing that we would be expecting.

            What I have here is basically a schematic of a standard or traditional hemodialysis system.  And I'll kind of walk you through it here.  I apologize if this is rudimentary review for anyone, but the blood typically is pumped from the patient through a drip chamber by the blood pump, goes through the hemodialyzer, and returns to the patient.

            Some of the key features that most systems do have is some form of saline pump.  This can be a gravity‑fed or it can be drawn in by the blood pump or maybe a separate pump itself.  Many systems have an anticoagulant pump that administers either a bolus or a rate of anticoagulant through the entire treatment.

            There is also a series of pressure transducers here, here, and on the venous side.  Different systems measure pressure differently.  Some measure multiple sites on the arterial side.  Some measure pre‑pump, as I demonstrated here.  Some measure post‑pump.  And some don't measure on the arterial side at all.  However, all systems are responsible for monitoring the pressures during the treatment.

            On the venous side, I will use the mouse so as not to blind anyone.  We do have an air detector that monitors for air embolism and keeps any air from being administered to the patient.  You also have a venous clamp that will clamp down on the lines in case of any warning states, any severe warning states, to prevent blood from being returned to the patient in an alarm situation.

            The system I've demonstrated here is known as a three‑stream system because you have the water, acid concentrate, and bicarbonate concentrate all coming into the system, being mixed by the system to form the dialysate of the prescribed conductivity.  Systems that use this type of technology would be required to have a conductivity meter to ensure that the appropriate composition of dialysate is being created.

            The dialysate is pumped by the blood pump in a counter‑current fashion through the dialyzer and back out to a waste.  I have also demonstrated some patient fluid being removed and would like to point out there is a blood leak detector on the dialysate outline.  This blood leak detector does not detect any disconnections or any loose connections in the blood tubing.  Rather, it connects broken fibers in the dialyzer itself.  Basically it monitors for hemoglobin or any pinkness in the dialysate line and will alarm if there is any blood in that line.

            Some systems do not use the three‑stream and will use either a premixed or some type of sorbent regenerated dialysate.  If this is the case, they can get rid of the mixing system that was there because the dialysate coming in is already expected to be the appropriate conductivity and concentrations.

            These systems may or may not have a conductivity meter, pretty much depending on whether they use the premixed or a sorbent regenerated system.

            What we have here is just a blow‑up of the dialyzer itself.  You can see the blood in and out and dialysate in the counter‑current fashion.  What I would just like to point out here is that if the pressure on the blood side and the dialysate side are approximately equal, your main soluble transfer is by diffusion alone.

            However, if you do have a higher pressure on the blood side, you have fluid removed of ultrafiltrate removed.  And you do have solute transfer by convection as well.

            During a hemodialysis treatment, the machine is responsible for monitoring the parameters listed here.  You can see that it monitors the blood and dialysate flow rates.  Typically does this by measuring the pump speeds or rotation permitted of a pump.  Again, it measures pressure in several different areas:  arterial, venous, dialysate, and waste.

            From these pressures, the systems calculate the transmembrane pressure, which is essentially the difference in the blood and dialysate side.

            Patient fluid removed is also monitored using a variety of methods.  And temperatures of blood and dialysate can also be measured.

            One of the main responsibilities for these machines is alarming in any type of error state.  And hemodialysis alarms usually come in two different varieties.  The first is a yellow or caution alarm.  These are the types alarms that indicate poor trends, high pressures, things of that nature, in the machine.

            These alarms usually can be mitigated, and the treatment can be resumed.  If they aren't mitigated correctly, they can progress to a red or warning alarm.  These alarms are more serious and may or may not require the termination of a hemodialysis treatment.

            Most systems have alarms for the following parameters.  And this is not an all‑inclusive list, but you do have temperature; blood leak detector, as I mentioned before; any flow rate mismatches, you know, the pump is going either too fast or too slow.

            You have several different pressure measurements:  arterial, venous, TNP, and dialysate lines.  You also have warnings if you have excessive ultrafiltration and too much fluid is being removed.  Air embolism, one of the more important warnings, prevents air from being returned to the patient.

            Systems that do mix or regenerate their own dialysate will have a conductivity or pH alarm to ensure that dialysate is of the appropriate conductivity and concentrations.

            Some systems communicate directly with an individualized water treatment system.  These devices may have alarms that generate poor water quality or when components of the water treatment system need to be replaced or serviced.

            There are also system‑level alarms.  They include many types of hardware or software failures, power failures, and things of that nature.

            I have included vascular access disconnection with venous pressure with a question mark.  And I've done this because most modern systems rely on the venous pressure to monitor for any type of needle pull‑out or vascular disconnection.

            The problem with using this method is that if a patient's natural venous return pressure is low and you are using small bore needles with long tubing, you can create enough resistance in the system to prevent these alarms from being triggered.  And we have had documented cases of needle pull‑outs in the clinic, and the systems have filed to alarm.

            I would now like to move into some of the accessory devices.  Again, these include water treatment systems, dialysis blood tubing, monitoring systems, and blood access devices.

            As Dr. Neuland mentioned, water treatment systems are classified under 876.5665.  And we do have a guidance document available for these devices as well.  This guidance document tells manufacturers of these devices what type of information we are looking for in a standard water treatment system.

            The primary job of a water treatment system is to convert potable water meeting EPA standards to purified water that meets AAMI RD:62 standards.  The AAMI RD:62 standard is not sterile water, but it does give the maximum level of concentrations of contaminants and things like endotoxins and bacterial contaminants.

            These devices can be designed for multiple patients.  And these are the types of water treatment systems that you see in the standard clinics that deliver to multiple patients or they can be single patient, more designed to interface directly with the dialysis system.

            It's important to note that the single patient systems often have or need some type of pretreatment to remove their sediments or chloramines, things of that nature.

            This is just a basic schematic or flow chart of a standard water treatment system.  These devices are highly customizable.  So not every water treatment will have each one of these components or may have additional components as needed.  It's a little cut off here because this is the EPA potable water, and you have any amount of pretreatment, sediment filters, or chloramine reduction.

            You then go to carbon filters, which are generally in two filters in an in series known as the worker/polisher fashion.  You then would go to a reverse osmosis and a deionization.

            The system I have shown here has an altered filter, which would be responsible for removing any residual endotoxins or lower, smaller contaminants.  And then you have your AAMI water coming out.  The system I have shown also has a data out, which could be used to interface directly with hemodialysis device.

            Hemodialysis blood tubing serves the simple function of being just the conduit for blood.  It interfaces directly with the patient and the dialysis system.  It has a larger section of the tubing that is known as the blood pump segment.  And it interfaces directly with the blood pump, the device.  The blood pump segment has to be fit in very snugly within the blood pump in order to get efficient pumping.

            The important thing to note about the blood tubing is that there are multiple connection points that the patient or care‑giver would be responsible for connecting.  That includes the access, arterial and venous, any pressure transducers that have to manually be fit on the machine.  They have to make sure that the tubing is securely fit into the air detector and into a roller or peristaltic blood pump.

            Blood tubing can also be either cassette or cartridge‑based, which interfaces directly with the dialysis system.  And these types of devices greatly reduce the number of connections that the patients required for making.  Generally they only have to do arterial and venous and maybe some priming manipulations.

            I would also like to point out that kinked tubing is a significant problem, can cause hemolysis, which can lead to death.  Again, we have had documentation of kinked tubing in the clinics causing patient death.

            Remote monitoring systems are basically software that interfaces with the dialysis system.  It can be built into the machine or it can be as an accessory or an add‑on.  They're used for basic data transmission.  They connect the machine to the internet by either a modem or some type of broadband connection or a local area network.

            These devices transmit real‑time alarms and/or completed treatment data.  So they can be used to transmit things to a nurse's station or over the internet to a remote monitoring station.  They can also be used to just send final data, such as amount of blood processed or length of treatment, that type of thing.

            I would like to point out that current systems are contraindicated as the sole method of monitoring a patient during hemodialysis.  So to date, FDA has required that there be some form of partner or some other monitoring for these devices.

            Finally, I would like to mention some of the blood access devices that are available.  They generally come in three different varieties.  There's long‑term cuffed hemodialysis catheters.  These devices are available both in single and double lumen models.

            Catheters contain alternating luer lock connectors that meet ISO standards for connection to the blood tubing.  This is important because catheters are less prone to any type of disconnection and completely remove the possibility of needle disconnection.

            There is also arterio‑venous grafts, which are implanted prostheses that are designed to bypass sections of native vessels.  These devices are accessed with a fistula or graft needle.

            And, finally, there are A‑V fistulas.  I'd like to point out these are a surgical procedure and, therefore, not a device regulated by FDA.  The fistula needles are regulated by FDA, are medical devices, and they use the same luer locks as the dialysis connectors but, again, aren't prone to any type of needle pull‑out.

            This concludes my talk.  I'd now like to introduce Dr. Michael Mendelson, who will be talking to you about human factors and dialysis.

            DR. MENDELSON:  Thank you, Josh.

                 3.  HUMAN FACTORS

            DR. MENDELSON:  Good morning.  I'm Mike Mendelson.  And I'd like to thank my FDA colleagues for the opportunity to speak on the impact of human factors on the safe and effective delivery of nocturnal home hemodialysis.

            Here are the topics I am going to touch on today.  First I want to introduce you to human factors.  I'm going to talk about the magnitude of use error as a cause of adverse incidence, including deaths.

            I'll talk briefly about methods used in the field of human factors.  And I'm going to talk about based on human factors the types of problems I anticipate in the delivery in the home.  And I'm going to give a brief list of the recommendations of our Human Factors Branch.

            First a definition.  This definition of human factors was proposed by one of the pioneers in the field, Alphonse Chapanis.  I won't read it to you, but I will mention that the emphasis in our field is on the behavior of humans, their limitations.

            As I said earlier, we are interested in use error as a cause of adverse incidence.  And here is a general definition of error, as found in a popular human factors textbook.

            You will notice again the emphasis on humans and their behavior.  And you will notice the words "effectiveness and safety," which are important to us at the FDA.

            Do we know if use error is really a significant problem in medicine?  Yes, we do.  I could spend my entire talk going over data, but here just a couple of pieces that might be interesting, Lucian L. Leape, who is a noted authority at the Harvard School of Public Health, has mentioned in an interview that you could load patients into one jumbo jet every day and crash it, resulting in all of their deaths.  And at the end, you would have 120,000 deaths, which is the number of people he estimates who die from medical error.

            And a more well‑known piece of data is this.  "To Err is Human" was released by the Institute of Medicine in November of 1999.  And it blamed use error for between 44,000 and 98,000 deaths in U.S. hospitals each year.

            Now, is it necessary for manufacturers to pay attention to human factors?  Yes, it is.  The quality system regulation, which took effect in 1997, requires that user needs be included in the design of medical devices.

            And hemodialysis units would be in class II, which is one of the classes to which the supplies ‑‑ and you'll notice I'm emphasizing the words "user" and "patient" and "actual" or "simulated" conditions.  I'll touch on that a little later.

            There are three phases in the design and development process of a device where human factors is required to be included:  the input phase, where a manufacturer would take into account complaints about previous devices or what is written in the literature or information known about good design; the output phase, where a comparison is made between what is known about design needs and what the design of the device is before it's actually constructed; and, probably the most visible area where human factors is required is ‑‑ and let's call it the validation phase, where an actual production unit is tested in a realistic setting with prospective users.

            Now, what areas in the medical environment do we look at in human factors?  Of course, we look at the device as the Center for Devices, but in human factors, we look at it as a system.  We look at the environment, where things can be very challenging.  We look at the user, whose abilities may be limited; and, of course, the device.  The outcome could result in safe and effective care or what we're trying to prevent:  unsafe or ineffective care due to use error.

            Now, if human factors is successfully applied to the design and development of a device, we would like to see the following.  Now, these are all areas of what's called the use interface, which is anything a user would interact with in order to deliver care with a device.

            Of course, we expect intuitive operation, what you'd call user‑friendliness, but also applies to all of these areas:  the displays, the controls, the connections.  You have an adverse incidence associated with all of these:  effective alarms.

            I have clear and effective labeling listed here, but I want to emphasize that by applying human factors, we hope to minimize the burden on labeling as a way of ensuring safety and effectiveness.

            This is noteworthy.  When I say "safe and simple installation, repair, maintenance, and disposal," I would like to point out that users aren't only the people who are operating the device to deliver care.  They're also the people whom you may not think about as users, anyone who interacts with the device in any way.  People have died because of interaction with cleaning staff, well‑meaning relatives, and others.

            If there is only one slide that I can present in this talk, it would be this one.  It's sort of our human factors mantra that some devices can actually invite people to commit errors because of poor design.  And these design deficiencies cannot thoroughly be prevented by labeling changes.

            There is an easy‑to‑read paperback called The Design of Everyday Things written by Donald Norman, which is quite often mentioned in the field as a source of seven principles of a well‑designed interface.  I'm going to touch very briefly on the seven.

            Here is an example of a device that does not provide good visibility.  It's a patient‑controlled analgesia device, a PCA pump, which was adapted to go home with pregnant women in order to delay labor.  It was used for the delivery of terbutaline.

            Now, all controls are limited to just these two toggles and a small display, which is about one centimeter by a centimeter and a half.  Through that small window, it was necessary to view approximately 50 nested menus.  And users quite often became lost and were unable to navigate their way back and have to start the process over again.  The authors of this study, Obradovich and Woods, called this an example of dumb automation.

            The second principle, communicate clearly.  Patients have been seriously injured and killed when well‑trained operators of medical devices did not know in one case what type of radiation and what the dose of radiation was when they were operating a radiotherapy unit.

            Mappings.  Probably the simplest way to mention mappings is to ask you, have you ever burned an empty pot on the cooktop because you flipped the wrong control?  It can have a serious result in other areas, including medicine, but one noteworthy area where mappings was a problem; that is, the relationship between controls and displays or controls and what you're controlling, is in the nuclear power industry.

            Don't be arbitrary.  Be consistent.  We all expect to turn a valve counterclockwise when we want to water our lawns.  This is actually a principle that is violated occasionally in medicine.  Older style anaesthesia vaporizers used in the operating room sometimes would require opposite turning of the concentration controls in order to accomplish the same thing, which is increased concentration.

            Simplifying tasks.  When Dr. Cooper asked us to please silence our cell phones, I was amazed at the number of steps people had to take in order to accomplish that task.  Both consumer devices and medical devices are filled with unnecessary features that increase their marketability but interfere with safe and effective use.

            Here is an example of poor constraints.  This is probably the most well‑known human factors disaster in medicine.  Well‑meaning family members and cleaning staff accidentally severely burned and electrocuted some infants when they plugged the electrode pins for infant apnea monitors into either a receptacle or an extension cord.

            FDA responded by requiring something called protected pins.  These pins on the ends of the cables could only be plugged into a dedicated connector for use with a monitor.  That's an example of a principle called protective incompatibility.

            Last principle, design for error.  If there's a critical step that needs to be performed, the user should be reminded in order to make sure that he or she doesn't accidentally perform that step.  For example, when you delete a file or a folder on your computer, you are asked if you really want to do that.

            Now, when we take a medical device into the home, there are reasons why human factors are particularly critical.  Looking first at our users, working at home, they do not have the support that one would have in a hospital.  They don't have biomedical engineers available to sort out problems.  They don't have supplies and repair parts.

            The users range in education from perhaps elementary school level to doctorate.  Bear in mind that the typical reading level of a person in the United States, you will hear different numbers, but I guess a good average is the sixth grade.

            For the very reason that users need to use their medical devices, their ability to operate them is compromised, particularly in the areas of vision and sense of touch and memory.

            This is a diverse country.  Many languages other than English are spoken, another reason why dependence on labeling may not be the wisest approach.  And even because of cultural differences, people do not look at devices the same way.

            This bullet is a little complex.  I'll explain it.  There are papers that we have read that indicate that hemodialysis can be delivered at the same level of safety as in the clinical environment.  But these authors, D'Amico and Bazzi, pointed out that based on their research, the clinical level of safety found in the home is due to the fact that the healthiest patients are doing this at home and they can better withstand the adverse outcomes that sometimes occur.

            Looking at the environment, when patients go into the clinic, they no longer have to take care of other people or accomplish other things.  In the home, they still have to take care of their families.  They have to worry about children and pets.  There are many stresses that they still face in the home.  Of course, the physical environment in the home is not as well controlled as it is in a hospital.

            Things as simple as the proper placement of device are often not possible because of crowding or unstable furniture.  Voltage and grounding can be problems because of old wiring.  Temperature and humidity are not well‑controlled.  The home environment is probably more dusty.  And there is variance in water quality.  Some homes do not have municipal water.

            Now I'm going to touch on two examples of problems with hemodialysis equipment that were not found in the home environment.  The reason I'm pointing these out is to show that in the clinical environment, where you have all of this backup, the results, the outcomes of these problems were minimized.

            We have to keep in mind that problems like this are going to occur in the home.  No device is perfectly designed.  And we have to be extra vigilant in order to prevent adverse outcomes.

            In the first example, one piece of hemodialysis equipment instructed the user that in the event that transmembrane pressure automatic control failed, the user would be alerted with a code that began with the letters "FL," and it would be possible to manually control it.

            Well, the users actually tried to do this, but for three of the fault codes, it was not possible to operate the device.  The result was a recall.  And the manufacturer's solution was labeling.  And I mentioned earlier how that can be unsatisfactory.

            With a second device, if a device was not plugged into a ground fault circuit interrupter, which frequently may not be found in the home, and certain other conditions existed; that is, certain other settings with the hemodialysis unit, it resulted in overheating.  Again, the manufacturer sent an important advisory letter to the users.

            Now, what I am going to do now is list some of the concerns I have with medical devices used in the home, particularly hemodialysis units.

            Hygiene is extremely important in the setup of a hemodialysis unit.  The unit should be designed to minimize the exposure to connections where hygiene is critical.  Josh Nipper mentioned the choice of using cassettes, rather than multiple connections.  I understand that some units may require 15 connections.

            As I said before ‑‑ and I will probably say it again ‑‑ we need to minimize the need for labeling.  Also, because both health professionals and lay users can easily forget their training and develop habits that may not be desirable, manufacturers may consider the need for retraining.

            Thick manuals are difficult to deal with for common needs.  It's more advisable to rely on on‑screen help or laminated cards attached to the machine.

            It's not certain at this point how the prescribed dialysate will be prepared.  The opportunity for error if it's there should be monitored so a user does not cause injury by using an inappropriate dialysate.

            The setup and adjustment of the device should be as simple as possible.  What I mean here, "ensure safety of consumables," I'm reflecting back on infusion pumps, which are a rich source of human factors disasters in medicine.

            There was an infusion pump which had an administration set; that is, a tubing set, which included an automatic valve that would pinch shut if the tubing were removed from the infusion pump.

            In one hospital, a woman in labor had an after‑market inexpensive tubing set attached.  And when the tubing was disconnected from the infusion pump, this after‑market tubing did not have this safety device.  And she was a victim of what's called runaway infusion.  She was overdosed on the medication, and she died.

            It's critical to prime the blood lines; that is, purge them of air.  And users in the clinical environment know how to respond to symptoms of air embolism.  Now, when a patient is asleep, we need to minimize the chance of this occurring and to alert the user of this particular problem.

            Interruptions in the delivery of hemodialysis occur in the clinical environment.  And users in the clinical environment know how to respond.  We have to be certain that the home user will know how to do this also.

            I may have mentioned before the possibility of inadequate room or opportunities for proper mounting of a device.  I know of cases where dialysis units were mounted on the floor, even though the manufacturer intended to mount them on a table because the user wanted to operate them from a bed and was unable to view the display.

            I mentioned that users are typically medically compromised.  And in end‑stage renal disease, they face the comorbidities of, of course, renal failure but quite often heart failure and diabetes as well.

            Touch screens are critical.  Users should not be probing for different controls.  Quite often, they are not using their glasses at night.  And they should not have to look up error codes.  They should learn what particular failure has occurred with their device without the need for separate labeling.

            Progressive disclosure refers to the principle of allowing users to get just the right amount of information.  If they're technophobes and they only want the device to operate, then they should not have to deal with a lot of technology.

            If, however, they want a device that is very transparent; that is, a device that lets them know exactly what is going on, particularly when the device is not operating properly, they should be allowed to get more information.

            And a critical problem with hemodialysis, of course, is the risk of exsanguination.  Based on the evidence available, it seems that a single‑needle system would allow the greatest protection from that because during the phase when blood is drawn from the patient, air would be detected in the line and the system would shut off.

            Patient abilities may be lowest at the start of the session.  What I mean by that is these patients are toxic at the beginning of their hemodialysis session.  Consequently, we should not expect their abilities to be as high as even those of an average patient.

            I mentioned in the home, power supply is not as dependable.  There are medical devices which have been tricky and caused problems that were difficult to detect because error codes were generated or the device went back to defaults that may not be appropriate.

            Design in virtual guardrails.  What that refers to is the principle of protecting the user from entering dangerous settings.  When a patient is asleep, of course, we want an alarm that is loud enough or produces the kind of auditory signal appropriate for awakening them.  But that doesn't mean that once they are awakened, their abilities are the same as they are in mid‑morning.  Things should be particularly clear for them.

            The human factors engineering process is what occurs when a medical device is designed from the concept stage taking into account the human factors needs of the user.

            In order to do it properly, manufacturers need to apply human factors from the time they first conceive of a new design.  The advantages are it's very easy to design in the needs of the user.  During their iterative process, it's easy to make changes.  It won't be as necessary to stick on warnings and increase the thickness of manuals.  And research shows that users appreciate devices designed this way.  And they last longer in the market.

            As I mentioned earlier, the validation phase refers to actually testing a sample finished device.  And I want to touch on something that I feel is critical:  the difference between a usability study and a clinical trial.

            The purpose of the useability study is to make sure that the risk of dangerous use error is minimized when a device is used by typical users in a realistic setting.  This contrasts with a clinical trial, where we are trying to demonstrate that a device is operating safely and effectively when used exactly as directed because, as we know, in an actual environment, devices may not be used where or how exactly they're directed to be used.

            And in my conclusion, I have five points.  We should not assume that users are using the devices in the exact environment that they are intended for.  There are compromises in the user and the environment.

            We need to minimize the burden of training and labeling and ensuring safety and effectiveness.  We should not leave the patient or user home alone without support.  They need the same kind of support, perhaps more, than is present in a clinical environment.

            When I mentioned design needs, I mentioned complaints from users were one area where information can be found to improve design.  I think it is critical to obtain as much feedback as possible from the users of the device so the next device can be better designed.  And our bottom line I'm just going to reiterate is we want to protect our users from dangerous use error.

            Thank you.  And I would like to introduce our nephrologist, Dr. Claudia Ruiz‑Zacharek.  Thank you.

               4.  FDA PRESENTATION

            DR. RUIZ‑ZACHAREK:  Good morning.  My name is Claudia Ruiz‑Zacharek.  The reason for meeting here today is to ask for your help on nocturnal home hemodialysis devices regarding optimal device design for actual use conditions, adequate labeling for minimizing error, appropriate training for successful treatments, risk analysis to minimize unforeseen problems, and clinical study design to demonstrate the safety and effectiveness of the nocturnal dialysis device under actual use conditions.

            Currently, as has been mentioned before, there are no devices that are specifically labeled to perform nocturnal dialysis.  In my talk, I will cover some background information on this issue, specific issues on nocturnal home hemodialysis.  And I will ask for your help for an optimal design of the clinical studies to address our concerns related to this modality.

            The background information will break it down in demographics, a brief review of the literature, definitions, and nomenclature.  The most common form of renal replacement therapy in the United States is in the form of hemodialysis.  This typically takes place in a center four hours per treatment three times per week.  This is what we call conventional dialysis.

            One of the main features of conventional dialysis that I would like to point out is the presence of medical personnel.  A patient has a passive role during the treatment from measuring the patient's vital signs to accessing the access or getting the weight to initiating treatment, troubleshooting through treatment, ending the treatment.  All of this is done by medical personnel.

            This is significantly different to our subject today, which is nocturnal home hemodialysis, where the treatment is actually performed at home by the patient, typically at night and while he sleeps.  This is done in the absence of medical personnel.  And that is one of our main concerns today.  And a patient is usually the performer of the treatment.

            The interest is shifting to different treatment schedules, away from the conventional hemodialysis described previously.  And the most promising change with strongly favorable literature accumulated over the last few years is daily hemodialysis.

            One of the forms of daily hemodialysis could be short daily.  Basically it is all in the same modality of providing more frequent, more intense hemodialysis than the conventional dialysis as we know it.

            Nocturnal hemodialysis also is referred as nightly hemodialysis.  The subject today specifically is the home, the nocturnal home, hemodialysis, which could be long nocturnal or sometimes also called slow nocturnal.

            In‑center nocturnal hemodialysis is not going to be considered today as the presence of medical personnel is still available.  So it's not much of a concern to us.

            So nocturnal home hemodialysis is performed at home by the patient in the absence of medical personnel.  Frequency has been reported to range from five to seven nights a week.  The length of treatment is about six to ten hours per night depending on the prescription.

            Blood flows tend to be also slower than conventional hemodialysis with blood flows reported to be 200 to 300 mL per minutes.  Dialysate flows range from 100 to 800 mL per minute, also depending on prescription, but the usual dialysate flows are about 300 mL per minute.

            According to the United States renal data system, the prevalence of patients on hemodialysis in the United States is about 281,000 patients, out of which 843 have home dialysis, are home dialysis patients.

            According to a publication by Dr. Lockridge in 2002, there are 115 nocturnal hemodialysis patients distributed in about 13 centers in North America.

            The characteristics across this population in North America, it's about 14 percent of diabetic nephropathy.  And I would like to compare this with the U.S. demographics of about 45 percent of diabetic nephropathy in the dialysis population.  Let me point out also that diabetes is the leading cause of end‑state renal disease patients leading to the need for renal replacement.

            This is important for us to consider because diabetic nephropathy or patients with diabetic nephropathy tend to have more comorbidities than patients with other etiologies.

            So there is a merging body of evidence that more frequent hemodialysis offers superior treatment than conventional hemodialysis.  And this is just a small sample of the body of literature available, but the reason I included this slide is to point out that these studies or these publications are non‑randomized trials.  Most of them are prospective.  Some of them are retrospective.  And the majority have a small number of patients.  Most of them are single arm, although there are a few of them that have matched controls.

            Both modalities of short daily or nocturnal dialysis have been associated with significant clinical benefits, including blood pressure controls.  Most of them report or, in fact, all of them report an improvement in blood pressure control to the point where patients need to decrease a lot of the anti‑hypertensive medications they're already taking.

            Calcium‑phosphorous control is also improved.  Some of the patients actually need to reduce the phosphate binders that they are taking.  Some reports also suggest that additional phosphorus is necessary to keep a normal phosphorus level.  That is actually a new thing for hemodialysis patients in which calcium‑phosphorus control is actually very difficult to control.

            Anemia, on the other hand, has not shown significant improvement on the patients on these studies.  Some of them have actually reported increased epoetin and iron mean.

            Other benefits associated with this type of more frequent dialysis include the reduction of number and severity of dialysis symptoms and the fatigue associated with the hemodialysis.  And the time to recuperate from a dialysis treatment is also reduced.

            Improved serum albumin.  This is also a good thing.  End‑stage renal disease patients tend to have low albumin.  And that is a poor prognostic factor.

            There are also no fluid or dietary restrictions reported in certain patients.  This is also in terms of improving the quality of life for a patient.  They also have improved the sleep patterns.  Pierratos in 1999 reported a reduction or correction of sleep apnea.

            So now we will concentrate on nocturnal hemodialysis issues that we would like you to consider.  Let me just remind you things that we are going to cover are the device design and components, human factors issues, water quality, use of a partner, and remote monitoring vascular access and extracorporeal circuit connections, labeling, and lay user training.

            Let me remind you again that our concerns with a conductor in home hemodialysis compared to conventional hemodialysis is the role of the patient in conventional dialysis is actually the passive recipient of hemodialysis while in nocturnal home hemodialysis, the patient has an active role.  He is the performer of the treatment.  The patient needs to troubleshoot.  Basically while he is asleep, he needs to wake up to the alarms.

            To review what Dr. Mendelson said on human factors, the objective to include this in the device design would be to improve human performance.  And for this, the device may need to be user‑friendly, reduce the likelihood of user error and patient injury, and to reduce the burden in training and labeling.  So please keep this in mind when considering device design.

            So we would like for you to consider the following features in the device design, such as redundancy to have a back‑up system in case a safety feature fails.

            What would be the right adequacy of the alarms?  The loudness and sensitivity and ease of understanding and correction and the possibility of additional safety alarms that are not present in conventional hemodialysis.

            Another important aspect in nocturnal hemodialysis is to consider water quality.  This is because the exposure to water is a lot higher in nocturnal hemodialysis.  Compared to conventional, hemodialysis patients are exposed to about 360 liters for week in form of dialysate.  Nocturnal hemodialysis doubles this amount to 600 to sometimes more than a liter a week depending on the device and depending on the prescription and the flow of dialysate.

            Just to let you know on the review of the standard water quality is a total viable microbial count of less than 200 colony‑forming units per mL and an endotoxin concentration of less than two endotoxin units per mL.  So should higher standards of water quality be required for nocturnal hemodialysis since the exposure to the dialysate is a lot more than conventional dialysis?

            So other issues to consider in regards to the modality itself are going on with water quality concerns, the type of water systems available right now, treatment systems, is reverse osmosis, deionization, or a combination of both.  So which system would be appropriate for a patient to take home and perform nocturnal hemodialysis?

            The water source on how to manage changes in the quality of municipal ‑‑ in the case of municipal water suppliers, another issue that is concerning to us is monitoring.  Raija in 2003 suggested that nocturnal dialysis could be done without a partner.  His study was 59 patients, prospective study.  Out of these 59 patients, 13 of them were actually in nocturnal dialysis without assistance.

            The conclusion was that there were no increased technical difficulties or increased alarms or safety issues in regards to this.  He based this conclusion in about 115 patient‑month experiences.

            In center hemodialysis, there is constant monitoring by medical person.  In home hemodialysis, we understand by that a patient that is awake.  So it's also not too concerning in regards to home dialysis.

            The London daily nocturnal hemodialysis study in 2003 suggests or the conclusion was that monitoring is essential for the initial three months of nocturnal hemodialysis therapy until the hemodialysis team is convinced the patient is stable and compliant.

            Now, they base this ‑‑ again, as I mentioned earlier, this is a prospective comparative non‑randomized study of about 23 patients with 22 matched controls.  However, only 14 of them were in nocturnal hemodialysis.  The rest were in daily dialysis.

            Their treatments ranged from 13 to 602 treatments based on these 14 patients.  There were reported in the time of the study about 4,000 patient nights, out of which there were about 5,000 alarms, reported alarms.  These resulted in calls, but most of the alarms were due to either slow response by the patient to the alarms or non‑response.

            Most of the alarms were due to arterial or venous pressure alarms due to the patient obstructing the blood tubing.  So there were not emergencies.  There was no need to call the designated contact person or emergency services.

            The average number of alarms per night decreased significantly over time.  So each progressive decrease from month 3 through month 18 was statistically significantly lower than the value at one month.  And that's why they concluded at three months.  After three months, monitoring may not be necessary.

            Our discussion wouldn't be complete without talking about vascular access.  Let me just review what is available right now for the hemodialysis and conventional in‑center or home dialysis.

            Long‑term cuffed catheters, it's actually very convenient in terms of using it immediately after placement.  But there is a disadvantage in the increase of thrombosis, increase of clotting, increase of infection.  So this is actually not recommended by the DOQI guidelines.

            The next vascular access available is a synthetic graft.  And this is a surgical graft.  Usually it's PTFE.  And it's usually available within weeks of placement.  It's superior.  It has a superior performance than the catheter but not as good as arterio‑venous fistulas.  And that's actually the recommendations by DOQI guidance.

            A‑V fistulas need some time to mature, sometimes up to three months and sometimes recommended to let mature longer depending on the surgeon.  But they have a much better rate, a superior rate of low infections compared to the other, to vascular access.

            The concern with the A‑V fistulas and daily dialysis, though, is that with increased punctures, would that affect the life of the fistula?  Quintaliani in 2000 reported an observational study of 24 patients in daily dialysis, follow‑up of 3.6 years.  They concluded that the life of the fistula is really not affected by the daily puncture, requiring daily dialysis.

            However, they used dual technique, rather than the single button hole technique.  Also, they were using different sites for puncture, rather than the same puncture, as the button hole technique does.

            Other things that we would like you to consider when it comes to nocturnal dialysis devices are vascular access location.  Would these have an effect in terms of disconnections, infections, or thrombosis?

            Connection to the device and the use of locking devices or interlinked devices, the self‑cannulation technique, the training that gets involved for the patient to self‑cannulate, full detection locking devices, as I mentioned earlier, with a connection to the device on fluid detection alarms to detect either blood or dialysate leaks.

            Pierratos in one of his papers and many others have reported the use of enuresis alarms to detect blood leaks or dialysate leaks.  In the same way, moisture sensors should be included or discussed and the technique of using single versus dual needle technique.  A lot of literature supports the use of the button hole technique because of the ease of doing it.

            So these are additional features which we think should be addressed by the device manufacturer as either part of the device or part of labeling.  Now, labeling is the operator manual that includes the warning, precautions, device specifications, instruction for maintenance, cleaning, and disinfection.

            This includes the physician's instruction for use, which it should also include relevant data from clinical studies and instructions for use in the caring of the device.

            This is basically the same thing as the patient instruction for use with the main difference and important to remember is that this is written for a person with no medical training or we should assume that they have no medical training.

            So training is defined as the teaching provided by the manufacturer.  So the medical expert can train the lay user.  And the lay user is able to use the device successfully.

            So the aim for the training should be to be able to conduct safe and effective nocturnal hemodialysis treatments.  The length of training has been reported to be from two to eight weeks.

            Agar in Australia in 2003 and Leitch in the London daily in nocturnal hemodialysis, NHD, this also depends on the past experience of the patient and prior exposure to home dialysis or hemodialysis in center or at home.

            Let me add this now.  For the purpose of testing the adequacy of the training, it would be ideal to have patients who have no exposure to hemodialysis prior to entering the trial.  The main reason for it is to test the adequacy of the training.

            So what we would like to see at the completion of the training is an appropriate use of the hemodialysis device, interpretation and use of safety features and accessory hemodialysis treatment itself, on evidence or a test to confirm the adequacy of the training.

            So, yet, we would like to minimize the burden on labeling and training.  So we have come up with a list of risk analysis that we would like you to consider and see the completeness of this list.  And maybe there are a few things that we might be missing.

            Inadvertent disconnections, blood loss from increased frequency of treatment, potential increased rate of vascular access infection, and the psychological effects.

            Let me tell you that none of this has actually been reported in the literature with the exception of blood loss.  And that might be one of the reasons why anemia has not shown a significant increase or a significant change, a clinically significant change in the daily treatments.

            Inadvertent disconnections have not been reported.  However, once the treatment goes into the broader marketing population, would the adverse events reported be still applicable to the wider population and the marketing population?  Would that be still reflective of the results of the trials and the publications?

            So disconnection could, as we all well know, be life‑threatening and if the patient is asleep may not even realize that he is bleeding.

            The potential increased rate of vascular access infections again has not been documented.  And the psychological effects could be positive or negative.  That is another thing that we need to assess.

            So now we are going to go into the clinical studies.  And I will divide this into the purpose of what we think a clinical study is needed, patient selection, inclusion, and exclusion criteria, to ask for your help, to ask for your help in the optimal design to test the device in actual use conditions.

            So the purpose is to demonstrate the safety and effectiveness of the nocturnal hemodialysis device under actual use conditions.  This is not intended to evaluate the long‑term morbidity and mortality of nocturnal hemodialysis as a therapeutic modality compared to conventional hemodialysis.

            So our concerns are the patient selection for trial.  Can this be reflective of the patient selection when the device goes for marketing?  Again, for the trial, in order to test the adequacy of training, we would like patients that do not have prior training or prior exposure to home dialysis.  And this is just to test the training.  Once it goes to marketing, of course, patients with prior experience would probably be preferable.

            We would like to make sure that the patient is able to perform the entire treatment and that the patient is actually able to attend the alarms.  So, again, the adequacy of the alarms becomes an issue.

            In terms of patient selection, there is different literature.  And let me go through these three citations.  Agar in 2003 ‑‑ he's from Australia.  He included in his prospective study 16 patients.  The selection criteria was they must be clinically stable for more than three months on hemodialysis.  They should be psychologically sound, technically adept, stable and have a supportive home and a history of compliance.  So this is already excluding a large part of the population.

            Alloati in Italy in 2002 also had a non‑randomized prospective study, including 18 patients.  Let me point out here the important thing is the diabetic nephropathy.  It's only one patient of 18 patients.  And the U.S. population, as I mentioned earlier, is 45 percent.  So, again, this type of selection is not representative of the marketing population in the U.S.

            Covic, this is a retrospective observational study of 286 patients in the U.K.  He basically started his analysis since 1960.  Initially the patients in nocturnal dialysis excluded older and frailer patients, exclusively excluded patients with diabetes, cardiac failure, and multiple myeloma, although that plan has changed.  And now they are including about 33 percent of diabetic patients.

            So should the following be incorporated into the selection criteria?  And that is availability of a partner or the possibility of monitoring.  Patient compliance, psychological well‑being, and home environment, which most of these items are already addressed by Medicare, to have an adequate water supply, adequate sewage, adequate electricity, adequate space, and social interaction.

            So once patient selection is discussed in regards to an optimal study design, what would the clinical endpoints be the best to assess this study design in terms of effectiveness and safety and adverse events?

            In addition to what is conventional dialysis, should a control group be necessary?  And should it be randomized or should the patient be their own control?  The length of follow‑up is also an issue and the sample size.

            Let me just briefly tell you that on home hemodialysis studies, what we have done is a small number of patients, less than 30.  The time frames are an observational period during in center; then the use in center of the device in question; then a wash‑out period, where the patient doesn't have this device.  And then they go home with the device, and there is continuation of the study.

            So these are treatment‑related issues include the dialytic composition, which may change with more intense and more frequent hemodialysis, the additives on the possibility of using phosphate, as reported by several of the clinical studies I've mentioned, administration of anticoagulation and the type of anticoagulation, what kind of dialyzer to use, and whether monitoring should be encouraged or required, and what type of monitoring, vascular access and what type of access would give the safest connection to the device and the practice of reuse.

            In conclusion, we would like your help for eventually creating a guidance document in creating an optimal device design for actual use conditions, adequate labeling to minimize error, appropriate training for successful treatments, risk analysis to minimize unforeseen problems, and clinical study design to demonstrate safety and effectiveness of the device itself under actual use conditions.

            Thank you.

            CHAIRPERSON TALAMINI:  Thank you very much for very clear presentations from all of the FDA personnel.

            I would like to now offer the panel the opportunity to question the FDA regarding these presentations.  In the interest of a smooth day, I would encourage you to not ask questions that we know are going to be discussed or are already within the scope of the questions that we are going to be going through as a panel, but try to ask questions that are either outside of that scope or are clarification questions about that which has been presented.

            So, with that caveat, are there any questions?  Dr. Blagg?

            5.  QUESTIONS FOR THE PANEL

            DR. BLAGG:  I would like to raise an issue about the definition of nocturnal home hemodialysis because there are patients ‑‑

            CHAIRPERSON TALAMINI:  A little bit closer to the microphone, sir, if you could.  Thank you.

            DR. BLAGG:  Okay.  I would like to raise a question about the definition of nocturnal home hemodialysis, which in this case is limited to a treatment frequency of five to seven days a week.  There are patients in this country and elsewhere doing it three times a week or alternate nights.

            And I think that the difficulty comes.  What we need to do is define nocturnal dialysis as dialysis which is done overnight to any frequency and define five or more times a week dialysis as nightly dialysis to clarify the difference.

            CHAIRPERSON TALAMINI:  So noted.  And I think through the day today, it appears to me as a non‑nephrologist, that there really are two global issues that we need to deal with.  One is the fact that this is being done at home, as opposed to in a center.  And the second is the issue of frequency and whether it's better for patients to have more frequent dialysis.

            But with that question, do any of the FDA speakers have a response to the issue of the definition?

            DR. MITCHELL:  Hi.  My name is Dr. Dianne Mitchell.  And I'm the chief medical officer for this division.

            I think those two definitions are fine.  What we need to keep in mind, though, is that we will be asking you to give us guidance for both types of dialysis you identify.

            CHAIRPERSON TALAMINI:  Other questions for the panel?  Dr. Sadler?

            DR. SADLER:  I would just like to make the point that a couple of presenters considered the material provided by the manufacturer to be the training or a large component thereof.  And the training actually is the responsibility of the medical team in the dialysis facility doing the training.

            And the manufacturers will provide background information, but my experience would say that the further the manufacturers go with the details of clinical steps to be taken by the team, the more they see themselves exposed to liability.  And so they're fairly reluctant to do that.

            And I don't think we should assign them this responsibility except for characterizing their product, its operation, and the rational hazards associated with it.

            CHAIRPERSON TALAMINI:  So noted.  Dr. Hoy?

            DR. HOY:  Dr. Mendelson said he felt that patients might actually be more toxic pre‑nocturnal dialysis than patients in the center.  In fact, that's not.  Perhaps I misunderstood what you said, that they're certainly not more toxic.  They're well‑dialyzed, much better dialyzed than our own center patients.  And their cognitive function is markedly improved.

            DR. MENDELSON:  Yes.  I think they didn't explain it correctly.  At the beginning of each session at night, they are more toxic than they would be at the end of the previous session.  I wasn't comparing the clinical setting to the home setting.

            I was simply saying their abilities may be compromised at the beginning of a treatment session relative to the end of the treatment session.

            DR. HOY:  I certainly ‑‑

            DR. MENDELSON:  Or relative to their recovery from the treatment session.

            DR. HOY:  In my experience, that is not the case.  These patients don't seem to show a major cognitive difference between the morning and the evenings because they are dialyzed 48 hours a week.  So I'm not sure that that is a real concern.

            DR. MENDELSON:  Okay.  Thank you.

            CHAIRPERSON TALAMINI:  Other questions?  Dr. Lockridge?

            DR. LOCKRIDGE:  Yes.  On the issue of the human factor ‑‑ and you cited the part about the hospital error.  There wasn't anything discussed about the fact that when you empower a patient, they become much more adept and much more responsible for their care.

            The factors of the 98,000 deaths per year in hospitals are based purely on the fact that personnel in the hospital are making mistakes.  And I think we have to be very aware that there is something that is very positive about empowerment of patients.

            DR. MENDELSON:  Yes, there is.  Now, we do know that error involving medical devices probably plays a ‑‑ it's not even half of those reported deaths.  It accounts for a minority of those deaths because so many of these deaths involve medication errors.

            But in looking at home use devices, we feel that the compromises of the patient in the environment need to be considered to the point where they play a significant role, just as the fact that although in the clinical role you'll have a highly trained, experienced person delivering care, that person may be overtired or may be distracted because of many different responsibilities or many devices, which pose conflicting operation methods.

            I realize there are patients who are so adept at use of their home use devices they sort of fall into a pattern.  Actually, some patients fall in love with their home devices, and they will not part with them, even though there are better designed, safer, perhaps even more friendly devices on the market.  I recognize that.

            CHAIRPERSON TALAMINI:  Dr. Aranoff?

            DR. ARANOFF:  I would like to come back to something that Dr. Blagg brought up, and that is the definition of this therapy because, especially if, as Dr. Neuland pointed out, we want to perhaps come up with a new document, advisory document, I think we need to think differently about the way that this therapy is defined because if we define it on the basis of blood flow, 200 versus 400, that is the wrong pathway because you can exsanguinate very quickly at 200 milliliters a minute, just as you can at 600 milliliters a minute, or if we define it by starting the machine after sundown versus at sunrise, that's also the wrong pathway.

            This therapy is fundamentally different than in‑center hemodialysis or previous home hemodialysis in the way that it is monitored, not how fast the blood goes or the dialysate flow rate or the time of day or the duration of the therapy.  It has to do with the way the treatment is monitored at home.

            Traditional home hemodialysis, not too differently than traditional in‑center hemodialysis, is monitored by more than one individually usually, by somebody who is awake and alert and can respond to traditional kinds of alarms and so forth.  This form of therapy is defined differently in that patients will be unconscious when the therapy is performed, largely.  And, therefore, the whole concept of how we view this therapy needs to be different than simply the superficial markers of blood flow, dialysate flow rate, duration, or time of day.

            Bob?

            DR. LOCKRIDGE:  I agree with you, but I think that the lower the blood flow rate, limiting the ultrafiltration, and going slow and long impacts on all the things that you're trying to monitor that we keep hearing are very bad in center.

            When you run somebody at a blood flow rate of 500, dialysis flow rate at 600, and you pull 6 kilos in the first 3 hours, they're going to all get sick.  The autonomic system is cranked up, and they get sick.  When you do at a 2 to 3 hundred ratio and you only pull 300 cc's an hour, they never get sick.  So I think it's clearly a difference.

            CHAIRPERSON TALAMINI:  Hold it.  Time out.  This is all valuable stuff, but I want to make sure that we ask the questions of the FDA that we want to ask them during this time period.  So this is all very valuable, but I want to make sure we take this time for clarifications or things that are outside the scope of the questions that we will already be discussing.

            Yes?

            DR. MITCHELL:  This is Dr. Dianne Mitchell again.  I was just wondering if I could make a comment in response to Dr. Sadler's comment about device training versus clinic training.

            We at the FDA do recognize that there is a difference between those two.  I think there was a reference to it in one of Dr. Zacharek's slides.

            The purpose of the discussion today regarding training is indeed for you to address what you think is appropriate for the device manufacturers to do in conjunction with training on their device.

            Thank you.

            CHAIRPERSON TALAMINI:  Thank you.

            Other questions for the FDA?  Dr. Weinger?

            DR. WEINGER:  Yes.  I think many of the issues with regard to the use of these by patients is going to depend somewhat on what percentage of this 280,000 patients ultimately would be subject to this.

            So are there any estimates from the FDA of what kind of patient population?  Right now there are 100 or 150 or something.  When these devices become available, how many are going to be on this?

            DR. RUIZ‑ZACHAREK:  Yes.  It's hard to predict.  Right now I wanted to mention that there is a very limited number of patients.  And these are highly selective in each trial.  When it goes to the marketing, that selection may or may not still be there.

            What I wanted to point out is the patients in the general population on dialysis in the United States have a lot more comorbidity than the patients included in these studies. With the impact of a device going into the market, specifically labeled for nocturnal home hemodialysis, we don't know.

            But we want to make sure we take the safety precautions to deliver safe and effective treatment among that population.

            CHAIRPERSON TALAMINI:  Thank you.

            Who has not?  I think let's go back to Dr. Hoy.

            DR. HOY:  Just a comment, Dr. Ruiz.  I think there has been an evolution of the selection of these patients, which isn't perhaps apparent in some of the published literature since it's usually two to three years behind what actually has occurred.

            In our center, we have trained 49 patients.  Seventeen of them are diabetics.  They have all of the diabetic complications, including peripheral neuropathy and difficulty with manual dexterity, blindness.  We have one legally blind patient hose husband is deaf.  The two of them can run this machine at home.  We have a matched set there.

            The patients, 41 out of 49 of our patients were over 200 pounds.  These are the hardest patients to dialyze adequately.  They were all excluded from the hemo study.  These are patients that we're using nocturnal dialysis as salvage therapy for.  They're the ones who have the unquenchable thirst, you know, who gain six to eight kilos between treatments and cardiomyopathies with ejection fractions in 5 of our patients with less than 25 percent.

            So I think that you have to be careful concluding that we are self‑selecting these patients.  There is only one absolute criterion that our patients have to have.  And that is they have to want to learn this.  And we have not yet found a single patient we could not train if they wanted to learn it.

            And our most difficult patient was a child psychiatrist.  She took 79 days to learn how to do this.

            DR. RUIZ‑ZACHAREK:  Yes.  As you said, the trends have changed.  In fact, most of the benefits or the people, some of the patients that most benefit from this type of modality are overweight patients, patients with sleep apnea.  Pierratos have included patients with overweight that demand a lot more dialysis than the conventional hemodialysis can provide.

            That's why sometimes diatolic flows go up to 800.  Sometimes the blood flows can go higher than what nocturnal dialysis tends to have, slower than conventional.

            So yes, the benefits, you know nobody can deny that the benefits will, the suggested benefits by the published literature, the patients that most will benefit from that are the patients that may have already been excluded by other clinical trials or by other publications.  But we cannot deny that they're probably going to be the ones that most benefit from it.

            CHAIRPERSON TALAMINI:  I would warn the panel our time for asking questions is running very quickly.  So succinct if you can, please?  Dr. Lockridge?  We'll just go around one more time, and then we'll ‑‑

            DR. LOCKRIDGE:  This is to Dr. Neuland.  It really comes out that the guidance document is one of the things that appears to be that we are going to try to develop here today.  Is there any direction that you would give to us as panelists of how to help you do that?  I mean, I think that's a critical issue that we are trying to develop.  Tell us what are our directions or try to educate us a little bit more about that.

            DR. NEULAND:  Well, I think that the two main issues have already been brought out, and that is whether you have any recommendations on design of a device that would make it safer for the medical community.

            Because in the guidance document, our overall plan is to put in issues that need to be addressed in a submission to the FDA for marketing clearance.  And that would include anything to do with specific issues related to the design of the device or with the clinical trials that we hope to have done to support these marketing clearances.

            DR. AFIFI:  Dr. Talamini?

            DR. NEULAND:  Plus, you're going to be addressing these, I think, ‑‑

            CHAIRPERSON TALAMINI:  I'm going to go around real quick.

            DR. NEULAND:  ‑‑ in the questions that you're going to cover today.

            CHAIRPERSON TALAMINI:  Dr. Blagg?

            DR. BLAGG:  I don't know whether we're going to cover more about the question of patient selection later on.  Are we or not because I would like ‑‑

            CHAIRPERSON TALAMINI:  Yes.

            DR. BLAGG:  Okay.

            CHAIRPERSON TALAMINI:  It is certainly within the questions.

            DR. BLAGG:  I will save the comment for later.

            CHAIRPERSON TALAMINI:  Yes, Dr. Gibson?

            DR. GIBSON:  Just a clarification.  Josh Nipper said that FDA labels current systems as contraindicated as the sole method of monitoring patients during hemodialysis.  Is that aimed at nocturnal hemodialysis?  Is the FDA now regulating that in some way or is this ‑‑

            MR. NIPPER:  Well, that was in relation to any remote monitoring systems.  That was part of our initial concern that the current information we had was limited to in‑center type devices.  So we were concerned when we see devices that connect to the internet and can transmit data remotely.  That was one of our concerns for nocturnal or just general home use.

            So both of those were contraindicated or all of the devices have been contraindicated as the sole method.  So as long as there is a partner there, we kind of are hands off on whether a device goes home under practice of medicine, but it is contraindicated without a partner.

            DR. GIBSON:  So you label it for that.  And if someone sends a patient home now without a partner, you would if you found out about it take some steps?

            MR. NIPPER:  Well, I mean, that would be right there under practice of medicine.  A physician can send any device home that they want to.  It does provide guidance to the physician that we do not think that the remote monitoring should be the only way to monitor the patient.

            CHAIRPERSON TALAMINI:  Perhaps I could ask somebody from the FDA at this juncture to clarify for us the practice of medicine issue versus FDA guidelines because that is going to play so vitally into the discussions today.

            MS. BROGDON:  The guidance that we will eventually be developing is guidance for the manufacturers, what they should submit to us in order to get the labeling and the clearances that they desire.  Regardless what gets cleared, it's the physician's right to use a device as he sees fit in his professional judgment.  So we regulate the manufacturers on what information and what devices they can supply.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Afifi?

            DR. AFIFI:  I have a fundamental question about policy that relates to a 510(k) pre‑market notification.  The question is probably for Dr. Neuland.

            When we are required to see whether something is demonstrated to be substantially equivalent to an existing device, the process is similar to an evaluation process.  And the question, then, are we evaluating just the process; in other words, the delivery of whatever the product is supposed to deliver, or also the outcome?

            And that, then, relates to a slid that Dr. Ruiz‑Zacharek presented, where she said that we need to demonstrate the safety and effectiveness of NHD but not evaluate the long‑term morbidity and mortality of NHD as compared to conventional procedures.

            DR. NEULAND:  Okay.  Hemodialysis equipment currently and all of the other medical device components are regulated under the 510(k) process.

            When those products are cleared, ‑‑ they're not actually approved; they're cleared ‑‑ we are determining whether the product is as safe and as effective as the other devices that are currently on the market.  It's more of a comparison.

            What Dr. Zacharek was trying to say in that slide ‑‑ and she can add to it if she would like.  Basically we did not feel that this equipment, that it was FDA's role to determine whether nocturnal dialysis as a modality in general was better than or worse than current conventional practice of medicine with the equipment that is currently used.  We were looking at whether the equipment as designed can do a safe and effective treatment basically or series of treatments for the patient.

            So that's why the studies for morbidity and mortality to us are more of a broad scope study, one the NIH might sponsor or something like that nature, not an individual manufacturer would have to demonstrate and prove before their product could go to market.

            DR. AFIFI:  I see.  I see.  And with that, Dr. Talamini, I have another point, but I think it can wait for the later discussion.  And that is the role of redundancy in the design.  That is something that can wait.  Is that correct?

            CHAIRPERSON TALAMINI:  Yes.  We will certainly be covering that later.

            Dr. Sadler?

            DR. SADLER:  I think I am remiss in not commending all of the speakers for being lucid and clear and making good presentations.  But I do have a few points to bring up that I think are important.

            I think this is still intermittent hemodialysis therapy.  It's a different application.  And it's a different site in some ways.  But it is not a completely different therapy.  It is possibly improved, but it is still the same thing.

            The point in that is that if someone should come forward and ask to have their device labeled for nocturnal hemodialysis, you can be sure that almost everybody who has one that's already approved for other hemodialysis applications will be in the same queue very shortly.  So we're still talking about hemodialysis and what goes on.

            When Dr. Mendelson talked about alarms, he talked about the sound and what goes on with them, but we have to realize that probably the worst thing about alarms is false alarms.  It frightens patients.  It frustrates patients.  And it causes them not to respond appropriately to alarms.

            As a former chair of the AAMI committee, I would be remiss if I didn't point out that in our comments about water, you must recognize that the increasing water quality requirements have nothing to do with problem‑solving.

            There is no one who has ever been harmed by water that met the original standard enunciated in 1982.  The standard has risen as our capability to purify water and to monitor what we have accomplished has increased.

            So that there really hasn't been any rational force for problem‑solving.  It's just a matter of admiring our capacity and responding to that.

            And the final point, I want to remind you that passive restraint levels of safety are still not absolute.  If you have an air bag, you still have to learn to wear your seat belt.  You have to learn to drive.  We have to build safer highways and avoid distractions.  There is no single passive restraint that is going to create safety in an automobile or when you are attached to a hemodialysis machine.

            And one last point, it's always been my experience since Dr. Blagg and I both did overnight hemodialysis with our patients almost 40 years ago that the patients dialyzing at home asleep had the same kind of sensitivity that the mother of a young child.  When the alarm goes off, they rouse, much better than someone who doesn't recognize that it is important to them.

            CHAIRPERSON TALAMINI:  Comment?  You can respond as if it were a question.

            DR. MENDELSON:  The time I had for my presentation was quite limited.  And one of the points I wanted to make about alarms ‑‑ well, several ‑‑ was it's my understanding that the false alarms outnumber the legitimate alarms.

            And what I am concerned about in the home at night is the effect on the patients psychologically and physiologically if he or she is roused over and over and over again.

            We know that interrupted sleep has an adverse effect on the patient's health.  And whether this problem is limited to the patient's health or the patient's health and their ability to operate the therapy, I'm wondering if people knowledgeable about the physiology of sleep can address that.

            Thank you.

            CHAIRPERSON TALAMINI:  Thank you.  So I think we will close the questions to the FDA and again thank the FDA presenters for very clear presentations.

            What I would like to do is call for a ten‑minute break, but let me just frame the rest of the day quickly before we do that.  We have a Herculean task to cover these eight questions.  We could probably spend one day on each question.  So it's going to be absolutely vital that we move along quickly.  And hopefully everybody will not hate me by the end of the day in driving the committee towards doing so.

            But I would like to take a quick ten‑minute break.  My objective is going to be to do two questions and then break for lunch.  I would remind the committee not to discuss the issues that are at play here outside of the room, as all of the proceedings here do need to be public and on the microphones.

            So we will reconvene in exactly ten minutes, at 11:34.  Thank you.

            (Whereupon, the foregoing matter went off the record at 11:24 a.m. and went back on the record at 11:34 a.m.)

            CHAIRPERSON TALAMINI:  I would like to call the panel back to order.  Dr. Ruiz asked for the opportunity to respond to some of the comments and questions.  So, Dr. Ruiz, if you could do so now?  Sixty seconds, if possible, please.  You could use this one over here perhaps.

            DR. RUIZ‑ZACHAREK:  I would like to make some comments on the concerns you raised about nocturnal dialysis and hemodialysis being basically the same modality.  It could be.  It's basically hemodialysis.

            And, as Dr. Pierratos said in one of these publications, any type of conventional hemodialysis machines could be used for nocturnal hemodialysis.

            The concern comes when we send these devices home with a patient that may or may not have the experience or the appropriate training.  So that's our concern.  It's not the fact that they have hemodialysis itself because this is the same modality, but, as somebody else pointed out earlier, it is the monitoring.  That's significantly different from conventional dialysis and home nocturnal dialysis.

            The other thing that I would like to point out is on the quality of the water.  Before all of these standards, actually, patients were being dialyzed with a little bit higher content of aluminum.  And they were having a lot of problems.

            So the question is now we haven't had any problems with the new standards, but would this be still applicable for patients that are exposed to a much higher volume of water in terms of dialysis or do we need to regulate that or do we need to change the quality in the water in former dialysis for patients that are undergoing nocturnal dialysis?

            And those are the issues that we would like to raise and have you discuss on them.

            CHAIRPERSON TALAMINI:  Thank you.

            DR. RUIZ‑ZACHAREK:  Thank you.

            CHAIRPERSON TALAMINI:  So all of those are points that we will be getting to.  So this meeting now continues with the panel discussion portion of the meeting.  Although this portion of the meeting is open to public observation, public attendees may not participate except at the specific request of the panel.

            Now, again, some comments on our task today.  We have eight questions, which all of the panel members had the opportunity to study prior to the panel meeting today.  And if you take the time remaining, the number of panel members, and the questions, that leaves about a minute and a half to two minutes per panel member per question.

            And, again, we could probably spend a day on each of these questions.  So I'm going to have to be very tough with folks, unfortunately.

            I would like to begin, however, by offering the opportunity for each panel member to offer any general comments.  I would ask you please to limit them to about 60 seconds, really the most important points.

            And I think since we have already had a fair bit away from the right side of the table, I'm going to start over on the left side.  If you don't have any general comments, feel very free to pass.  We would love you for it, but we would also love to hear your comments.

            Ms. Moore?

            MS. MOORE:  Yes.  My only comment is that in reading the material, my major concern was that we have this technology, which is customarily used in a hospital setting or a clinic of some sort, with professionals in charge.  And I was still concerned.

            And I was very much interested in the doctors over there who mentioned the fact, you know, that they had had experience with patients who were on nocturnal programs, but I know that we are going to have patients with varying amounts of abilities, ages, all kinds of conditions that may mitigate against some of these people using the technology to the best that the technology could offer.

            And so I guess my major concern was some kind of safeguards so that these people who may not be as educationally able or the elderly persons or people of that sort would be able to use this treatment to the best of his or her ability.

            And one other thing, the contact, you know, what happens if things go wrong.  I was looking to see if there was any consideration given to making contact with someone without being automated.  In other words, if there is an emergency, who do you contact and will not get an automated message but you would get a person?

            CHAIRPERSON TALAMINI:  Thank you.  You mean not get caught in a voice mail trap.

            Dr. Duffell, your general comments?

            DR. DUFFELL:  Yes.  I think one thing that is going to be important when considering the questions today, which has already come up briefly in Dr. Neuland's remarks, is about the design of these products.

            CHAIRPERSON TALAMINI:  A little closer, please.

            DR. DUFFELL:  And we need to be mindful in our deliberations that we probably shouldn't be trying to actually design products here at this panel but, instead, raising the issues which the manufacturers need to address in their designs because they may have creative ways that we haven't considered.

            And then my second point is also to keep in mind that it also seems like from one of the prior FDA speakers is that the issue that we are really looking at here, in addition to safety and effectiveness of the products, is also the useability of the products.

            I think that one of the speakers brought to mind the difference between a clinical trial and a useability study.  So I think what we are really looking at here is making sure that these things are useable more than they are effective.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Schulman, general comments?  Sixty seconds, please, if possible, or we can come back to you.

            DR. SCHULMAN:  If you can come back, that would be good.

            CHAIRPERSON TALAMINI:  Sure.

            Dr. Sadler?

            DR. SADLER:  I think I made mine, but in response to Dr. Ruiz, I do believe that if you look at the water standards, the 1982 standard was fairly complex and comprehensive.  And what we have done is to incrementally approve it.

            But, as I say, there hasn't been a problem to force those increments.  We have only improved it as our ability to do so occurred.  And I guess in admiration of our accomplishments, we raised the standard.

            You said in your comments that you worry about the training.  And I'm afraid that FDA has to relax about that because you can't regulate it.  And we just have to get that done.  And that becomes our responsibility in our field.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Sadler.

            Dr. Afifi?

            DR. AFIFI:  I have no general comments at the moment.  I will have several relating to design especially later on.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Aranoff?

            DR. ARANOFF:  One additional thing that we haven't mentioned is that although there are no predicate devices for nocturnal home hemodialysis, there is a great deal known about nocturnal peritoneal dialysis.

            And some of the concepts, although the consequences of failure of the treatment or danger of the treatments are very different with peritoneal dialysis and less so, but some of the concepts about design of monitoring and so forth might be applied from that, might be drawn from that.

            CHAIRPERSON TALAMINI:  Great.  Thank you.

            Dr. Kalota?

            DR. KALOTA:  I think we need to keep in mind that this technology is never going to be for everyone.  And we can't try and make it for everyone.  There are still physicians who are going to make the decision whether a patient they feel is competent or not and then in the training, are they still competent to do it and that we need to make it safe for those who both patient and physician feel are appropriate to do so.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Gibson?

            DR. GIBSON:  Just very briefly.  In most of the material that we have when psychological is mentioned, it's mentioned in terms of risk and monitoring for detrimental effects.

            I would like to congratulate Dr. Ruiz‑Zacharek for mentioning the possible benefits of home dialysis, which may be considerable and which I don't think we should lose sight of during the deliberations today.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Weinger is our panel expert on human factors.  So I would void my 60 seconds and offer you a few minutes if you need them for your introductory comments.

            DR. WEINGER:  I'll just make two short comments.  First, the data that I have reviewed on this shows tremendous promise.  And I think that this methodology and the devices associated with it are going to be a superior method for a subset of patients.

            But the data that to date is based on experience of individuals who have many years who have evolved a practice ‑‑ and the reality in the future with these devices is they are going to be aggressively potentially marketed to clinicians who don't have those years of experience.  And they are going to be trying it for the first time.

            Our obligation as advisers to the FDA is to promote patient safety.  And I think that we need to develop guidance, strong guidance, to industry about the design indications and training that maximize safety under those circumstances.

            I have to disagree with Dr. Sadler about training.  In fact, training is well within the FDA's purview.  And, in fact, carotid stents is a good example, which there are now regulations that say every user, in this case clinicians, rather than patients, every user, must undergo rigorous simulation‑based training before they can use the device.  And it's within the FDA's purview to make similar requirements for patients, not that I am necessarily advocating that at this time.

            The last point I wanted to make in response I think to Dr. Lockridge was that clinicians make user errors.  In fact, there's a study by Baden in 2001 involving all of France where they showed the use errors were the most frequent cause of all medical device incidents and the second most common cause of deaths associated with medical devices.

            Now, these are experienced, highly trained, highly educated individuals using a full range of medical devices.  Now we're talking about patients, who are going to be far less educated and perhaps less well‑trained, but the issue is how trained should they be, going to be using the complex device.

            So use error is clearly a problem.  And we have to help industry and the clinicians deal with it.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Gillespie?

            DR. GILLESPIE:  I would just like to say I hope people will keep in mind as we go through this the potential pediatric applications of these machines.  And I hope that the pediatric uses will be considered part of the kind of overall design of the machine, part of the initial labeling, and not an afterthought or an off‑label use because there are a number of good reasons why children would be excellent candidates, in some cases even better than adult candidates for this.  And certainly there is a question all the way back to the first home nocturnal dialysis patient.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Gillespie.

            Dr. Blagg?

            DR. BLAGG:  I would just like to remind you as we talk about these machines, that the first home dialysis machine developed in 1964 was the first single patient machine.  It was the first machine which was monitored.  And it's really the basis of all of the machines we use now.  So a home dialysis machine is where it all started.

            And the second comment, just a brief one in terms of patients, we did a study once when we had a psychologist come in and do IQs on 100 successfully, consecutive successfully, home‑trained home patients.  And the IQ ranged from 87 to 147, with an average of 103, which he told us is just where it ought to be as we did not take the really low‑level effectives.

            Like some of the previous speakers over here, I think almost anybody technically can do this if they are motivated and everything else falls into place.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Blagg.

            Dr. Moran?

            DR. MORAN:  I would like to echo Dr. Kalota's comments.  There are two places where you can make a choice about sending the patient home.  First of all, the health care team can look at the patient and say, "Does this patient look suitable for home therapy?"

            And then you have the training period, which is two or four or six weeks.  And, again, the patient can be assessed every day during that training period.  And we can abort the home decision at any time during that training period, and we do.

            I think that the other point that I would like to make is we keep on talking about professionals doing it in the center and nonprofessionals doing it at home, but at home you've got one person operating, one device on one patient.  And that is very, very powerful.  The same person is sticking the same access.

            It's very different from a patient coming into a center, who could have any one of 20 professionals who may or may not have seen this access before.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Moran.

            Dr. Lockridge?

            DR. LOCKRIDGE:  Well, I think to me the most overall comment would be that the FDA has invited the clinicians who are doing this and that this whole process is a process of physician‑patient decision‑making to see whether or not it's better or worse for the patient.  And we are looking for the FDA to give guidance to manufacturers to try to certainly make things safe, but it's that balance.

            I need to have the choice with the years of experience that I have to choose with patients the FDA needs to protect the overall public.  So we need to work together to come up with conclusions.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Hoy?

            DR. HOY:  Just I'm more confused than ever as to whether the jurisdiction is simply the concerns with the machines or access or ROs or water or dialysate or what or monitoring.  I mean, it sounds like the FDA ‑‑ in fact, you could go in any direction you wanted, but it's to go to the manufacturers ultimately I guess, right?

            CHAIRPERSON TALAMINI:  Well, I think that that is actually a good way to conclude this part because what we really do need to figure out is what the objective of the day is and how we are going to get there.

            So we have this set of eight questions, and we need to go through them.  You have all looked at them.  And they do cover a broad array of topics and issues, some of which are directly on point of the machines and how they work and some of them have to do with access and other issues related.  So our job for the rest of today is to do our best to answer those questions and to provide the information.

            Now, again, that is a Herculean task.  Let me tell you how I hope that we can do it.  And we will try this with the first question.  What I hope to do is go around the table, give everybody a minute and a half to address some piece of the question, hopefully a piece that has not already been addressed earlier going around the table.

            If you've got nothing to say, a pass is just fine.  By doing that, hopefully we will get all the way around and have most of the information and still have time to go back and see if there are things that we have missed or other things that we need to discuss.

            I would also make it clear to the panel and to the audience that we are not voting for approval or disapproval of any specific device today.  There may be times where I as the chair ask for a straw vote from the panel to gauge opinion of the panel as a whole, but I don't want in any way for that to be construed as any sort of official panel vote of approval or disapproval.  If we do that, it would only be to try and get the temperature of this group here today.  Now, if that strategy does not work as we move along through the day, we'll modify it as best we can.

            So the way we will actually go through this is to read these questions one by one and then go around the panel.  And we'll start at a different spot with each question and go clockwise so that everybody has an opportunity as much as possible to respond.

            So with that, if we could put up the first panel discussion point, which is on device design?  And I will read this question, "Standard hemodialysis delivery devices contain monitors and alarms to assess blood pressure, pulse, venous and arterial pressures, blood and air leaks, temperature, dialysate and blood flow, ultrafiltration rate, acid and bicarb pumps, end of treatment, and other parameters particular to the specific device.

            "Please consider and discuss the need for the following additional safety features in nocturnal home hemodialysis, NHD, treatments and provide suggestions for additional features.  You should take into consideration the importance of human factors when discussing these features."

            Next slide.  "Blood Access, a) additional safeguards to prevent blood access disconnections; b) alarms to detect fluid, blood or dialysate, leaks, and a moisture detector at the site of hemodialysis access."

            Next slide.  "Central monitoring, c) software incorporated into the NHD device allowing connection to the internet for remote monitoring; d) central monitoring of treatment and patient parameters, such as blood pressure, pulse, venous and arterial pressures."

            Next slide.  "User‑friendly design, e) instructions displayed on the machine itself that are clear and easy to follow for treatment setup, discontinuation, troubleshooting, and disinfection of the device; f) sensitive and loud alarms with clear explanations of what they mean and how to respond; g) a justification for leaving out any standard alarms or features found in traditional hemodialysis equipment."

            So that's quite a mouthful.  It's probably the biggest question, but our goal is to try and answer that question over the next 30 minutes or so.

            So, Dr. Hoy, if you could pick a spot in there?

            DR. HOY:  Do them all or just a through ‑‑

            CHAIRPERSON TALAMINI:  Well, I think it would be impossible to go a through and still be on time.  So perhaps if you could think about what piece of that you would like to address?

            And if it has already been addressed, please try and pick another piece.  And that way, hopefully we will get all the way around and get back to the place where we can summarize and rediscuss some of the points.

            DR. HOY:  So I guess I get to select one out of those.  I would like to comment on remote monitoring since we do remote monitoring over the internet in real time and collect the data from the dialysis and the alarms.

            We allow the patients to respond to the visual and audio alarms.  And our observer who is watching the screen for 33 patients at the current time responds after two minutes if the patient doesn't fix the alarm.

            I believe that remote monitoring is useful.  I don't think it is absolutely necessary.  And I know that there are definitely circumstances where it is not going to be feasible in patients who are too poor to have a second phone line and an internet connection.

            For us, we find it very helpful.  It's a very useful way to download information about patients and what is going on with their treatments and maintain a clear sense of where they are having problems.  It also helps us with some preventive maintenance on the machines.

            It also allows us to have some sense of who is running and not running each night because patients when they are well‑dialyzed may tend to take time off.

            I don't believe that you need a partner at home if you have remote monitoring with an observer.  The New York State Department of Health, which is, believe it or not, harder on us than the FDA or CMS, feels that we meet the criteria for a partner at home using this as a virtual partner.

            CHAIRPERSON TALAMINI:  Dr. Hoy, just to address Ms. Moore's question, do your patients have a human they can get in touch with at night?

            DR. HOY:  There is a human being at the other end of that alarm system from 9:00 p.m. to 7:00 a.m. every night.  And if they do not respond to that alarm within two minutes, he or she will call them up on their second phone and talk to them.

            Remote monitoring does not prevent bad outcomes of catastrophic events.  And we have had one patient have a cardiac arrest on dialysis.  Within 15 minutes, the emergency squad was there, called both by the patient's wife and by the observer.  The patient was taken to the emergency room and died.

            We have had one patient have a ruptured kidney while on dialysis.  And he was at home alone.  And the patient was saved by the observer, who had the emergency squad called and had the door broken down so that they could get in.

            And we have had one patient who had ‑‑ that's another issue.  Never mind.  That's it.

            CHAIRPERSON TALAMINI:  Okay.  Thank you very much.

            Dr. Lockridge?

            DR. LOCKRIDGE:  I would address the blood access.  First of all, I think it's the physician and the patient should decide the blood access associated with the types of dialysis.  It's not the manufacturer or FDA I don't think.

            I think there are really three different types of blood access that need to be addressed:  number one, a catheter.  Certainly about 27 out of 33 people or 25 out of 33 people I have at home now have a catheter.  The infection rate is much less because of an interlink device and a disconnect device and how the patient is handled.

            So I think standard insulin or catheter use is different in center versus at home.  I think it's a safer modality.  I'm not saying that that should be the primary access, but I think that FDA needs to have the companies look at some way of assuring a disconnect.  A simple clam shell, a simple interlink device would be of significant benefit.

            As far as the A‑V fistula and the discharge of the needle or the graft, I think that the A‑V fistula using a button hole technique and just simple tapering does prevent the needle from coming out.

            We have done 41,000 treatments at home.  We had never a needle come out in 41,000 treatments with a simple taping procedure, the arm wrapped with a fishnet, and then the lines anchored to the upper arm.

            So there are real techniques that make needles in the arm, whether it be in a fistula in the lower arm in the upper arm, safe.  The same thing that goes with a graft.

            So I think, in summary, vascular access should be chosen by the physicians with the patient.  I think all three are safe.  The real focus is on whether or not, how we assure the connection.  And there are techniques out there with clam shells that can prevent that.

            Thank you.

            CHAIRPERSON TALAMINI:  So, Dr. Lockridge, then just looking at point A, do I hear you saying that additional safeguards are not needed to ‑‑

            DR. LOCKRIDGE:  No.  I think that clearly in the home setting, Fresenius has developed a disconnect system that clamps on the tube.  We have a clam shell.  There is a disconnect device or a tape.

            But if you use that and the patient applies it ‑‑ and I think human error does happen.  In one case, I had a gentleman not go to sleep but put himself on and did not put his clam shell on.  He was watching TV.  And his catheter became disconnected.  And he had to stop the treatment.

            So that's out of 41,000 treatments.  But it's a patient error.  So if the manufacturer can create a way, a red thing that has to be taped around the catheter or there's a connection device that is built into the tubing that connects to the catheter or the needle, the needles once they are taped are okay.  So I think we just have to emphasize that the patient should be responsible in that way.

            CHAIRPERSON TALAMINI:  How about alarms, which is the second point of the access?

            DR. LOCKRIDGE:  In the 41,000 treatments, when the patient puts or sets up or builds the unit, they screw the blood line into the top of the kidney and in the bottom of the kidney.  And then there are valves that connect to the kidney, the dialysate.

            We have had one experience where the person over‑screwed the tubing arterial.  They went on, went to sleep, and through the night, there was a slow drip of blood that resulted in about a unit of blood loss over a night.

            We have and Pierratos and everybody now has a water detection, moisture detection, device, which I think should go under the kidneys or where it is or under the kidney machine, which would alarm and prevent that from happening.

            Once again, that is human error and if the manufacturer can figure out a way that when you anchor those lines in place there is a pop or something like that.  So the machines that are built with cartridges or kind of packed in there, that decreases the number of connections the patient has to make.

            But I think that a moisture detector device underneath the kidney would prevent that.

            CHAIRPERSON TALAMINI:  Great.  Thank you.

            Dr. Moran?

            DR. MORAN:  I think in view of the eloquence of my colleagues, I should yield my 60 seconds.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr Blagg?

            DR. BLAGG:  I would just like to comment again on central monitoring because I'm pleased that Chris doesn't think it's essential because I don't think it's essential at all.  And many of the programs that are doing nocturnal hemodialysis at this point in time do not use central monitoring.

            I think to have an internet connection so that you could download information as to what happened during dialysis at a convenient time would be a nice addition where it's available, but it's nothing essential.  And in our program, we have one of the training nurses on call 24 hours a day to answer all patient questions.  And we haven't seen any problems.  So I don't favor central monitoring.

            Another point I would like to make is that there are state issues here, too.  I was at a meeting recently where somebody from Texas was explaining that for nocturnal hemodialysis in Texas, they're supposed to take half‑hourly blood pressures throughout the dialysis, which would certainly not help patients sleep.  Apparently the Medicare surveyors in Texas will actually want to find evidence that this is being done.  So there may be state regulations which may have different effects on some of these things.

            Thanks very much.

            CHAIRPERSON TALAMINI:  So we have already heard two different opinions on central monitoring.  Perhaps I could just get the temperature of the panel on that one issue.  How many would favor central monitoring being a requirement of such systems?

            (No response.)

            DR. GIBSON:  Could we make that dependent on whether the patient is alone or has a partner, which I think makes a big difference?

            CHAIRPERSON TALAMINI:  I don't want to make it too complicated, but ‑‑

            DR. SCHULMAN:  I think that's essential.  I think you should make that difference.

            CHAIRPERSON TALAMINI:  Okay.  So let's say if the patient does have a partner.  How many think monitoring is ‑‑

            DR. WEINGER:  Just to get more complicated, are we assuming the partner is in the same bed or in some room at the other end of the house?

            CHAIRPERSON TALAMINI:  Present in the house.

            DR. WEINGER:  Just in the house?

            CHAIRPERSON TALAMINI:  Yes.  So how many with that stipulation, a partner present in the house, think central monitoring should be a requirement?

            (Whereupon, there was a show of hands.)

            CHAIRPERSON TALAMINI:  Okay.  Now, without a partner in the house, how many think it should be a requirement?

            (Whereupon, there was a show of hands.)

            CHAIRPERSON TALAMINI:  So I think there were six who said yes to that.  Okay.  So, again, just a straw poll to get the temperature of the group.

            Dr. Gillespie, you're ‑‑

            DR. LOCKRIDGE:  Can I comment on the central monitoring since the experience I have ‑‑ basically we have not done central monitoring.  I would like to have ability to download it, but in those 41,000 treatments, we have not identified a single event that central monitoring would impact on the safety of the patients or the response time to the patients.  And I think that is critical.

            We have a 24‑hour nurse on call that directly is called and talks to the person.  And we have 8 of our 33 patients at home dialyzing by themselves at home with nobody else in the home.

            So I'm saying that it appears to be safe, but I understand the viewpoint of the panel.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr.  Gillespie?

            DR. GILLESPIE:  Just a quick comment on alarms.  I think we have to think very differently about alarms than we do in in‑center dialysis, where we have alarms going off all the time, but because of the slow blood flow and slow ultrafiltration rates, it's much less common.  So it may not be as much a concern about disturbance of sleep and everything.  In the studies we were given, they said there were an average of about one to two alarms a night.

            I think in patients after they kind of completed the initial training and got used to it but certainly an important concern that we should look at when we are evaluating these systems is, how often do they alarm?  And what does that do to the patients' sleep?

            CHAIRPERSON TALAMINI:  Terrific.  Thank you.

            Dr. Weinger, again I would offer you more than a minute and a half since so many of these issues are human factor issues.

            DR. WEINGER:  Well, thank you.  Thank you for the comment on alarm.  I don't have to say as much about that.  So I'll just say that what alarms are present need to have a high positive predictive value.

            And false alarms are almost as concerning as the absence of an alarm.  In the presence of a false alarm, at best, they're disruptive.  At worst, they're ignored or even disabled in some way that can impair safety.  And alarms need to be designed and evaluated on any device.  And IEC 60601‑1‑8 is perhaps the best source for that.

            More generally, I want to talk about ‑‑ there's this whole section on user‑friendly design here.  And I wanted to talk about that in a general way.  And that is that ‑‑ and I know this point has been made, but I think it's critical that useability testing is absolutely critical for these devices.  And it has to occur before the clinical trials and, in fact, needs to occur throughout the design phase.

            The process results and documentation of the human factors engineering for the device need to comply with current national and international standards, which include ANSI AAMI HE 74, IEC 60601‑1‑6, and ISO IEC 62366.

            The final thing with regard to design is that there are a lot of questions later about labeling and training, but I want to emphasize the importance of designing the product right in the beginning to minimize the need for patients to have training on how to attach all of these connections, on how to run the session, on how to respond to alarms.

            The design should be sufficiently robust that errors are prevented wherever possible, that when errors occur, that there is safe and easy recovery, that when the device fails, it fails safe.  And then you have issues of alarms.

            The other thing to think about we will talk about more is just‑in‑time training, where, rather than just a beep going off, that the device provides information about what is wrong and what one needs to do to fix it at the time the event occurs.

            And I think I will stop there.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Gibson?

            DR. GIBSON:  The only other thing I want to add is that consideration should be given to having instructions displayed, not just clearly in English but clearly in other languages.  At least in New Orleans, we have many people who might well benefit from this who do not speak English very well and don't read English at all.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Gibson.

            Dr. Kalota?

            DR. KALOTA:  Nothing more to add.

            CHAIRPERSON TALAMINI:  Thanks.

            Dr. Aranoff?

            DR. ARANOFF:  A couple of brief comments on each section.  When it comes to the choice of blood access, that's going to have to be a medical decision based on the physical characteristics of whether the patient can have a fistula conduit or a catheter.  And the devices should be designed to function equally well with all three blood access, potential blood access.

            When it comes to monitoring the blood access, redundancy has to be the key.  However it is monitored, there have to be redundant systems because it is inevitable that a blood disconnection will occur at some time.  And there must be multiple system failure before someone exsanguinates.

            Central monitoring, whether or not it's mandatory or not, it should never be allowed to be used as the only way of monitoring patients.

            With regard to user‑friendly design, Dr. Mendelson's talk could be used as a check‑off list for the design of these devices.  They should be as automated as possible.

            They should not be designed in a way that allows for use error.  And patients should not be allowed to move from one point of setup from treatment to cleanup without completing all of the steps sequentially in order.

            And the devices should be designed in a way so that alarms or the appropriate process cannot be short‑circuited, as they so frequently are in dialysis units or in intensive care units.

            CHAIRPERSON TALAMINI:  Terrific.  Thank you, Dr. Aranoff.

            Dr. Afifi?

            DR. AFIFI:  Yes.  I would like to look at it from the point of view of a public health professional.  The concern I think should be for taking the optimistic view that Dr. Blagg offered that this could actually be better than existing current available technology.

            I hope this will be the case, in which case, though, what we want to pay special attention to is something that Dr. Weinger said a minute ago.  And that is, if the device fails, we want to design it so that it fails safe.

            And that I think is a fundamental point from the point of view of looking at the overall results when analyzed on a population‑based level that building in the redundancy that Dr. Aranoff is something that perhaps the FDA could think about requiring in some way.  I don't know what the actual implementation of that would be.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Afifi.

            Dr. Sadler?

            DR. SADLER:  Two quick points.  One of the items on here says, should we justify leaving out standard alarms?  I would say no.  The standard alarms are good, and we need them.

            Secondly, the blood access is the greatest source of anxiety and the greatest opportunity for improvement.  This is the one place where a central catheter without needle access would be desirable.  And if product development is going to take place, a central catheter that is more securely attached less prone to thrombosis and infection would be an advantage.

            CHAIRPERSON TALAMINI:  Just prerogative of the chair here, does anyone disagree with Dr. Sadler with regard to point G, justification for leaving out any standard alarms or features found in traditional equipment, just so we can be clear on that point?  Dr. Lockridge?

            DR. LOCKRIDGE:  Yes.  I think the traditional alarms, there's so much that has been on a dialysis machine:  sodium modeling; conductivity; Kt/V, which has become standard modeling.  All of that stuff needs to go.  It needs to be very simple for the patient.  And that's built into a lot of alarms there.  So that needs to go.

            But as far as the standard alarms, we need to keep them.

            CHAIRPERSON TALAMINI:  Any other comments on leaving out alarms?  Dr. Hoy?

            DR. HOY:  Our patients don't have a problem with the alarms.  We only have about 1.3 alarms per patient per night.  What they have a problem with is that sometimes they're more sensitive than they need to be.  And specifically for a catheter, they would like a longer time before the alarm goes off.  And for a fistula, we would like a shorter time before it goes off.

            So we would like some variability built in that allows us to set the alarms to go off dependent on the access.

            CHAIRPERSON TALAMINI:  Great comment.

            Dr. Lockridge?

            DR. LOCKRIDGE:  The other issue that we found in training, hearing is a big issue as you get older.  And none of the machines are designed to deal with a hearing deficit.  We have developed a vibrating connector to the machine that vibrates and puts under the pillow.

            So I think technology is out there.  The manufacturers need to look at that to deal with hearing deficit.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Weinger?

            DR. WEINGER:  Just like any other feature on these future devices, the alarms that are chosen need to be tested in the end users.  And so one can vote in favor, as I would, of retaining alarmable conditions, but the specific alarms and how they are triggered and what they sound like probably will be very different for a home device than for a device in a dialysis unit.

            CHAIRPERSON TALAMINI:  Dr. Hoy?

            DR. HOY:  One last comment.  When we asked our patients how they would approve the alarms, some of them have already done it.  We use a Fresenius 2008H machine, which has buttons, not a touch screen.  And they have put tactile felt or something else on the buttons to identify the ones that most commonly go off, the arterial pressure button, so they can reach out and just shut it off as they turn over and go back to sleep.

            Tactile probably is helpful, not useful with a touch screen.

            CHAIRPERSON TALAMINI:  Great.  Thank you.  Those are great comments.

            Dr. Schulman?

            DR. SCHULMAN:  Well, I would also agree with Dr. Sadler that the most worrisome issue is worrying about a blood disconnect.  And I think that to the extent that the manufacturers can provide us with a way to have these safeguards be set so that the machine can't, the patient can't be treated unless their force should be an important feature of design.

            Secondly, I think I would like to know about the quality of the treatment.  So in central monitoring, I think you can have a memory stick with a USB port on the machine to capture it with the parameters of the treatment, including things like Kt/V, that could be brought in for review periodically.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Schulman.

            Dr. Duffell?

            DR. DUFFELL:  I'll pass.

            CHAIRPERSON TALAMINI:  Ms. Moore?

            (No response.)

            CHAIRPERSON TALAMINI:  That is terrific.  Great job.  We did that in 20 minutes.

            So just a couple of other issues.  I don't think anybody discussed specifically point C in central monitoring, and that is software as involved with central monitoring.

            Do any of the panel members have a specific comment with regard to software?  Dr. Lockridge?

            DR. LOCKRIDGE:  Yes.  I think that the point about bringing in and clinically monitoring these people, if you could have the machine designed to be able to just take a phone line and plug it in, hit a button, and it dials up and downloads the stuff to my computer in the home training area, then I can monitor each treatment each day, the nurse can, and look at that.

            The other thing is central monitoring is critically ‑‑ in my population, I looked at it.  Forty percent of the people that I've sent home could not afford two lines of internet connection.  So when we basically create a central monitoring, we're going to select out the people, the haves and the have nots.  And that is not who is doing dialysis.

            So I think we have to figure out a way to not do that.

            CHAIRPERSON TALAMINI:  Dr. Hoy?

            DR. HOY:  A comment on this issue of the software.  This is done very differently all over the world.  About half of the units that are doing nocturnal dialysis in the world are doing monitoring.  Let me take that back.  It's about a third of the units, but they constitute half of the patients.

            For example, in Holland, their monitor, their observer, is Lifeline.  They find so few alarms and so few problems that they outsource it to Lifeline.  They use people who are not trained in any way, shape, or form to know anything about dialysis.

            In Toronto, Dr. Pierratos brings people in off the street who have no contact at all with dialysis and trains them to be the observers.

            This isn't rocket science.  It really is safe to do.  And you can do it.  And you can have a set of protocols that allow us to know when to call the nurse on call, when to call the doctor on call, when to call the technician on call, and how to intervene.

            So that there are definitely software that is useful.  Fresenius has come up with a wonderful system called Eyecare.  There is another system out there by Sebernius, which is what we use.  These are very good systems.  They're not perfect, but they work.  And then you have protocols which are implemented by the people watching.

            CHAIRPERSON TALAMINI:  Let me do Dr. Blagg first.  Then I will come around to Dr. Sadler.  Dr. Blagg?

            DR. BLAGG:  A couple of quick comments.  If you actually use central monitoring and add it to the cost of the treatment process ‑‑ and that may be a concern ‑‑ the same thing is that Dr. Pierratos does not feel that remote monitoring is essential to all.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Sadler and then ‑‑

            DR. SADLER:  I just wanted to make the point that among the issues we raise, this is of interest, but this is secondary to the more important ones.

            CHAIRPERSON TALAMINI:  Dr. Aranoff?

            DR. ARANOFF:  We're actually talking about two aspects of monitoring here.  One is the safety of the treatment.  And the other one is to access clinical information about adequacy and the characteristics of the dialysis treatment.

            My comments about safety and monitoring are that we should never use central monitoring as the only way of monitoring safety.  And others have recognized that it may not be necessary at all.

            And, secondly, I would point out that the majority of home hemodialysis or peritoneal treatments now done in the United States do not download information to any central source.  And, in fact, the vast majority of in‑center hemodialysis treatments provided in the United States do not download this information electronically to any central source.

            CHAIRPERSON TALAMINI:  Terrific.

            Dr. Kalota?

            DR. KALOTA:  He answered my question.

            CHAIRPERSON TALAMINI:  Okay.  I would just like to ask the FDA ‑‑ this might be a spring on these folks ‑‑ if there are other issues that the panel had either missed or that you would like us discuss in more depth with respect to device designs.

            MS. BROGDON:  Thank you.  I was checking with the staff on that question.  And Dr. Ruiz would like to ask something.

            CHAIRPERSON TALAMINI:  Okay.  Dr. Ruiz, please?

            DR. RUIZ‑ZACHAREK:  And I'll be very quick, too.  This question is for Dr. Hoy.  You said that partners are not required among your patient population.  You reported also some patients that have some serious events, not treatment‑related but basically a kidney rupture, which is pretty danger and you need to act quickly on it.

            What would happen if there is no central monitoring, no partner, and someone who lives far away?  In the cases that you presented, emergency personnel showed up in 15 minutes.  If these patients live in remote areas, what type of safeguard would we have or should we restrict the nocturnal dialysis to patients who live close to hospitals?  How would we address patients that live far away?

            DR. HOY:  This is a perfect technique for people who live far away from a dialysis unit in a hospital.  If a catastrophic event takes place in an in‑center dialysis unit and the emergency squad gets there in 15 minutes, which is about what it takes, it doesn't change the outcome most of the time.

            Those events, we have incredibly sick patients, you know, and they die.  And they die suddenly sometimes.  The fact that we have had a mixed bag, we had one patient die, despite the observer, the partner, and emergency response quickly.  We had one patient whose life was probably saved by the observer and the rapid emergency response.

            We send out a video to the emergency squad for every patient.  And we have the number of that emergency squad.  And that video shows them, this machine, and how to disconnect it because that is all they need to know.

            If the patient has a partner who wants to learn how to do this, we teach the partner to do it.  But the main reason that there are only 98 people in New York State on home hemodialysis out of 20,000 and less than 1,000 in the entire country is because of care‑giver burnout.

            CHAIRPERSON TALAMINI:  So would you contend, then, that the emergency response system of the country in general as it exists is already sufficient for whatever would happen with this population?  It sounds like that is what you are saying?

            DR. HOY:  Yes.  I don't think this population is any different than any other population that's at home and sick or in center and sick.

            CHAIRPERSON TALAMINI:  Other opinions, particularly contrary opinions, on the panel?  Dr. Sadler?

            DR. SADLER:  I simply refer to my earlier remarks about passive restraint levels of safety.  There are some things you cannot guarantee.  The worst thing that can happen to one of these patients is not to get his dialysis.

            So that to be remote from a hospital and a dialysis center and to have the ability to get it done at home without having to travel, it may involve some risk, but it avoids that very large risk.

            DR. RUIZ‑ZACHAREK:  Thank you.

            CHAIRPERSON TALAMINI:  Dr. Duffell?

            DR. DUFFELL:  Similar to the last comment, I just think FDA needs to keep in mind that we really can't regulate totally the use of these things.  I mean, clinical judgment has to come into play here.  Patients are very complicated.  So there may be extenuating circumstances.

            I think the most you can do is probably in the professional labeling give the benefit of the wisdom of the things that you see in MDR reports and complaints and stuff that come through of points to consider.  But the clinical judgment has to come into play.

            So if that patient is 100 miles away from an emergency response team, there may still be compelling reasons to do this.

            CHAIRPERSON TALAMINI:  Yes?

            DR. BLAGG:  Just one rapid comment.  Patients may go home with a helper, but the helper may not stay around.  And, for example, our patient with an IQ of 87, measured by a psychologist, not by me, we didn't find out until he was transplanted that for the last 15 months, his roommate had not been with him.  He had been dialyzing alone very successfully.

            So we can't control that the helper stays there.

            CHAIRPERSON TALAMINI:  Yes.  Thank you.

            So that will close question one.

            DR. LOCKRIDGE:  There is the issue about the water I don't think we really addressed.  That is in this section.  I think as far as design purposes for the machine, I think AAMI standards are very critical and we need to be able to have a patient‑friendly port that looks at the water post ‑‑

            CHAIRPERSON TALAMINI:  That is actually question two.

            DR. LOCKRIDGE:  I'm sorry.

            CHAIRPERSON TALAMINI:  So we're about to get to that.  So let me be the ugly surgeon and get you guys to do one more question before we break for lunch.  So if we could do question two, which refers to device design with respect to water?

            "The quality of the water to be used to prepare dialysate is crucial for any dialysis treatment.  Please discuss the water purification needs for the NHD procedures, including the following:  a) type of water treatment equipment for preparation of water and verification of its quality appropriate for nocturnal home use; b) due to the potentially higher exposure of patients to the processed water, consideration as to whether the water quality recommendations should be different for nocturnal home use, as opposed to conventional, in‑clinic use; and c) procedures on how to handle changes in the water quality and composition by municipal water suppliers.

            Dr. Lockridge, since you have started in, go ahead.

            DR. LOCKRIDGE:  All right.  I think water deals with 60 percent of the problems that I have with the patient at home.  I think it is a major issue that we have to address.  I think we have to meet the AAMI standards.  And I think that if we do that, that is where we need to be.

            Clearly, I think whether we choose to do an RO versus a DI is not the issue.  I think what we do need to have is the highest quality water because these people are seeing more water per week than they are when they are doing three times a week in‑center dialysis.

            There needs to be, you know, the carbon filter, a pre‑filter that protects the sediment that gets into the water, well water, before the carbon tank, the carbon tank to do with the chloramines, and then a post‑DI or if you use DI microfilter.  And then you need to have another filter to do the dialysate or to create ultra‑pure water, I think.

            So I think that we need easy ports for the patients to access the water at each one of those points so that we can test the water so it meets AAMI standards and the clear CMS requirements once a month for colony counts and LALs and how often we need to look at the dialysate.

            So I think it's critical, easy for patient to access.  We're not going to send technicians to go out and do that.  We need the design to do that.

            CHAIRPERSON TALAMINI:  Terrific.

            Dr. Moran?

            DR. MORAN:  I think the new AAMI standards have it specifically the same for in‑center and home use.  And I would agree with that.  I think I disagree about the concern about the potentially higher exposure of patients to processed water.  I think if they make the AAMI standards, that should be fine.

            Procedures how to handle changes to the water quality, that's a never‑ending saga, both for centers and for home patients.  And it's one of the biggest things we struggle with at home.

            We have to have redundancy in the system.  That's why we have two carbon tanks.  They're testing the water between the two carbon tanks.  And when the first carbon tank shows evidence of a breakthrough, then they change that tank.  It's a fail‑safe system, but it is a big issue.

            One of the concerns I have about AAMI is saying that the water should be tested once a month at home.  That's much more difficult to do than it is in center, especially for some of the machines.  I think at home, where you are dealing with a single established system which doesn't have storage tanks, as you do in center, no blind loops, as you do in center, there is much less risk of water problems.  And, therefore, the testing should be done at some period less than once a month.

            CHAIRPERSON TALAMINI:  Terrific.  Thank you.

            Dr. Blagg?

            DR. BLAGG:  Just a brief comment, following up on Bob.  I think that there is a tendency to move towards ultra‑pure water for dialysate, particularly in Europe but coming in this country, too.  And it's something that we should think about.

            I agree with Dr. Moran that it is very difficult.  The way we handle our home patients is we try to contact the local water company, wherever they may live, every six months and also make sure that they know our phone number.  But it's a problem you can't always solve.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Blagg.

            Dr. Gillespie?

            DR. GILLESPIE:  I think I'll pass and defer to my colleagues on this.

            CHAIRPERSON TALAMINI:  Thanks.

            Dr. Weinger?

            DR. WEINGER:  Just a general comment.  To the extent that patients have processes, procedures, and devices related to water purification and assessment, those need to be designed for patients and evaluated for their ability to use them effectively.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Gibson?

            DR. GIBSON:  Yes.  This is more in the nature of a question for those of you who know more than I now do about the technical aspects of hemodialysis since I haven't taken a dialyzer apart or set up a water treatment system in a long time.

            But it used to be at least theoretically possible for back diffusion to occur in a high‑flux system.  And by that, of course, I mean dialysate passing into the blood compartment.

            So my question is, is that a real possibility?  And would that make a difference in your view of whether AAMI's standards are essential or whether we should recommend ultra‑pure water?

            CHAIRPERSON TALAMINI:  Dr. Sadler, do you want to address that?

            DR. SADLER:  I don't think there's any question but that back diffusion occurs in dialyzers.  And I don't think there's any evidence that it's harmful.  And, as I said before, I am still waiting for any evidence that anyone is harmed by water that meets AAMI standards, even the original AAMI standards.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Kalota?

            DR. KALOTA:  Pass.

            CHAIRPERSON TALAMINI:  Dr. Aranoff?

            DR. ARANOFF:  Just a couple of brief things about water.  First of all, I agree with my colleagues that there is no evidence that water that meets the AAMI standard would be inadequate in any way for this form of therapy.

            However, with regards to the increased amount of ultra‑pure water that is processed, we need to make sure that the home systems that are used are capable of handling the increased volume for the creation of dialysate that may not be considered.  You're running a lot more water through the systems.

            And so issues of number of hours on the device, issues of preventative maintenance and so forth, those issues need to be considered in the use of the device because they may need to be done more frequently than we are used to than with current home devices.

            And, again, with regards to procedures to handle water quality from municipal systems, many of our home patients live in very rural areas.  If they have city water at all, the people who make that city water may not be as highly trained as in larger cities.  And they have a tendency to dump things into the water to make it taste better or clearer.

            And there needs to be some way of monitoring when alum is dumped in to take out particulates, when chloramines rise and may foul the carbon tanks.  And so that needs to be monitored, I think, as Dr. Lockridge pointed out.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Afifi?

            DR. AFIFI:  No comment.

            CHAIRPERSON TALAMINI:  Dr. Sadler, further comments?

            DR. SADLER:  Only two.  One, we need to remember that the AAMI standard is a performance standard.  It's not a prescriptive standard that says, "Thou shalt use" this device or that but that the water should come out meeting the standard.

            I agree completely with Dr. Weinger that we haven't thought enough about the displays on water treatment systems for the benefit of patients and that should be considered.

            And beyond that, nothing except to remind you that a private well doesn't have to worry about a municipal operator fouling it up.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Schulman?

            DR. SCHULMAN:  Most of what I thought of has been said already, but just for completeness, I would add that for the areas that have hard water or water softener should be part of the water treatment.

            CHAIRPERSON TALAMINI:  Okay.  Dr. Duffell?

            DR. DUFFELL:  I'll pass.

            CHAIRPERSON TALAMINI:  Ms. Moore?

            MS. MOORE:  Pass.

            CHAIRPERSON TALAMINI:  Dr. Hoy?

            DR. HOY:  As usual, I have a lot to say.  We have had a lot of different experiences with different water sources.  Several of our patients are in rural areas.  And we have discovered that if you are downstream, if your aquifer is downstream, of agricultural lands, there are real issues with increased phosphates when people are fertilizing and when there is a lot of rain.

            We also have problems with droughts.  And so patients have had to come in center for that.  We have also had people who were downhill of somebody else's septic system.  Their well was downhill with someone else's septic system.  These are real problems, but we have managed to settle all of them when we need to.

            We have urban systems that are old with a lot of particulate matter.  You have to have pre‑filters.  We do water softeners and carbon filters.  We use ROs because they are less expensive and they're less cumbersome and bulky.

            We have had problems with suburban homes with patients getting constantly reinfected.  And it turned out that the home itself had E. coli infection and other gram‑negative bacteria in the piping of the system.  We had to shock the whole house to get rid of the problem.

            The supply, especially for ROs, needs to be more efficient.  You throw away half of the water we use in RO.  We want to get up to the industrial efficiencies, where 75 percent is recovered for dialysis.

            You have a problem with the adequacy of septic systems for effluent.  You're dumping a lot of water every day into a septic system.  You often need a larger drainage tank.  There is an ongoing cost, increased cost, to the patient of water, though in our area, it's not huge.

            Reverse osmosis machines if they have stagnant loops often get contamination.  And they need to be designed so that water is always flowing between the dialysis machine and the RO.

            Incorrectly collected cultures are the bane of our existence.  In our unit in March, five patients had to be treated in center for at least one dialysis and in April, four patients.  And half of those were because of incorrect or because of falsely positive cultures.

            Our patients point out that they cannot collect the cultures if there is an open window, a fan, or a ventilation system going on in the room at the time they collect them.  It would be very nice to have what Bob suggested, which is an easy port through which those cultures can be collected safely and cleanly because it a huge inconvenience to the patients to have to go in center while we are waiting for cultures to come back and be proven to be incorrect.

            The new AAMI standards don't worry me particularly.  I think the issue of ultra‑pure water is a problem that you should be addressing for all dialysis patients.  If you told someone that you were exposing them to two liters of water a day that they drank and then you told them that you were exposing them to 576 liters of dialysis weekly because they go to dialysis 3 times a week for 4 hours and then have a dialysate flow rate of 800 cc, you wouldn't even worry about whether the 1,400 liters that we expose our nocturnal patients to is an issue.

            I think you need to address the issue of ultra‑pure water for all of our dialysis patients.  And you don't require it for our in center patients, and we should probably.

            Maintenance of ROs is incredibly cumbersome.  They should be made simpler.  They should have timers.  They should have automation.  They should be online so that we can download preventive maintenance.  They're not reliable.  They need to be more reliable.

            DI tanks we find give more efficient water use, but they're more expensive over the long run.  They do have lower maintenance.  They're bulky, and they have to be replaced every three months.

            CHAIRPERSON TALAMINI:  Thanks.

            Dr. Weinger, would you be willing to address a human performance way to help the false positive culture problem?

            DR. WEINGER:  I'm not sure I fully understand it, but what I think I heard from one of you was the idea of having this dedicated port.  And certainly if this is critical to the success of the overall treatment, then that ought to be part of the waster component, needs to be part of the overall system.  And built into that should be a mechanism to take the cultures easily and effectively as necessary.

            I think that I have a general comment, if I could, at this point ‑‑

            CHAIRPERSON TALAMINI:  Sure.

            DR. WEINGER:  ‑‑ about the fact that what we see with other medical devices, whether it's in the home but even in the hospital, is that when the functions that need to be accomplished require multiple components, particularly if they are manufactured by different companies, the risk of use there and injury increases substantially, probably exponentially.

            And so something to think about is whether home systems need to be a comprehensive integrated system that includes the monitors, the water treatment, et cetera, all as one unit, rather than the ability or the obligation, as happens now, where the practitioners are cobbling together devices and solutions from multiple sources, which I think substantially increases the risk.

            CHAIRPERSON TALAMINI:  Thank you.

            Let me just try and get again the temperature of the committee with respect to question B here because it sounds like there have been slightly different opinions.  And that is whether water quality recommendations should be different for nocturnal home use, as opposed to conventional.

            I think I heard mostly a no.  So I'll ask the opposing question.  Is there anybody who thinks that they indeed should be different for nocturnal home use.  A show of hands, anybody who thinks they should be different?

            DR. HOY:  Dr. Talamini, do you want the consultants to vote or not?

            CHAIRPERSON TALAMINI:  You know, I don't think it much matters here because we are really just getting the opinion of this committee.  It's not an official vote.  So don't think ‑‑

            DR. SADLER:  Dr. Talamini?

            CHAIRPERSON TALAMINI:  Yes, sir?

            DR. SADLER:  When we go from normal kidney function to dialysis, we increase water exposure 200‑fold.

            CHAIRPERSON TALAMINI:  Right.

            DR. SADLER:  When we go from conventional dialysis to frequent nocturnal dialysis, we double it again.  So that's not an order of magnitude, and I don't think the difference is significant.

            CHAIRPERSON TALAMINI:  Thank you.

            I would ask the FDA again if there are other issues you would like the panel to address specifically on this question before we close the question.  Yes, sir?

            DR. MENDELSON:  I have a question.  This doesn't pertain directly to human factors but just a thought.  Does the presence of plastic piping, which is probably more prevalent in the home, place a higher bacterial load on the filtration equipment that is going to be used to treat the water?

            CHAIRPERSON TALAMINI:  Any panel members have insight or response?

            DR. LOCKRIDGE:  You know, you're talking about microfilming that happens in plastic, but I just think that that is probably not an issue for the amount of water that runs.  I just can't ‑‑

            DR. ARANOFF:  Functionally in dialysis units, all of the piping is plastic in dialysis units.  We have to use PVC piping.  Otherwise, ultra‑pure water will etch metal pipe.  So every fitting, every pipe in a dialysis unit is already plastic.

            CHAIRPERSON TALAMINI:  Okay.  Perfect.

            Any further questions from the FDA?  Yes?

            MS. MOORE:  I have a question.  I just have a question based on the last comment that was made.

            CHAIRPERSON TALAMINI:  Ms. Moore, if you could just bring the microphone in?  Thank you.

            MS. MOORE:  I have a question just on your last comment.  Does that mean that the old houses, you know, with still the copper piping and that kind of things, there would have to be a change or something?

            DR. ARANOFF:  No.  The difference is that in a dialysis unit, you process the water before the dialysis unit.  So, actually, what is delivered to the dialysis unit is essentially ion‑free ultra‑pure water to be mixed into dialysate at the machine.

            So from the point that the water is purified in the dialysis unit in the basement or on the ceiling or on the roof, wherever, from that point on, it has to go through plastic piping.

            In a home, in a home setting, the water is purified closer to the machine.  But from the point, that point on, it has to be plastic piping.

            CHAIRPERSON TALAMINI:  Dr. Neuland?

            DR. NEULAND:  Yes.  I still need a little bit more clarification on the views on the ultra‑pure water because your question, Dr. Talamini, was, do you think that the systems in the homes should be different than the clinic?  But I heard a number of people here say, "I think ultra‑pure water should be used, and then it should also be used in the clinic."

            So, I mean, I'm not getting a good picture on whether you really believe ultra‑pure water should be used in the nocturnal home hemodialysis.  I mean, I feel like we have kind of evaded the exact question.

            CHAIRPERSON TALAMINI:  Well, I think this panel is unwilling to say that there is a difference between the two.  But I think it would be fair enough to ask for the temperature of the committee with respect to that question, whether overall ultra‑pure water should be used in dialysis, period.

            Is that fair enough?  Can I ask the panel that?  Dr. Lockridge, do you have a better suggestion or ‑‑

            DR. LOCKRIDGE:  No.  You know, I think that the literature from Europe and some literature now that is coming out here would suggest that ultra‑pure water is better, that we're looking at more of an inflammatory marker, but I agree with when you're looking at 1,000 patients versus 300,000 patients and trying to make a difference when it's just a twofold difference, I would hope that the designers or the manufacturers would have what we call a dialysate filter before their filter or they're using columns to remove it to make it ultra pure.

            I would think that would be a goal, but if FDA requires it for home patients to be a goal, then I think that is inappropriate unless they are going to require it for in center.

            CHAIRPERSON TALAMINI:  Dr. Gillespie?

            DR. GILLESPIE:  I think this is still a developing field of research.  I mean, intuitively cleaner water sounds better, but we still haven't totally figured out how it plays out clinically.  And we also have to consider when we're talking about outcomes that I think most of us would agree that more frequent dialysis has a tremendous effect on outcomes.  That's kind of the big picture.

            And so issues about the water purity are probably just the detail that may not have as much of an effect as just the transition from less frequent to more frequent dialysis.

            CHAIRPERSON TALAMINI:  Other comments regarding ultra‑pure water?  Dr. Sadler?

            DR. SADLER:  I think I have already made it clear that I believe the increase in the quality of water, the standards reached for have nothing to do with the treat to patients that we can define.  It has to do with the fact that we are capable of making a higher quality of water.

            There is indeed some literature that says perhaps there will be a lesser inflammatory response with ultra‑pure water, but I don't think that data is anything like strong enough to change the practices across the field.  And I think we have to be cautious about requiring things for which there is no clear need.

            You know, this is a group that is a mixture of zealots and skeptics.  And so we're always going to have differences.

            CHAIRPERSON TALAMINI:  Right.  We'll take a one‑hour lunch break at this point.  I would like to thank the panel for really terrific work and concise comments.  And hopefully we'll be able to continue the same.

            Yes, Dr. Duffell?

            DR. DUFFELL:  Just to the audience in my role as industry rep, if there are any members of industries out there that have any matters that are going to be discussed with the panel today that you would want to discuss with me, please approach me during lunch.  Thank you.

            CHAIRPERSON TALAMINI:  And I would remind the panel again that this is an open proceeding.  So may not discuss any of these issues that are before the panel during lunch.  So I would ask you to be disciplined.

            So we will reconvene at 1:46.  Thank you.

            (Whereupon, at 12:46 p.m., the foregoing matter was recessed for lunch, to reconvene at 1:30 p.m. the same day.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 


         A‑F‑T‑E‑R‑N‑O‑O‑N  S‑E‑S‑S‑I‑O‑N

                                       (1:48 p.m.)

            CHAIRPERSON TALAMINI:  I think we will go ahead and reconvene the panel at this time, again in the interest of getting done in a timely manner.  So the meeting now reconvenes with the continuing panel discussion portion of the meeting.  And we will begin with the third question.  So far the processes seem to work well.  So we'll stick with it.

            Question number 3 has to do with risk analysis.  FDA has identified the following potential risks associated with NHD, in addition to those related to conventional hemodialysis:  a) increased risk of inadvertent disconnections, the increased blood loss from increased frequency of treatments, c) potential increased rate of vascular access infection due to increased use of access, and d) psychological effects; e.g., impact of treatments on patients, such as loss of social interaction, and the impact of increased responsibility on the patient, requiring need for adjustment or discontinuation of therapy.

            To aid FDA in the development of a guidance document on NHD, please comment on the completeness and appropriateness of this list.  I think we left off with Dr. Moran if you're willing to pick a piece of that and comment for a minute and a half.

            Again, I would applaud the Committee on their clarity and brevity this morning.  And if we could continue to push that, we'll make the taxis and so forth, the flights this afternoon.

            Dr. Moran?

            DR. MORAN:  I think this is a very good list.  I agree that the risk of disconnection is the single most serious problem we should be dealing with with nocturnal hemodialysis.  And that has to be addressed, whether it's catheters, as Dr. Sadler wants to do, or with fistulas and button holders I want to do with precautions against disconnect.

            I agree it's a very serious issue.  I think technically the best answer to that would be a single needle technique, either a dual/single needle or a machine that is providing a single needle.

            Increased blood loss, I can see that that occurs.  I don't think it is an issue in the age of intravenous iron and erythropoietin.  I think the benefits of frequent dialysis far outweigh this risk.  And conventionally most people find that erythropoietin usage actually goes down with nocturnal hemodialysis.

            Vascular access infection, I would make one point.  In our home patients using the button hole techniques six times a week, we have had two serious episodes of staph aureus septicemia in young men with fistulas using the button hole technique.  And I can't remember the last time I saw staph aureus septicemia with a fistula.

            We know there was a break in technique from one patient.  We know the other patient was doing things like dialyzing himself at 3:00 a.m. and probably, therefore, careless.  We have since then very strongly emphasized skin prep and haven't had any more problems.

            But the button hole technique is wonderful.  The patients love it.  But I would caution about the increased risk of infection.

            Psychological effects.  I don't see those.  The psychological effect I think you have to be careful about is the partner.  I think it's like living related donor transplantation.  You have to take the partner aside and make sure the partner understands what they are committing to and make sure that they have a real commitment, they're not just saying yes because they don't know how to say no.  I think that is a major issue.

            I think that is the end of my comments.

            CHAIRPERSON TALAMINI:  Okay.  Thank you, Dr. Moran.

            Dr. Blagg?

            DR. BLAGG:  I agree with all that has been said.  I think the bit I would add is that our experience at least has been that vascular access problems of all sorts are not increased with more frequent sticks and that this reflects the fact that the patient or one other person is doing all of the sticks, rather than the 20 people that somebody mentioned a little while ago.

            And in terms of the psychological effect, we still try to persuade the patient to do as much themselves.  And in our program, about two‑thirds of the home patients, the patient is the prime mover and does most of everything or everything, but psychological problems do occur from time to time obviously.

            The second patient that we trained in Seattle should have been divorced.  Unfortunately, the patient died, but it really destroyed his marriage.  And that made us aware of these things very early on.

            CHAIRPERSON TALAMINI:  Thank you.  Dr. Gillespie?

            DR. GILLESPIE:  I agree with all the preceding.  I would just add that on item C for the potential increased rate of vascular access infection, I would broaden that to just any vascular access complications because if you are worried about potentially thrombosis or other failures resulting from increased use.  And, of course, the literature hasn't really borne that out, but if you're looking at risk, that would be what you would want to consider, I think.

            CHAIRPERSON TALAMINI:  Okay.  Dr. Weinger?

            DR. WEINGER:  One item that isn't on here that we have mentioned before is acute decreases in water quality, for whatever reason.  In terms of useability issues, I would add to the list risks of misconnection, particularly related to disposable components, issues related to family, friends, pets, and other entities in the environment interacting with the device, and then the general issue of response to abnormal conditions and either a failure to respond or an inadequate or inappropriate response to emergency or just atypical condition.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Gibson?

            DR. GIBSON:  Well, I'm glad to see psychological effects on the list at all.  In terms of adaptation to chronic illness, there is much more that we don't know than we do know in terms of how people cope with the stress of chronic illness.

            As I said before, I think that there are some potential benefits to home dialysis that need to be considered along with the risk.  And those benefits would be a sense of mastery that increases self‑esteem and psychological well‑being.  And I don't think that is trivial.

            I think other people have already emphasized another aspect of this that is extremely important.  And that is if negative effects are seen, it is more likely to be on the family, rather than on the receiver of the dialysis solely.

            Now, I think that was shown beyond any question.  And what limited data is available from studies of conventional home hemodialysis and peritoneal dialysis is that care‑giver burnout, which someone has already mentioned, is one of the more significant negative aspects, negative psychological aspects, of dialysis.

            And it's not just limited to the care‑giving partner, but it involves the whole family in new ways of learning social interaction and new ways of scheduling activities and new limitations on things that they can no longer do that they used to do.

            Now, I don't think any of these are reasons for the FDA to not approve this as a method.  As I've said before, these are things that we don't know the answers to on conventional dialysis.  And the purpose, as I understand it, of our meeting here is to determine safety and effectiveness, not to determine long‑term effects.

            So I do not like to see us use psychological exclusions, arbitrary psychological exclusions, as reasons not to put people on nocturnal hemodialysis.  I was very impressed with what Dr. Hoy said earlier, which was that the main determinant of who can go on nocturnal dialysis is a person's motivation for doing it.

            I don't think psychiatrists are able to predict with any degree of accuracy who is smart enough and adept enough and intellectually aware enough to do this successfully over the long term.

            CHAIRPERSON TALAMINI:  Thank you.

            Perhaps to add a little bit of clarity, it seems ‑‑ and anybody can correct me if I am wrong on this ‑‑ the key question here is whether any of these four or any that aren't on this list represent incremental or additional risk for nocturnal home dialysis, as opposed to dialysis as it's now practiced.

            So it sounds like we've got a set of comments from the beginning folks on all of these issues, but as you think about this question, I think that is the real thing we need to get at is whether the FDA has to be concerned about incremental risk with doing this at home, as opposed to the way it is currently practiced.

            Dr. Kalota?

            DR. KALOTA:  I think Dr. Gibson elaborated my comments much better than I would have.

            DR. ARANOFF:  I think to address that specific question is that the incremental risk is related to the incremental number of treatments and the time that are incrementally increased on dialysis.  All of these things could happen in conventional hemodialysis.

            I would also like to point out that we probably need to give some consideration to risk analysis of the problem of unintended consequences of this therapy.

            So, for example, we have identified the potential unintended consequence of worsening of anemia because there is a defined amount of blood loss with each treatment.  And is that overcome by the better management of uremia per se?  And that is a question to be answered.

            We have also identified the unintended consequence of abnormal electrolyte values because there's increased dialysis exposure, but there may be other unintended consequences.

            So, for example, although some of the group favor the use of catheters for this procedure, there are others who would passionately argue that catheters are not a good dialysis access for anyone for other reasons, because of plastic and inflammation and the body of evidence for that.  So perhaps there ought to be some consideration that ensures that all access devices would be useable for this form of therapy.

            And, finally, other unintended consequences on things like drug clearance, the dosing of drugs in patients that are getting vastly increased clearances of small molecules, the dosing of the drugs, and those recommendations will need to be considered as well.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Afifi?

            DR. AFIFI:  Yes.  There's a close connection between this question and the next one, which is study design.  Specifically, risk analysis should be I think, partially at least, aimed at identifying how analyzing the potential risks can help us avoid adverse events.

            And also in my mind from a statistical point of view, we need to classify the risks and the adverse events later when we get to them as to whether they are life‑threatening or not because the kind of thinking I'm sure from the clinical point of view, but what I know more about is from the statistical analysis point of view, the analytical approach would be different for a life‑threatening event versus a non‑life‑threatening event.

            So these issues I think are ones that need to be clarified.  And the connection between this question and the next I advise FDA to think about and spell out more clearly.

            CHAIRPERSON TALAMINI:  Okay.  Thank you.

            Dr. Sadler?

            DR. SADLER:  I have to say that these potential risks are presumptive, rather than demonstrated.  And I believe that they are itemized because of the potential for a worse outcome if they happen than if they happen in the presence of a number of other people.

            And so it would be very important to tabulate these events because I rather doubt that they happen more frequently in this form of dialysis than anywhere else.  They might actually happen less often because there will be more consistent procedure by people who have more motivation to do it right.

            I guess I have to correct George, Dr. Aranoff.  I did not say that I supported increased use of central catheters but simply that because there are benefits of them in this environment, we should see if we couldn't improve that device or we should hope that someone would try to improve it because I suspect we will always need some.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Schulman?

            DR. SCHULMAN:  We'll be at the end of the line here.  Most of what I thought about has been said already.  However, unless and until devices to prevent disconnection occurring at nocturnal dialysis, that is something that probably we have to be concerned about as above the standard dialysis treatment.

            CHAIRPERSON TALAMINI:  Okay.  Thank you.

            Dr. Duffell?

            DR. DUFFELL:  Like Dr. Afifi, I am drawing a connection with this risk analysis to the next question, which is study design.  And I guess the cautionary remark I would make is that some of these things, especially when you're talking about infection rate or psychological effects, are probably long‑term assessments that have to be made and probably not something that you would want to force on a manufacturer to do in a prospective trial prior to approval of these technologies to go out.

            So, in other words, I think they are probably more of an issue for post‑market surveillance, rather than a prospective study to get to market.

            CHAIRPERSON TALAMINI:  Thank you.

            Ms. Moore, pass.  Dr. Hoy?

            DR. HOY:  I don't think that we're seeing disconnections at the patient's access.  However, we've done about 27,000 treatments, and Dr. Lockridge has done 41,000 treatments.  And both of us have seen a patient inadvertently either cross‑‑thread or over‑exuberantly tighten the blood connection at the dialyzer.

            And my patient slowly lost blood until he woke up in the middle of the night.  And the blood dripped down the tubing and fell away from our moisture protector underneath the machine.

            So we've both seen the potential for a very strong person in this case to mis‑cross‑thread the connection at the dialyzer.

            CHAIRPERSON TALAMINI:  Can you bring the microphone in just a little bit?  I'm sorry.

            DR. HOY:  I'm sorry.  I'm sorry.  We have both seen a serious blood loss from a cross‑threading where the blood connections are made to the dialyzer.  I'm not seeing any disconnections from the arms.  I'll be happy to show people what our patients' arms look like.  They're a human factor nightmare.

            But they work.  And we have moisture protectors.  We have end uresis detectors.  And we check them.  And they work.  And we also put an end uresis detector underneath the machines.  And in this particular case, the patient lost two units of blood.  And it wasn't picked up until he woke up and discovered this, and we were very lucky in that respect.

            So I think that there is an issue here.  In 68,000 treatments, we have each seen a cross‑threading or a misthreading or an over‑exuberant threading of the connections at the dialyzer.  You know, how often does that happen in center?  Probably not as often.  And maybe that's something we need to address.

            I'm not seeing inadvertent disconnections at the arms.  I don't think we see increased blood loss from treatments.  Actually, our hemoglobins on our patients stay up and are higher than they were when they first come to us.

            I think there may be a potential increased risk of vascular infection due to button holes.  I think we need to collect the data.  I certainly think there is an increased risk of infections from catheters, despite Dr. Lockridge's incredible data.

            CHAIRPERSON TALAMINI:  I'm sorry?  You said you certainly do think so?

            DR. HOY:  I certainly do think there is an increased risk of infection, no more than we see in the in‑center patients, but I think catheters innately have a higher risk of infection.  So I don't think it's an additional risk for the nocturnal patients, but it certainly is a risk.

            And, finally, I think the psychological effects are ‑‑ I, frankly, am much more concerned about what happens to my patients in the center than I am about what happens to my nocturnal patients.  I think there is some potential adjustment for care‑givers.

            I think that patients who suddenly have a lot of free time do have an adjustment to make, but they don't seem to have much of a problem figuring out what to do with it.

            I am much more concerned about what the isolation and pacification of 300,000 in‑center patients does as they lose control over their lives and their therapies and they're too physically ill to socialize, work, exercise, or make love.  The emotional well‑being that's engendered by feeling normal in the nocturnal patients, by their free daytime hours, by their greater energy, unlimited diets and fluids, and improved cognition more than compensates most nocturnal patients for the loss of the companionship of their chronically ill and dying in‑center fellow inmates.

            CHAIRPERSON TALAMINI:  Thank you.  Well said.

            Dr. Lockridge?

            DR. LOCKRIDGE:  I think it has been well‑said.  I think the bottom line is this is a risk analysis.  And there are risks in doing all of this.  And these are outlined very well.  But the benefit is so much greater for this with the same experience that Chris and I have that this is about patients doing better, feeling better, living normal lives.

            I think we have to have industry figure out a way to prevent disconnects.  I think that is still an important issue, but it is no greater than what happens in this country.  And I don't think it's the risk of putting people at home to do this any greater than in center.

            The blood loss has been resolved as far as just give them enough iron and enough EPO and their hemoglobins are higher than they were before with less EPO.

            The psychological issue I think is to think that people ‑‑ there are a few people that socialize in center, but we are basically doing it to people.  We are making them robots.  We're sticking needles in them, and we're saying, "Come to this time."  By empowering people to care for themselves, we are making their lives become more normal.  They're in control for the first time.

            CHAIRPERSON TALAMINI:  Thank you.

            So I'm going to push the panel just a little bit.  And, again, these are not any sort of official votes, but the question it seems to me the FDA wants to know whether they need to be concerned about the marginal risk of these four things in nocturnal dialysis, as opposed to the way it's practiced.

            So let me just ask for a show of hands?  How many people think for A that there is an increased risk of inadvertent disconnections, a marginally increased risk for NHD, as opposed to normal dialysis?  Anybody?

            (Whereupon, there was a show of hands.)

            CHAIRPERSON TALAMINI:  So we have eight.

            DR. AFIFI:  I'm not voting, by the way, because I'm not an expert in this.

            CHAIRPERSON TALAMINI:  Right.  And, again, these aren't official votes.  We're just trying to get a temperature here.

            How about increased blood loss due to the increased frequency of treatments; again, marginal risk for NHD?  Anybody feel that that's ‑‑ a show of hands?

            (Whereupon, there was a show of hands.)

            CHAIRPERSON TALAMINI:  Four.

            DR. ARANOFF:  It happens, but does it matter?  That's the issue.

            CHAIRPERSON TALAMINI:  Well, the reason I'm not being more specific is that the FDA wants to know, did they need to be concerned about this?  Is this something that they need to be additionally concerned about, particularly with the concept that a document is going to be crafted here?  So I'm just trying to help crystallize this for the purpose of the day.

            Comment, Dr. Moran?

            DR. MORAN:  You just rephrased your own question.  Your first question was, was there a marginal increased risk of blood loss?  The answer to that is yes.

            And then you rephrased it by saying, should we be concerned about it?  And the answer to that is no.

            DR. ARANOFF:  That's correct.  That's right.

            CHAIRPERSON TALAMINI:  Okay.  Fair enough.  Fair enough.

            How about the third, potential increased rate of vascular access infection due to the increased use of access?  How many feel that's a marginally increased risk with home?

            DR. DUFFELL:  Well, but should the question really be again a risk?  Is this what we are worrying about?  Should not the question, though, be restated as before.  Is it clinically relevant?

            CHAIRPERSON TALAMINI:  Sure.  We can phrase it as a clinically significant risk.

            DR. DUFFELL:  Yes, exactly.

            CHAIRPERSON TALAMINI:  How many feel it's a clinically significant increased risk?

            DR. SADLER:  Are we just talking about infection or are we talking about all manner of vascular access problems because I think we have all said that we worry about all of them?

            CHAIRPERSON TALAMINI:  I'm just trying to keep it to what it says in C, "potential increased rate of vascular access infection."  So no hands?

            (Whereupon, there was a show of a hand.)

            CHAIRPERSON TALAMINI:  I've got one taker.

            DR. WEINGER:  I think the fact that patients are doing this and they're doing it way more often has the potential for an increased risk of access problems.

            DR. LOCKRIDGE:  I think that is not a truism, so to speak.  I think they are better than any person that's in center that's sticking, a nurse or a technician that is sticking.  So I think it actually goes the other way.

            And I think there is clear literature that has looked at using these accesses.  It says that there is no difference.  So I don't think it's an issue.

            CHAIRPERSON TALAMINI:  And that's probably the predominant opinion, it appears, of the panel.

            MS. MOORE:  May I?

            CHAIRPERSON TALAMINI:  Yes, ma'am.  Ms. Moore?

            MS. MOORE:  I recognize the experts, and I know that that is the predominant opinion of the panel.  But speaking from a consumer's point of view, even though you may consider the risk minor, I think that it is something that patients ought to be aware of that it is, you know, a risk.  And if something should happen to them, then they would recognize that they had already been forewarned that this is a risk.

            So I guess I would as the non‑expert on this panel disagree with the experts.

            CHAIRPERSON TALAMINI:  Okay.  Thank you.

            Dr. Sadler?

            DR. SADLER:  I think that the point that Dr. Lockridge was making is that they do recognize this is a risk.  They do recognize how important their technique is.  And they do recognize it is their health that hangs in the balance.  And that's why they generally do a better job than our staff in the facility does.

            So, you know, there is a much increased awareness of the importance of every step they take.  And so they generally take it more carefully, with better habits.  And because it's one person doing it every time, I think that there is not really the increased risk.

            And so, you know, my concern is not so much that the stress gets them down, but the euphoria of freedom may affect their judgment detrimentally.

            CHAIRPERSON TALAMINI:  And I would rush to say that simply because this panel doesn't think it represents incremental clinical risk doesn't mean that the FDA is going to forget about it and not bring it up as an issue with these treatments.  But I'm just trying to force the point and crystallize this particular question.

            So the last one, the psychological effects, how many panel members feel that the psychological effects of being home, as opposed to the standard treatment, represents a clinically significant marginal increased risk?  Any panel members?

            (Whereupon, there was a show of a hand.)

            CHAIRPERSON TALAMINI:  Ms. Moore?  So we have one.  Do you have a comment?

            MS. MOORE:  No comment.  Just ‑‑

            CHAIRPERSON TALAMINI:  No?  Okay.  So there were two other issues that were brought up.  And that is unintended consequences and acute decompensation in the water quality.  Do either of those represent significantly increased risks for this that panel members want to comment on further?

            (No response.)

            CHAIRPERSON TALAMINI:  No?  Dr. Lockridge?

            DR. LOCKRIDGE:  Yes.  The one thing that was brought up about antibiotics, I think that because nocturnal dialysis, we do dialysis so much longer, two and a half times or three times more than what we're doing now.  There are certain clinical issues, like we dose vancomycin every other day instead of every five days or every seven days.

            So the actual clearance that does occur, the FDA needs to be aware of this, and the companies need to be aware of it in their labeling, I think.

            CHAIRPERSON TALAMINI:  Okay.  Dr. Hoy, comment?

            DR. HOY:  I think that the dilemma here is that we really don't know what we're doing.  We know that what we do in center doesn't work very well because we haven't changed the mortality rates at all.  But we really don't know what the net effects of doing a to of extra dialysis are.

            And we do see some both potential and detrimental side effects or perhaps they are side effects ‑‑ I don't know ‑‑ of doing a lot more dialysis, especially with large high flux dialyzers.

            For example, we have now started to see a lowering of beta‑2 microglobulin levels in our patients, which doesn't occur at first very much but then later on seems to be coming much more real.

            We're also seeing for the first time, both in center and in the nocturnal patients, lower B‑12 levels, which we have never seen before.  And dialysis patients almost never get low B‑12 levels because it's excreted by the kidneys normally, and they usually have high B‑12 levels.  We're starting to see elevated MCVs and lower B‑12 levels in these patients because we're using these huge high flux dialyzers.

            So there's an issue here of the inadvertent consequence, unforeseen consequence, of this good action, namely using bigger dialyzers.  It's probably a risk for both sets of patients, not just nocturnal.

            We also dialyze off a lot more magnesium.  We dialyze off a lot of phosphate.  What are the net effects of hypomagnesemia chronically?  What are the net effects of hypophosphatemia chronically we've never seen before in dialysis patients?  I mean, we should be lucky to have this problem.

            CHAIRPERSON TALAMINI:  And I would certainly say that in the absence of a lot of hard data, our job here is to offer opinions, which you are doing well.

            I would ask the FDA again, being almost out of time for this question, whether there are other issues with respect to the risk analysis that you would like the panel specifically to address.  And it would need to be quickly.  Any issues?  No.

            MS. BROGDON:  No issues.

            CHAIRPERSON TALAMINI:  Okay.  Thanks.

            So we'll move on, then, to the fourth question, which regards clinical study design.  "The FDA proposes that manufacturers evaluate NHD devices in clinical trials.  The purpose of these studies is to evaluate the incidence of adverse events, the device's ability to deliver prescribed treatments, and the patients' ability to conduct the treatments as prescribed in a home nocturnal hemodialysis setting after appropriate training.

            "Please consider the following aspects of the clinical study design and provide input on the following elements of the study:  a) study design; e.g., retrospective or prospective; b) need for a control group and, if so, appropriate type of control; e.g., prospective, historical, patient at their own control; c) appropriate sample size; d) inclusion/exclusion criteria; e.g., exclusion of patients previously on home dialysis; e) frequency and duration of treatments; f) study duration; g) safety endpoints; h) effectiveness endpoints; and i) length of follow‑up."

            I would turn and ask Dr. Afifi to begin this discussion as our panel expert on these matters.  Dr. Afifi?

            DR. AFIFI:  Thank you.

            I think we should look first at the three items in the second sentence:  incidence of adverse events; the device's ability to deliver prescribed treatments; and, the third one, the patients' ability to conduct the treatments as prescribed or described.

            Item one let me put aside for a moment.  And I think the other two items, the device's ability to deliver the treatment and the patient's ability to do it correctly, we can think of those as appropriate to use a pre/post design; in other words, have patients begin who are currently in clinic settings and follow them up for a while and measure those quantities and then do a wash‑out, as the FDA person earlier today described ‑‑ I think it was Dr. Ruiz ‑‑ and then do the same thing for the nocturnal and then compare the performance, both of the machine and the patient.  So that would be an appropriate thing.

            As far as adverse events, we need to look at them by whether they're life‑threatening or not, as I mentioned earlier.  If they're not life‑threatening, then a count of how many adverse events happen for a given period of time.  And then there are standard statistical methods, plus one or negative binomial and so on, that gets pretty technical.  So I will leave that alone.  But there are standard ways of doing that.

            I think that would need to be a clinical trial with a control group not having the patient as her or his own control.  In that case, historical controls could be useful if the data are available in the form that we need them; in other words, how long for a given period of time.  If not, then a prospective study comparing home to clinic settings with a two‑armed clinical trial would be the appropriate thing to do.

            For the catastrophic event, life‑threatening, and then the event itself would either be death or potential death, the appropriate analysis is what we call survival analysis.  In that case, we don't count how many times the patient died obviously.  We measure, rather, how long until that catastrophic event has occurred.

            And then the relationship of that to whether this is a home patient or a clinic patient and a bunch of other co‑variates to be adjusted for would be the job of the analysis.

            So this will take me more than a minute and a half.  Is that okay, Dr. Talamini?

            CHAIRPERSON TALAMINI:  Yes, yes, sir.

            DR. AFIFI:  Thank you.

            So this is as far as the study design on the options and the potential control groups.  Sample size computations would depend on the particular outcome.  So there is really no recipe as such.

            Each study would need to be handled differently perhaps.  Inclusion and exclusion criteria, I would obviously rely on the clinical people in such studies.

            The frequency and duration of treatment, it seems already the FDA has proposed a standard.  And I think the experts in this area need to review that, whether five to seven days a week.  And I think this is what is meant by this point unless it is something else.

            Study duration and the length of follow‑up are closely related.  In general, we in designing a clinical trial, especially one that will look at length of time until a catastrophic event occurs, we look at the half‑life, if you will, what is the median survival, the median disease‑free or survival time, because we want to come close to actually observing half of the patients have the event occur to them.  Otherwise there would be too much censoring in the terminology of such studies, so some review of the literature to look at how long does it take until ultimately the patient dies and how long does it for half of the patients to experience that.  So that would be my advice on the length of follow‑up.

            Effectiveness endpoints is something I think that is best discussed first by the clinicians but, then, also with input from other experts, such as statisticians, epidemiologists.  And, as I mentioned earlier, I'm glad to see the FDA thinking of using epidemiologists more.  They can't contribute more than just the clinicians or the statisticians can.

            So these are my comments.  And there are more technical things that I don't think we need to go into now.

            CHAIRPERSON TALAMINI:  Thank you very much.  That's very helpful.

            We'll just start from there and go around.  Additional comments on any of those points and perhaps particularly the effectiveness endpoints?  Dr. Sadler?

            DR. SADLER:  Since this is a clinical trial to evaluate a device, when the question of retrospective versus prospective comes up, your immediate assumption is, well, it should be done prospectively.  But if somebody comes with an existing device that has been used a great deal in this arena, I would imagine retrospective data if it is good and complete and accurate would be very useful.

            I don't think we can have an ABA switch with these people because in the first place, if you get people in the dialysis center and they get habituated, they don't want to change anything.

            And if you get somebody dialyzing at home, he doesn't want to come back into the center.  So you take one or the other.  You put them there.  And so you may take historical controls or some sort of additional control group that would be carefully matched.

            I'll leave the size of the group to the statisticians.  And I would think that this could be both people on daily and people on less frequent nocturnal dialysis.

            It certainly doesn't need to go on for an extended period of time to demonstrate that the device is effective, that it doesn't induce user errors, so I would think something like six months, which would get you close to 100 treatments on alternate days and over 100 treatments on every day.  And certainly once that had been established, the time would probably shorten for subsequent devices.

            The safety endpoints certainly should be user errors, should be machine failures, should be false alarms and real alarms, and interruptions of therapy due to device‑related or user mishaps.  And those would also be much of our endpoints, plus demonstrating that the patient's meter exceeds the DOCI guidelines.

            CHAIRPERSON TALAMINI:  Terrific.  Thank you, Dr. Sadler.

            Dr. Schulman?

            DR. SCHULMAN:  I'll just make two comments.  One thing is that the endpoint, one of the endpoints, should be the delivery of adequate therapy.  And I would recommend using a GOCHA standard, Kt/V method of kinetics be considered in such a study.  That I think is an important way to quantify this.  Even though this isn't a survival study, it would show efficacy of the actual treatment.

            And the other comment is that because this study isn't going to ‑‑ you cannot do a really randomized trial in a study such as this.  So there is likely to be bias in the way the patients are selected.  And you are going to have to rely on your statistician to employ the necessary techniques to account for that bias.

            CHAIRPERSON TALAMINI:  Thank you.  Perhaps you could expand on why you couldn't do a randomized trial.

            DR. SCHULMAN:  Well, I mean, you really couldn't take ‑‑ a lot of this, the experience of Lindsay is that when he tried to recruit patients for the study, it was really patient‑motivated, the ones that would want to actually come out of the unit and go into the home and do this.

            And so you really couldn't pick a truly equally randomized group.  You had social issues and different level motivation in the patients that went into the home, these nocturnal therapies.

            CHAIRPERSON TALAMINI:  It seems to me with the right funding mechanism, you might be able to convince patients, but I don't know.

            DR. WEINGER:  Can I follow up on that?  I have heard this from several people, and I am not understanding it.  Since this is only going to be for certain patients anyway, if the entry criteria is those patients who are interested and then you promise everybody "Ultimately you'll get it, but we're going to randomize half of you as the control group to stay in clinic and the other half get it" and then use an intent‑to‑treat approach, I don't understand why that wouldn't be a reasonable design.

            CHAIRPERSON TALAMINI:  Dr. Hoy has a comment.

            DR. HOY:  Can I respond to that?  You know, the NIH and CMS are actually trying to do this.  The problem is that there has never been a successful dialysis study that has randomized people to two different dialysis modalities.

            In other words, we have never compared CCPD, continuous cycle peritoneal dialysis, to CAPD.  We have never compared home hemo to in‑center hemodialysis.  We have never compared transplant to hemodialysis.  And we are now proposing to talk about doing a randomized controlled trial of in‑center hemodialysis against short daily hemo in‑center dialysis and nocturnal home hemodialysis.

            You know, I've got to tell you I'm participating in this because I think it is the gold standard.  And I think, actually, if CMS and NIH can get together, probably you ought to think about trying to get information from them as well.  But there has never been a successful study that has collected patients for two different modalities.

            CHAIRPERSON TALAMINI:  But if it could be done, it would be a worthy goal, I would assume.

            DR. SCHULMAN:  You could do things like develop propensity scores for the patients and do certain techniques like that, but that's the way you get around.  I don't think you'll get a truly randomized trial.

            CHAIRPERSON TALAMINI:  Okay.  Well, Dr. Lockridge?  Then I need to get us back on track.

            DR. LOCKRIDGE:  Right.  I think we really are focusing on the wrong thing here in the sense of a clinical trial that is going to prove efficacy.  I think that in the initial definition, we were talking about a trial that looked at a device that delivered the same capability as another device that already existed and whether it was safe.

            And that doesn't include whether the endpoint is death or not.  It includes whether or not the machine is safe and delivers the same form of therapy.

            So I think we are way out here to think about a six‑month trial or a two‑year trial randomized.  I think we are really looking at a much different trial.

            CHAIRPERSON TALAMINI:  Thanks.

            Let me go back to going around the room, even though we have sort of gotten ahead of ourselves a little bit.  Dr. Duffell?

            DR. DUFFELL:  I don't know if my comment is going to pick up where you left off or not, but my thought behind this is the presumption that a clinical trial is indeed needed in order for a manufacturer to put one of these products on the market.

            Is that a true assumption?  Will the data bring forth a knowledge base that will enlighten us in some way that it is going to really help in the marketability of these products?  Because from the conversations I have had with you all around the table, it seems like everybody wants these things.  They want them pretty much now.

            So, you know, the question you need to ask yourself is, is what we're going to get out of this trial really going to enhance our knowledge and change what we are doing because it is going to delay the availability of the technologies to you all as clinicians.

            So is there a return on this investment, in other words, from a business standpoint, not whether the questions are valid, because the questions maybe could be addressed somewhere else, some other method, such as post‑market surveillance registries and things of that sort, where you gather information and revise labeling based on what you learn that way?  So do you need it prospectively now in order to put these things out there or can we get it some other way?

            CHAIRPERSON TALAMINI:  Let me just let Dr. Sadler make a comment quickly.

            DR. SADLER:  Very quickly.  We already have them.  And it may very well be that the manufacturers will not ask.  And they can't be refused if they don't ask.  And so none of this will come to pass.

            If somebody wants to get it, they will have to go through the hoops, whatever those hoops happen to be.  And, as I said, if it's a machine that has already been out there and has a significant track record, they can probably do it with retrospective data if it's complete and accurate.

            CHAIRPERSON TALAMINI:  Ms. Moore?  Pass.

            Dr. Hoy, further comments?

            DR. HOY:  Just one.  I think there's a huge amount of data from Canada.  And I know it's a foreign country, but it might be worth thinking about trying to see whether or not the retrospective data that we have in the United States combined with the data from Canada actually gives us some ideas about the safety of this method.  I mean, there is a lot of data out there.

            CHAIRPERSON TALAMINI:  Dr. Lockridge, further comments?

            DR. LOCKRIDGE:  Yes.  I'm still missing the point here.  I think it's not about the safety of the modality as much as the safety of the machine.  The safety of the machine and the modality related to nocturnal dialysis, the patient asleep night over night.

            There are over 500 articles written about more frequent dialysis.  Seven years ago, people didn't believe it was better.  It's no question that the nephrology community believes that more frequent and longer dialysis is better.  The real question is, can a manufacturer make a machine that can safely deliver nocturnal dialysis if they apply for that?

            So I think we need to answer that question for the FDA.  What kind of study does that do?  I think it's a prospective.  I mean, similar to what Aksys did with their study, it's eight weeks in center, training period, eight weeks at home, see how many alarms there are, how many times the machine is down, is it safe, does it disconnect.  I mean, it's a pretty simple thing.

            To randomize to approve mortality in the NIH study, they're saying we have to do 1,500 people and follow them for 2 years.  No business people are going to do that to deliver a machine to the community.  We have got to really be realistic here about what we're asking for.

            CHAIRPERSON TALAMINI:  Yes.  Those are good comments.

            Dr. Moran?

            DR. MORAN:  Next time I'm going to sit upstream from Dr. Lockridge.  I do have to agree with him.  I think the study design should be what was done by both machines that have undergone trials for home use, a period of training in center, and then a period being followed in center, and then a wash‑out at home, and then collection of safety data in the home.  So each patient is their own control.

            I entirely agree we will never get a randomized study if we try and do that.  I think the only way to design the study is to have each patient in each arm.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Blagg?

            DR. BLAGG:  I agree with both Dr. Moran and Dr. Lockridge, and I also take up John's point that, after all, we have all of these machines out there now that were never designed to do home dialysis, but they're working perfectly well with home dialysis.

            So if a manufacturer wants to get specific approval for home dialysis, I think the sort of approach that John Moran is talking about and Bob is a way to do this:  brief, simple.

            And the only question I think is, should you use patients who are already home patients to do the study on or should you use patients who are in the center but who are willing to go home?

            I think that may be very difficult to achieve in terms of patient numbers in a reasonable amount of time.  So I think that the minimum study is what we should be looking at.

            DR. GILLESPIE:  And just on a slightly different note, in terms of inclusion and exclusion criteria, I would hope that the age range will be very broad and that the criteria will be very broad.  That would help improve the numbers of the study as well, but it's actually young people that have perhaps some of the most to benefit from this modality and might actually be a larger proportion overall using it because of the benefits on nutrition and growth and on reduced school absenteeism.

            So there is certainly no reason to exclude people under 18 or anything like that from these kinds of studies.

            CHAIRPERSON TALAMINI:  Thank you.

            Let me just take the prerogative of the chair for a second.  So if I hear the majority and what they're seeing, it's that the more frequent dialysis is no longer an unanswered question.  That's a question that has been answered in the literature.

            And, like usual, the key to doing the right study is asking the right question.  And the question here is simply, are the machines safe enough to do it at home?  Is that what I'm hearing people say?

            DR. ARANOFF:  I think you also have to consider, will the specific machine provide the treatment also?  I mean, the modality is proven, but we're talking about individual machines at some time.  So the machines are going to have to demonstrate that whatever that machine is also provides a treatment.

            CHAIRPERSON TALAMINI:  Right.  Dr. Schulman?

            DR. SCHULMAN:  I don't disagree with the suggestion of Dr. Moran about how this study should be done and what the purpose of the study is, but I don't want the FDA to go away thinking that it's a slam dunk that longer dialysis is better.

            The whole nephrology community thought that a higher Kt/V was better.  And in a real trial, the null hypothesis was fulfilled.  So let's not go into that part of things.

            I think I just want to make that cautionary note.

            CHAIRPERSON TALAMINI:  Okay.  It looks like we're going backwards now, but let me get Dr. Kalota, and we'll get back over.

            DR. KALOTA:  Well, I just see it as two separate issues.  One, whether daily hemodialysis is better, longer hemodialysis is better, is a medical issue.  It's not a device issue.  And I thought we were here for the device, which is whether it's safe or not, whether it's comparable or not, not whether to decide whether we should be dialyzing more often, longer, slower, or faster.

            CHAIRPERSON TALAMINI:  Okay.  Dr. Gibson, do you have a comment?

            DR. GIBSON:  I just support the idea of the minimal studies about the device.  That's all.

            CHAIRPERSON TALAMINI:  Dr. Weinger?

            DR. WEINGER:  Yes.  I think we have to separate safety from efficacy.  And if you simply are testing the device, it might make sense to have for efficacy a trial of this new home‑specific device and an older not home‑specific device used in the home.  That would test whether it was as good as existing.

            But safety is a different thing.  And I want to reiterate the importance of useability testing before you try and do efficacy testing.  That, then, gets at the issue of inclusion and exclusion criteria.  The temptation whenever you are doing a clinical trial is to have as narrow a group of patients, the best possible patients so that you can show the best possible result.

            But if you are trying to look at safety in such a clinical trial, you're going to get a best case scenario of safety.  And so you have to pair that with separate useability tests with a broad spectrum of potential users and use environments so that you really understand what the constraints are in terms of safety of the modality.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Moran, further comment?

            DR. MORAN:  I think that is entirely the wrong attitude.  What you are suggesting is that we should test these machines in patients who the doctors don't think are safe to go home.  People entered into this study need to be the people who are potential users of the machine, not the worst‑case scenario.

            DR. WEINGER:  But you misunderstand me.  In a useability test, what you want to know is what are the potential failure modes, errors, and such that patients can make?  And if you pick the best possible patients and you do a limited clinical trial, you are going to pick up some of those things, but you won't pick up a sufficient number that will reduce the risk of in post‑market seeing catastrophic events.

            And that is why useability testing, where you are looking at worst‑case scenarios, what if there is a disconnect, what if there is this and that in simulated environments, mind you.

            Then you can tell, did my design or warnings or whatever you do really ‑‑ is it going to work at preventing a catastrophic outcome.

            CHAIRPERSON TALAMINI:  Well, I don't feel as good about where we came to at the end of that question's discussion because of the complexities involved, but it sounds at least like the panel's opinion is that the prime objective of the study needs to be to test the safety of the machine.

            It sounds like there is a difference of opinion as to whether these should be naive patients or experienced patients.  And is that worth discussing a little bit more for such a study or not?  Dr. Sadler?

            DR. SADLER:  Well, one of the key characteristics of a good machine for home is that it's easy to learn.  So certainly part of the testing ought to be while the patients are training.  And you can't train somebody who is already trained.  So I would think yes, it ought to be incident patients, rather than prevalent patients.

            Again, I think because most of the existing machines are already in use for this purpose and no specific labeling for this purpose has ever existed, I think it is highly unlikely that there is a motivation for the manufacturers to come forward and seek it.  And we're probably just having an intellectual exercise here.

            CHAIRPERSON TALAMINI:  Dr. Afifi, summary comments?

            DR. AFIFI:  Yes.  I think the concept of substantially equivalent that the FDA uses says that, indeed, what we want to do is find the comparison machine, the currently used one in clinics and the new proposed one.  And, therefore, having the patients be their own controls, start there, wash‑out period, then do it at home and compare the results.  I think that that is an appropriate way of thinking to my mind.

            The question of long‑term effectiveness and long‑term safety could I think be thought of as a post‑marketing concept.  And that is the best that I could offer you.

            CHAIRPERSON TALAMINI:  Terrific.  Let me go on to question 5, which is more ‑‑ yes?

            MS. BROGDON:  I believe the staff had one question they wanted to ask.

            CHAIRPERSON TALAMINI:  Okay.  Please do.

            MS. BROGDON:  It was just answered.  Thank you.

            CHAIRPERSON TALAMINI:  Oh, okay.  Terrific.  So question five is with regard to clinical study design, "Please address these additional issues and discuss whether or not they should be considered in a clinical trial:  a) need for dialysate additives, such as phosphate, and monitoring requirements for these levels; b) type of anticoagulant appropriate for home use; e.g., dose, bolus, and monitoring issues; c) type of hemodialyzer membrane and permeability; d) type of monitoring; no partner present, partner awake, no partner but using remote monitoring, or no partner or remote monitoring; e) vascular access choice and location, and risks associated with these; f) practice of hemodialyzer reuse; and g) psychological effects; e.g., impact of treatments on patients, such as loss of social interaction and the effects of the increased responsibility on the patient, and the impact on and the reaction of the family members living in the house).

            I guess what I might suggest that we do as we go around the room is if you don't feel these need to be studied in the context of what we have already discussed, state that.  If so, then state perhaps why it should be part of such a study.

            Let me start with Dr. Sadler and go around so that Dr. Afifi can summarize after this question.

            DR. SADLER:  Okay.  In this particular question, I think that A, B, C, and E are clinical practice and not part of a device investigation.  My own feeling with regard to D is that I would be very reluctant to send anybody home without a partner.  I just think that to do so is to believe that Murphy was wrong.  And I believe that things do go wrong, and that is when the partner is needed.

            With regard to F, for most patients dialyzing at home, reuse is more trouble than it is worth.  And there are effective single use dialyzers these days that they could utilize.

            And with regard to G, that's not really a factor in the device so much as it is clinical practice again.  There are both beneficial and stressful effects on patients at home.  And without observation of the individual patient, it would be inappropriate to generalize those.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Schulman?

            DR. SCHULMAN:  With respect to these dialysis additives, I think it's known that the phosphorous levels, for instance, may go down.  And they're going to have to be supplemented.  That is something the patient is going to have to do.

            And the device, it should be, that behavior should be, tracked how easy it is for them to do, how often, does it affect the curve, because the phosphorous level went to low and so forth.  So I think that that is part of monitoring that should be done in a study design.

            So I do disagree with Dr. Sadler.  With respect to most of the other things he said, I'm in agreement.

            CHAIRPERSON TALAMINI:  Okay.  Dr. Duffell?

            DR. DUFFELL:  I agree with Dr. Sadler's remarks.

            CHAIRPERSON TALAMINI:  Thanks.

            Ms. Moore?

            MS. MOORE:  No additional remarks.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Hoy?

            DR. HOY:  I agree with Dr. Sadler's remarks about A, B, C, and E being clinical decisions that the physicians have to deal with.  The monitoring issue, to me the patient has to be monitored in some way, shape, or form, whether it's by a partner at home or by a virtual partner.  It doesn't matter to me because I think they're both equally safe or equally risky depending on how you look at it.

            The practice of dialyzer reuse is unnecessary and inefficient.  Ask the largest dialysis company in the world about that one.  They don't do it anymore.  It would just interfere with the ease of being at home.

            I do think the issue of the impact of this treatment on other family members needs to be looked at, care‑givers, other care‑givers, spouses, significant others, and children.  It is worth investigating further.

            CHAIRPERSON TALAMINI:  So, Dr. Hoy, let me just pin you down and perhaps Dr. Sadler a little bit more on this.  This issue of monitoring and partners, is this an issue worthy of study or no in a trial that the FDA would participate in?

            DR. HOY:  No.

            DR. SADLER:  I think not.  It's been demonstrated that both work.  There is preference on the part of individual nephrologists who will train patients.  And they will practice in conformity with their beliefs, but the evidence is already out there that people can do it both ways.  It's just a matter of how much of a margin of safety exists.  And I don't think that it's appropriate for this study.

            CHAIRPERSON TALAMINI:  Thank you.

            DR. HOY:  I'm sorry.  One last comment I forgot to make about the issue of not reusing dialysis.  There is the issue of first use reactions.  And we have not run into trouble with that because we recirculate blood for 10 to 15 minutes before we actually hook the patient up.  But that is an issue to be looked at if you're not reusing dialyzers.

            CHAIRPERSON TALAMINI:  Dr. Gibson wants to make a comment quickly.

            DR. GIBSON:  Yes, very brief, about the psychological effects.  I agree with the industry representative, Dr. Duffell, that it should not be incumbent on the industry to prove that this is psychologically safe.

            But if I understand Drs. Sadler and Hoy to say that we should not collect data on the psychological effects on the family during the study, I think that would be wrong because, even though it's my opinion, like theirs, that there will not be any psychological effects that are harmful during this period, I think we should collect the data.  And there are some fairly simple instruments that at least have a validity among the behavioral health people with regard to outcomes versus the process of coping, which is a different story.

            But as far as outcomes are concerned, that data could be collected.  My hypothesis is that it will not show any adverse effect, but I don't know that.

            CHAIRPERSON TALAMINI:  Yes.  I was actually pinning them down on the monitoring issue.  I think Dr. Hoy actually was in favor of studying the psychological aspects.

            DR. SADLER:  Well, if I may comment, I didn't say at all we shouldn't study it.  I just said that I thought it was not something to be generalized on small populations because people will respond in different ways.  It certainly should receive attention because it is always important when we put this kind of a responsibility on a patient.

            CHAIRPERSON TALAMINI:  Dr. Lockridge?

            DR. LOCKRIDGE:  Yes.  I probably disagree a little bit about what should be done as far as the dialysate additives.  I think nocturnal dialysis is different than daily dialysis, short daily, or in center 3 times a week in the sense that it offers 2 to 3 times more dialysis or clearance, clearance of about 30 to 40 cc.

            So in our practical experience, I think that phosphorous is a real issue.  And I think new machine designs to the field should address that and figure out how to make the phosphorous at 1.5 to 2.  In other words, that should be part of the machine design.

            The issue of magnesium that Chris brought up, everybody is hypomagnesemic, I think that needs to be looked at in this document that says what you should do to bring this to fruition as far as the manufacturer.  Magnesium is an issue.

            And the third issue is calcium.  I think they have clearly shown that when you dialyze this amount of dialysis is actually calcium wasting, there are no phosphate binders or no calcium supplement that patients take.

            So, actually, with hemo filtration or removal of the fluids, you're removing calcium.  And people become calcium‑depleted.  So I think that we standardly start people on three milliequivalents per liter of calcium.  The standard bath in center is 2.5.

            So I think calcium is another thing that really has to be looked at.  You really don't see the changes on the bones for a year, but I think we need to have the manufacturers be able to deliver a 3, a 3.25, and a 3.5 calcium bath.  And it needs to be proved that they can do that in this trial because that is going to be needed to care for these people.

            DR. SADLER:  Bob, isn't that incumbent on the concentrate manufacturers, not the machine manufacturers?

            DR. LOCKRIDGE:  Well, it depends on what machine and who is ‑‑ I agree if you're Fresenius and using the standard proportion machine.  If you're Aksys and make your own concentrate or if your renal solutions that are using absorbent cartridge or NxStage you mix the fluid in a bag.  They have to bring that to the marketplace.

            CHAIRPERSON TALAMINI:  Dr. Moran, which of these items should or should not be included in a study?

            DR. MORAN:  Reuse should not be included.  I don't think anybody is doing reuse in the home patients.  I apologize.  I agree with Dr. Sadler.  It's more complicated than it's worth.  I apologize.  I'm upstream of you.

            Water‑soluble vitamins is the other thing I think we should be looking at, too, in this long‑extended dialysis.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Blagg?

            DR. BLAGG:  I basically agree with Dr. Sadler's comments.  The other comment I would like to make, though, is we really need to look or somebody needs to look at what this trial is going to cost the manufacturers to do the more of these things that get put into it because I think that will be one of the reasons why manufacturers may never want to go and have their machines approved for nocturnal hemodialysis.

            CHAIRPERSON TALAMINI:  And that is one of the beauties of being on an FDA panel, that we're not supposed to think about cost.

            Dr. Gillespie?

            DR. GILLESPIE:  Just to comment on the reuse thing, I think it doesn't make sense in the traditional sense with traditional machines, but it's possible that people are going to come up with different technology.

            I know there is one machine right now that reuses not only the dialyzer but the entire circuit.  And it's self‑cleaned each night.  So I think we couldn't write off the reuse question entirely if somebody comes up with a novel technology for it.

            I wouldn't see people like collecting the dialyzers and carrying them back into the center to reprocess, as was done in the past.  Like I said, if the machine is something that is categorically different, then we may have to look at it differently.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Lockridge, did you want to make a comment?

            DR. LOCKRIDGE:  Yes.  The heparin issue is a real issue, too, because because you dialyze people six or seven or eight hours, the design, you can bolus, but the kidney is going to clot.  When my patients forget to cut on the pump, their kidney clots during the night.

            So I think it's FDA should be having the manufacturers have a machine that has a heparin pump on it that will deal with that anticoagulation.  And how that heparin pump works and how it is monitored needs to be monitored in this clinical trial.  And that makes this different from nocturnal.  I mean, nocturnal is different in that way.

            CHAIRPERSON TALAMINI:  Yes.  Right.  That's helpful.

            Dr. Weinger, which issue should be ‑‑

            DR. WEINGER:  I'll pass.

            CHAIRPERSON TALAMINI:  You'll pass?

            Dr. Gibson, further comments on which of these should be included in a study and which not?

            DR. GIBSON:  Pass.

            CHAIRPERSON TALAMINI:  Pass?

            Dr. Kalota?

            DR. KALOTA:  I can't comment on the things specific to the hemodialysis, but I would say for the test, we should have some kind of monitoring present to answer the question as to whether or not we need to have monitoring present.

            CHAIRPERSON TALAMINI:  Okay.

            Great.  Dr. Aranoff?

            DR. ARANOFF:  I think it's less important for us to consider which ones of these individually specifically should be studied and more incumbent upon us to comment about whichever ones are studied for a particular device, that that then be covered in the labeling of that device.

            So, for example, we may have technology where a particular machine would use a dialyzer that has heparin covalently bound to the membrane.  And so it wouldn't need bolus or infusion heparin to keep it one.  That ought to be covered, then, in the labeling for the device as it is studied.

            If a device is studied with both remote monitoring and a partner, if that manufacturer chooses to use their machine and study it in that way, then that ought to be included in the labeling, that this was shown to be safe with a remote monitoring and the partner.

            If, on the other hand, a device manufacturer chooses to study their device without remote monitoring or without a partner and still proves it to be safe, then they ought to have that indication.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Afifi, summary thoughts about these issues and the clinical study question in general?

            DR. AFIFI:  Yes.  First, it seems that none of the items A through G under this area need to be thought of as randomization variables.  In other words, we don't need to randomize do you have a monitor or not, do you have an in‑resident person or not, et cetera?

            I think it would be wise to collect the data on them.  I don't think that will be a burden, an excessive burden.  I think the point that if we then collect those variables and look at them as covariates and perhaps do some stratification or subgroup analysis, we will learn about whether some of these items, some choices for these items, are more helpful than others for the success of the whole operation.

            And, as Dr. Aranoff said, I think these could be helpful conclusions to be included in the labeling later on.

            CHAIRPERSON TALAMINI:  Terrific.  Thank you.

            I would again ask the FDA if there are further issues the panel can address to help with this clinical study design question.

            MS. BROGDON:  I need to confer with the staff for just a moment if you would allow that.

            CHAIRPERSON TALAMINI:  Sure.  There was a housekeeping detail that somebody mentioned about the front desk and needing to ‑‑ okay.

            DR. SADLER:  Well, while we have a gap, let me put one loose comment into it.

            CHAIRPERSON TALAMINI:  Absolutely.

            DR. SADLER:  It's been a thought that has been rattling around since we started that we talk about training laymen to do dialysis at home and dialysis in center being done by professionals.  But those professionals are ordinarily technicians who have a highly varied and relatively short period of training.  The difference is not so much the quality of the people doing the dialysis but the presence of oversight.

            CHAIRPERSON TALAMINI:  Good point.

            Yes?

            DR. RUIZ‑ZACHAREK:  Yes.  We actually wanted to elaborate on the psychological effects.  One of the concerns we had ‑‑ again, this might just fall in the practice of medicine and individual prescription by each nephrologist.

            Our concern was, how stable would a patient need to be to go home with a dialysis machine to perform nocturnal dialysis.  And our concern was on patients who have prior suicidal attempts or some psychological instability.

            CHAIRPERSON TALAMINI:  So let me ask perhaps if four of the clinical nephrologists could address that for us.  What four clinical nephrologists would like to address that question?  Dr. Lockridge obviously would.

            DR. LOCKRIDGE:  I think there are criteria for who should not go home.  We treat nocturnal dialysis as a privilege in our facility.  And we look at it similar to a transplant.

            In a transplant, basically if you're a drug user, you have to be drug‑free for six months with randomized studies.  So that's dealing with the drug user.

            If you're an alcoholic, you have to be alcohol‑free and be tested.  And if you smell of alcohol, it's looked within the document, in six months, you don't get a chance to dialyze, to do nocturnal dialysis.  And if you have mental health issues, you should not be allowed to go home.

            We have found that education is not an issue.  We have taken cards.  We have sent people home with third grade or less education to do it.  It is the desire to do it.  If the patient wants to do it, they can do it.

            So I think that teaching, taking the same approach that this is a privilege and treatment it as if it's a transplant patient and having the same requirements for drugs, mental illness, and alcohol is the thing.

            As far as the issue of stability, we feel that having the social worker makes a social work visit in the home by herself with the family to try to get a feel for the family interaction, which we think makes a difference as far as what kind of input.

            We have trained people that weeks into training the wife said, "the machine or me in the bedroom."  And certainly the person took the wife in the bedroom, instead of the machine.

            So I think there are limitations of what we can do, but I think that is where I would structure it.

            CHAIRPERSON TALAMINI:  Dr. Gibson?

            DR. GIBSON:  I would like to start by saying what I said before, that I would not like to see arbitrary exclusions based on the psychological data because, frankly, psychiatrists are not all that good at predicting who is going to do what in the future.  And nobody is all that good at predicting human behavior with any certainty.

            With that said, I think Dr. Lockridge is right on target that the evaluation needs to be center‑specific, much like transplant programs, where transplant programs do exclude people who are actively using substances and people with unstable mental illnesses.  And that is certainly appropriate to do, but I would certainly argue that there are very likely people with stable mental illnesses who do extremely well on home hemodialysis of any modality.  So that's really all I would like to do, is just make a plea for not excluding people based on what you think is going to happen.

            And there is one other point.  I mean, the active substance abuse and unstable mental illness I would view as exclusions from entering the program, but when people are in the program, of course, they're going to be trained.  And during that period of training, you have an opportunity to evaluate how people are actually doing.

            And my prediction is that, of course, like everything else, we'll be surprised that some people you will predict will just pick this up in a flash and do well, will flunk, and never go home and people that you will predict are just dumb as dirt will learn how to do this procedure just fine and will stay at home and never come back.

            CHAIRPERSON TALAMINI:  Which leads nicely to the next question.  Any other burning comments in response to that?  Great.  We've got a process comment.  We're doing great, but we've got to pick it up just a little bit.  So on to question number 6 with regard to training.

            The training of patients and their partners, if applicable, is crucial in performing NHD treatments.  Please comment on the training needs for this modality in terms of the important aspects to be included in the training program, how long the training period should be, and what criteria should be used to determine if a patient has been adequately trained and is ready to begin self‑care at home.  Please also consider who should do the training and how they should be qualified.

            Dr. Blagg, let's start with you.

            DR. BLAGG:  Okay.  First of all, training needs obviously have to cover various things.  The first thing, in our program at least, is to teach the patient to stick themselves or if family members can do it, to teach them because our experience is it is difficult for patients to learn until they feel at least reasonably comfortable with sticking themselves.

            They need to know about their disease.  They need to know about dietary management and so on.  They need to know how to do the dialysis.  They need to know the complications that are likely to occur and how to handle them, how to contact for support and when to see their physician and so on.

            How long a training period?  In our program, once the patient can stick where we take three to four weeks to train the patient, we train them three days a week.  And the other two days, they come in and go to classes and do other things, like work on the equipment and so on.

            Criteria to be adequately trained.  We have a series of tests of various procedures that the patients do that the nurses do with the patients on a regular basis.  They have a final exam, if you like, on which they have to reach a certain minimum score before they're allowed to go home.

            During the last couple of so dialyses that they do in the training program, they're in the room with the door shut, and they're talking to the staff by the telephone, like you would at home.

            Staff.  I think the important thing ‑‑ I don't think it really matters whether it's a nurse or in some cases a technician.  I think what you need is somebody who has a lot of experience of dialysis and who is capable of being a teacher, not just a treater, and keep their hands off patients if they make mistakes, let patients learn from their mistakes.  And I think that covers ‑‑ how they should be qualified, well, I guess each individual unit is going to decide how to qualify their staff.

            Those are my comments.

            CHAIRPERSON TALAMINI:  I guess I would ask the panel also to think about how much of this falls into medical practice and how much of this should really be ‑‑ I don't want to use the word "legislated" but should be part of whatever is done in terms of the FDA.  As Dr. Weinger pointed out, there are now procedures where training is absolutely essential before you can perform the procedure.  So think about that as well in your comments.

            DR. BLAGG:  I would suggest that it's the physician in the dialysis unit's responsibility to take care of these things.

            CHAIRPERSON TALAMINI:  Dr. Gillespie?

            DR. GILLESPIE:  I agree with everything Dr. Blagg said.  I don't think I could say anything better.

            CHAIRPERSON TALAMINI:  Dr. Weinger?

            DR. WEINGER:  Yes.  I think that the issue of where is the boundary between the manufacturer's responsibility for their specific device and the clinical environment is a very important one.

            The manufacturer certainly to assure success in the use of their device and, thus, the sale of more of them is going to provide general guidance for the clinical aspects of it, but, really, the focus has to be on the device.  And my advice there is that the manufacturer needs to establish clear performance criteria for the key elements and components of their device and perhaps then provide suggestions and some support materials perhaps for helping the clinician attain those performance criteria, helping the clinicians to train the patients to attain those performance criteria.

            But, really, it's about the patient needs.  This would be typically what one would call a useability objective.  Patients needs to be able to tend to and respond appropriately to X alarm in Y amount of time correctly.  Those are the kinds of performance criteria that a manufacturer needs to identify.

            And hopefully they will have in parallel with developing the device developed training materials and tested these things themselves so that they know that those are effective materials to attain those performance criteria.

            CHAIRPERSON TALAMINI:  Dr. Gibson?

            DR. GIBSON:  I like the idea of the performance objectives and agree with everything Dr. Blagg said and nothing else to add.

            CHAIRPERSON TALAMINI:  Dr. Kalota?

            DR. KALOTA:  Nothing else to add.

            CHAIRPERSON TALAMINI:  Dr. Aranoff?

            DR. ARANOFF:  I think Dr. Blagg described home training very well.  And I don't think that nocturnal home hemodialysis with regards to training is fundamentally different than routine home hemodialysis with regards to the training characteristics.

            And I think that could even be applied to the psychological issues that we discussed a moment ago.  A patient who is appropriate for home hemodialysis or home therapy, whether it's home hemo or PD, should be a candidate for nocturnal as well.

            So I don't think this is fundamentally different than the guidance the FDA has already given for training.

            CHAIRPERSON TALAMINI:  Terrific.  I haven't heard much about who should do the training other than the physician.  So perhaps as we go around, we could address that as well.  Dr. Afifi?

            DR. AFIFI:  No comment.

            CHAIRPERSON TALAMINI:  Dr. Sadler?

            DR. SADLER:  I agree with Dr. Blagg that the training should be done by somebody who is interested, a good communicator, a teacher, and well‑experienced in dialysis.  And it doesn't matter what credentials they hold.  They can be very effective.

            I do believe that this represents a function of the dialysis facility.  It's the practice of medicine.  And while it was perfectly reasonable to ask the facility to certify that the patient has been trained, has demonstrated capability, and been tested on this, you shouldn't get into that box to see what's in it.

            To quote Mark Twain, "No generalization is worth a damn, including this one.  And we ought to be entitled to have a little bit of variation."

            CHAIRPERSON TALAMINI:  Dr. Schulman?

            DR. SCHULMAN:  Well, I agree with what has been said already.  I would like the FDA to know that the people who do home training, whether it's CAPP or hemodialysis as it exists now, often among the motivated personnel.  And they generally tend to do a good job and are given the latitude by most units to spend as much time as they need with the patient.

            So I don't think that should be a major concern of the FDA.

            CHAIRPERSON TALAMINI:  Okay.  Terrific.

            Dr. Duffell?

            DR. DUFFELL:  I agree with Dr. Sadler.

            CHAIRPERSON TALAMINI:  Ms. Moore?

            MS. MOORE:  Yes.  I agree with Dr. Blagg and the others who spoke about the training for home care, but I think that the manufacturer should take some kind of responsibility for guidelines for training the medical staff, who will then be responsible for training the persons who are going to be doing this at home.

            CHAIRPERSON TALAMINI:  Okay.  Great.

            Dr. Hoy?

            DR. HOY:  I don't think that the manufacturers do very much to help us train our technicians.  There may be some very broad stuff they give us, training materials, but basically we train our technicians to do dialysis.  And those technicians are actually no different in education or background usually than many of the patients who come to us.

            Several comments about the training process that we use, one, we find that we can't train the first week.  We dialyze them four times a week for the weeks that we do the dialysis.  And we do it very aggressively because by the end of the first week, they're less anxious.

            We're sort of describing what we're doing.  We've also gotten to know them a little better.  We've decided which one of our three trainers is going to be a personality fit with those patients.  And we've got them cognitively improved by the end of the week so they remember what we're saying.

            Secondly, we also do everything we can to chart self‑care in the dialysis unit in center before they come to us.  And we start cannulation training before they come to us.

            CHAIRPERSON TALAMINI:  Great.  Thank you, Dr. Hoy.

            MS. MOORE:  Excuse me.

            CHAIRPERSON TALAMINI:  Yes, Ms. Moore?

            MS. MOORE:  I think I mentioned the responsibility of the manufacturer because I remember being on a prior panel where there were machines and the manufacturer had some very clear training packages for the physicians because each machine then I think is designed in a special way, maybe very simple training, but it seems to me that a physician might need some kind of guidelines when there is something new introduced in his office, such as a new machine.

            CHAIRPERSON TALAMINI:  Thank you.

            I think probably what we're caught up in a little bit again is the difference between nocturnal home dialysis and standard in‑hospital dialysis, where the practices are already pretty well‑established with regard to physicians and others using the machine.  But that is an excellent comment.

            DR. WEINGER:  Can I follow up on that just briefly?  Based on other devices that have been put in the home and certainly drugs that are being used at home, the manufacturer would probably have substantial ‑‑ it's a two‑way sword but probably have substantial liability if they left all of the training of the patient to the clinician.

            So any manufacturer who didn't have a comprehensive patient training program would be in trouble.  So you can anticipate they're going to have that.  And, therefore, we want to make sure that it's effective.

            DR. DUFFELL:  One other comment.  All manufacturers have an FDA responsibility to them to provide adequate instructions for use.  That's just an absolute given.  So they have to be there.

            CHAIRPERSON TALAMINI:  Right.

            DR. SADLER:  But one quick comment.  Now, we get surveyed all the time by several different agencies that come and tell us that we have to follow the manufacturer's recommendation.  I'm sorry, but I was doing home dialysis training before these people started manufacturing machines.  And I really don't want them telling me how to train people for home dialysis.

            CHAIRPERSON TALAMINI:  So noted.

            (Laughter.)

            CHAIRPERSON TALAMINI:  Dr. Lockridge?

            DR. LOCKRIDGE:  Well, Dr. Sadler, I may have to disagree with you here, but there are less than 1,000 people doing home dialysis in this country right now.  There are 300,000 people doing it.  We are rigging new devices to the community that have new ways of doing things.  The Aksys machine is totally different.  The renal solution is the next stage.

            I think it's FDA's responsibility to clearly have the manufacturers produce training manuals that are very detailed and complete.  The CMS as far as who can train or who cannot train, it's clearly stated that an R.N. has to oversee, that L.P.N.'s can train but an R.N. has to oversee a home training program.  That's the standard of care.

            Now, you can certainly use technicians.  You can use L.P.N.'s.  Two of the ladies that train in my unit are L.P.N.'s.  They're very good, but they're overseen by an R.N.  CMS has already covered that.

            But I think that either we are going to get to in this country with just 1,000 people centers that are specialized in home training, like Wellbound, or we're going to have to have standardization of care if we're going to make a dent in this care.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Moran?

            DR. MORAN:  I agree with what has been said.  We have full independent home training centers.  And our first criterion when we hire a new nurse is, are they going to be good trainers.  Experience with a particular machine comes second to that.  We have all three of the new machines in our centers.  And we expect the nurses to be able to teach any patient on any of those machines is a strict policy.  I think that is something else we have to do.

            We don't want to get into the old system, where we had PD nurses and hemo nurses and never the twain shall meet.  We want the nurse to be able to train a patient on PD in the morning and the Aksys machine or the NxStage machine or the Fresenius machine in the afternoon in a different patient.

            CHAIRPERSON TALAMINI:  Okay.  Dr. Blagg, further comment quickly?

            DR. BLAGG:  I would like to disagree with you.  I think if you have enough patients, you have a separate program for PD and home hemodialysis.

            I also think that the physician who cares for the patient probably will have very little to do with the trainings, maybe the medical director of the facility, but we have 45 physicians in Seattle.  And most of them don't know anything about training.  We just hope they will refer their patients to be trained.

            And a final comment that we haven't mentioned is that we believe that, at least every year and preferably every six months, one of the training nurses should go out and see the patient do a dialysis in the home to make sure they haven't acquired too many bad habits.

            CHAIRPERSON TALAMINI:  Excellent point.

            Other questions regarding training?  Let me just ask specifically the FDA whether they would like the panel to address other issues regarding training.

            MS. BROGDON:  I think we have enough information.  Thank you.

            CHAIRPERSON TALAMINI:  Terrific.  Thanks.

            So the last two questions the panel is asked to address are with regard to labeling.  Question number 7, device labeling directed towards the patient should include information on NHD and on the device, including instructions for the use and care of the device, how to deal with alarms and how to run treatments.  Please discuss other important aspects of NHD and the value of including them in the lay user's manual; e.g., treatment, not device risks, psychological effects of the treatments, vascular access, information.

            Dr. Gillespie, would you be willing to start the round?

            DR. GILLESPIE:  That's kind of a tough question to answer right off the top of my head.

            CHAIRPERSON TALAMINI:  We can pass if you'd like.  No problem.

            DR. GILLESPIE:  Okay.  Appreciate that.

            CHAIRPERSON TALAMINI:  Dr. Weinger?

            DR. WEINGER:  First, labeling, just so everybody is on the same page, is not just a physical manual for use.  It includes all aspects, written and electronic, related to the use of the device, including, to some extent, warning and instructions physically attached to or displayed by the device.

            And so the bottom line with regard to the patient, which is the focus of this question, is again that one needs to test that whatever the piece of information is, that it effectively transmits that information to the patient in the condition and situation where that information needs to be used.

            And, as well, I want to make sure and appreciate that there are lots of modalities and for lay people, often instruction manuals in the traditional sense are the least effective way of transmitting that information.

            Again, the drug companies I think have a fair history now of various programs from videotapes and DVDs and Web sites and telephone hotlines, et cetera, that need to be considered in the design of these kinds of instructions, labels, and such.

            DR. GIBSON:  I don't see any value of putting anything on the label about possible psychological effects.

            CHAIRPERSON TALAMINI:  Thank you.

            DR. KALOTA:  Pass.

            CHAIRPERSON TALAMINI:  Dr. Aranoff?

            DR. ARANOFF:  I think in 25 years, I've only had one patient, one home patient, ever quote anything out of an instruction manual to me.

            (Laughter.)

            DR. ARANOFF:  The patients ignore them.  They learn from the training what to do.  And they don't read the instruction manuals.

            However, with the ability now to have instructions embedded into the useability of the machinery, that is where our attention should be focused.

            And the instructions should be simple.  They should be sequential so that patients are not given information that they don't need at the very moment that they are doing something.  That is, if they are setting up the machine, the instructions should lead them through it, "Push the green button.  Now push the red button.  Now push the yellow one.  Now you're ready to start."

            And then once they start their treatment, so on and so forth, and then when they finish their treatment, then come off the machine and clean it out or whatever, that should be sequential.  And fundamentally the machine should diagnose what is going on and tell them what to do about it.

            CHAIRPERSON TALAMINI:  So that is an excellent point.  With that point in mind, let me just ask the FDA or any panel member who knows to what degree embedded software that is actually part of the patient interface can be considered labeling or is it never to be considered labeling?

            MS. BROGDON:  We would consider it labeling, but that's the extent of my knowledge on this.  I can confer with the staff if you want to give me a couple of moments.

            CHAIRPERSON TALAMINI:  No.  I only ask it for the benefit of this discussion because certainly as these machines become more sophisticated, indeed much more of the information that is needed is in the software and not a piece of paper.

            MS. BROGDON:  It is certainly a developing area for FDA.

            DR. WEINGER:  It is part of the user interface and, therefore, subjected to all of the good manufacturing practices and other design controls.

            CHAIRPERSON TALAMINI:  Let's keep going around the room.  Dr. Afifi?

            DR. AFIFI:  I was just looking at the list of things here.  I didn't see water purification.  I'm wondering if that could be added to it.

            CHAIRPERSON TALAMINI:  Dr. Sadler?

            DR. SADLER:  I would like to reinforce what Dr. Aranoff said about having the machine's software lead you through this.  I think that question seven is inevitably tied to question eight.  And the question eight is going to produce a manual for professionals.  And then you turn that over to some experts, who translate it down to a fifth grade reading level.  And that will satisfy the requirements.

            But, as he says, it's not very important to patients because they may keep it as a reference to look something up sometime.  And so it ought to be laid out so that you can find things in it.

            The most important part of the training is the materials that are provided by the person who does the training and by how the machine gives them feedback as they operate it.  So that I believe that doing it as a software function is much more desirable.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Schulman?

            DR. SCHULMAN:  I have nothing further to add.

            CHAIRPERSON TALAMINI:  Dr. Duffell?

            DR. DUFFELL:  I agree with Dr. Aranoff, but I would also add that our professional staff doesn't read manuals either.

            (Laughter.)

            DR. DUFFELL:  I mean, I think we fool ourselves.  It makes us sleep better at night to have things covered there, but the realization needs to be back to the design of the product to make it intuitively obvious.

            CHAIRPERSON TALAMINI:  As soon as the manual becomes complicated, the first thing you do is make a phone call anyway, rather than thumb through the manual.

            Ms. Moore?

            MS. MOORE:  Pass.

            CHAIRPERSON TALAMINI:  Dr. Hoy?

            DR. HOY:  I have nothing to add.

            CHAIRPERSON TALAMINI:  Dr. Lockridge?

            DR. LOCKRIDGE:  Yes.  I think FDA needs to think about this entirely differently.  I think that we need to tie the labeling and the training manual and what is embedded in the machine altogether as one thing.

            I think that we have rewritten our training manual six times now over the last seven years.  We have little cards that we use now.  And when we call the patient, they do refer to the manual if it's used in a training manual.  So to have a label like what the FDA requires on a drug is worthless for the patient.  The patient doesn't use it.

            But if you have a labeling that ties in the training, the machine, and the embedding of the ideas in there, that is very useable to the patient.  And I think that is what we should be moving toward.

            So I would favor the combining of labeling, training manual, and software in the machine.

            CHAIRPERSON TALAMINI:  Excellent point.

            Dr. Moran?

            DR. MORAN:  I think our experience says that having a graphical user interface and a touch screen are absolutely wonderful for training, establishing confidence in the patient.  And in a situation where an alarm occurs, if the machine tells them, as Dr. Aranoff said, what the problem is and what the steps to follow should be, one, two, three, and they don't even have to think, I would point out that if they are in bed at night and they are awakened by an alarm and you've got a graphical unit user interface that automatically lights up ‑‑ I beg your pardon.  I was saying if they are in bed at night, if they are awakened by an alarm, if you have a graphical user interface that automatically lights up, and they're not fumbling for the light switch and everything like that, they can just reach out and do the appropriate ‑‑

            CHAIRPERSON TALAMINI:  Great.  Thank you.

            Yes?

            MS. BROGDON:  I just wanted to say to Dr. Lockridge we support the evolution of user and patient labeling.  We do review all of that information.  And we try not just to require paper labeling just because FDA has always gotten that.  We're open to more innovative types of labeling and also then training.

            CHAIRPERSON TALAMINI:  Dr. Blagg?

            DR. BLAGG:  I agree very strongly with Dr. Moran about the touch screen, and I have nothing else to add.  I agree with particularly Dr. Aranoff.

            CHAIRPERSON TALAMINI:  Great.

            Dr. Gillespie, further comments?

            DR. GILLESPIE:  I agree with what people have said, that the importance of printed manuals is dwindling in a world where more interactive types of things are becoming commonplace.

            I think also part of the problem with printed manuals is that a lot of them are really not readable and not very useable, even by professionals.  And that's why people don't use them.  So certainly we support any progress in making things that are more user‑friendly.

            The question also asks about other aspects of nocturnal hemo and whether those should be included in the user manual like just general treatment risks and psychological effects and vascular access.  And I don't think so.

            I think the manual should stick to just the machine and try to keep it concise.  And those other areas, like things about access are a separate issue that people should discuss with their health care professionals or with other materials made for that purpose.

            CHAIRPERSON TALAMINI:  Okay.  Summary comments about the patient labeling because in a moment we're going to discuss physician labeling?  Dr. Weinger?

            DR. WEINGER:  Yes.  I just want to follow up.  Some of my colleagues provided a reasonably effective design solution to a specific user need with regard to specifying a touch screen and such.  I think we need to step back from that because especially the FDA's job isn't to design devices for manufacturers but, rather, to see if manufacturers designed devices effectively for use.

            And so the important thing is for the manufacturers to have an effective human factors engineering process where up front they identify what are the key issues with regard to labeling and instructions and the ability of the patients to respond to alarms, for example, and then to establish performance criteria and then, whatever their design solutions are, to demonstrate those solutions meet the performance criteria.

            One can be misled.  I'll give you an example.  AEDs we all know.  And the human factors folks put them forward as examples of here are devices that were designed with human factors engineering involved so that eight‑year‑olds can do defibrillation of strangers in airports.

            But we have some data that shows that the voice, which is a very powerful guide to getting the end user of that device to do what needs to be done, is, in fact, too powerful sometimes.  And we have videotapes of providers who are shocking people over and over and never getting around to pushing on the chest and actually getting some circulation going.

            So, again, you have to really evaluate your solutions with the end goals in mind.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Lockridge, did you have a comment?

            DR. LOCKRIDGE:  Yes.  I think that I have a mixed feeling about this.  We talked about how much information should be in the manual, the patient's manual, about NHD and how much should not be.  And I feel strongly that the physician and the patient should be making that decision.

            And I think that that is what you were saying, Dr. Gillespie, that this is a clinical decision.  But I think we need ‑‑ you know, how much do we do?  How do we do that?  And I don't totally know how to do that at this time.  So I'm mixed.

            CHAIRPERSON TALAMINI:  Okay.  Questions from the FDA for the panel with regard to patient labeling?

            MS. BROGDON:  Dr. Mendelson had a comment he would like to make on human factors.

            CHAIRPERSON TALAMINI:  Okay.

            DR. MENDELSON:  My comments will be concerning labeling and human factors in general, what our activities are at FDA.  I guess I want to reinforce what Dr. Weinger said.

            I talked about a process when I did my talk.  And I want to emphasize that we try not to be burdensome and prescriptive.  We never tell a manufacturer what shape knob to use, what type of sound a user should hear.  What we like to do is look at the manufacturer's process.  So if they're labeling, if their design works according to their human factors evaluations, that is satisfactory to us.  We don't tell them what to do.

            CHAIRPERSON TALAMINI:  Great.  Thank you.  Appreciate that.

            Okay.  Final question, number eight, with regard to physician labeling, the physician labeling, prescribing information, typically includes the indications for use, contraindications, warnings and precautions, instructions for use, and clinical data on the use of the device.

            For NHD, there is other important information that clinicians need to know, such as the need for alarms that may not be part of the NHD device; e.g., circuit disconnect alarms, fluid leak or moisture detection alarms, the need for a partner or for remote monitoring, vascular access requirements, and the need for dialysate additives; for example, phosphorus.  Please discuss whether or not this and other treatment information should be part of the physician labeling.  A lot of these are issues we have already discussed, but this is with respect specifically for the physician labeling.

            Dr. Weinger, would you like to begin?

            DR. WEINGER:  Sure.  I have two points I wanted to make.  One is that one of the things it seems to me ought to be part of this is guidance to the clinician on patient selection and evaluation.  My colleagues on this side of the table have vast experience, but that needs to be transmitted efficiently to those who are trying to do this for the first time.

            The second thing that I haven't heard about and I think needs to be in this area is a proper way of evaluating the home environment, which includes a variety of issues that have been mentioned in our early presentation beyond water quality, which I heard about, but lighting, noise, distractions, clutter, other stressors, and issues of when and who and how that home environment evaluation would occur.

            And then, finally, I just want to put on the record because I didn't hear it mentioned and I think it may be important in terms of the design of the device for the patient to respond to a crisis situation, and that is fatigue and circadian effects.  And I heard somebody early on having sleep people involved in the conversations, but that may actually be a real design issue.

            CHAIRPERSON TALAMINI:  Great.  Thank you.

            Dr. Gibson?

            DR. GIBSON:  Actually, I find this question difficult and don't quite know how to respond to all of this.  So I think I'll pass for that.

            CHAIRPERSON TALAMINI:  All right, sir.

            Dr. Aranoff?

            DR. ARANOFF:  Well, I'm not sure exactly how much comes under physician labeling and would fit into a technical manual, for example, of the issues that would be very device‑specific.  For example, the home environment was mentioned.

            Different devices will require different water pressure.  For example, whether that goes in the physician labeling, or whether that is a technical specification in an owner's manual of some sort, I don't know but that sort of information.

            I think in this, the one comment I would like to make about physician labeling is that it ought to reflect how the device was proven to be safe and effective.  That is, it should describe the situation in which it has been safe and effective.  And I made that comment about monitoring and so forth earlier.

            CHAIRPERSON TALAMINI:  Dr. Afifi?

            DR. AFIFI:  Nothing to add.

            CHAIRPERSON TALAMINI:  Dr. Sadler?

            DR. SADLER:  With regard to the additional points made in this question, my answer is no.

            CHAIRPERSON TALAMINI:  Thank you.  That's clear.

            Dr. Schulman?

            DR. SCHULMAN:  I think some of the issues that are listed here, for instance, of phosphorus down in this case, don't necessarily have to be part of the labeling for the physicians.  Just some of these functions should take place as a CME type of a thing, you know, where we learn the procedures from experienced people, like Drs. Lockridge, Hoy, and Blagg.  But I think these things don't have to be labeled for the physician.

            CHAIRPERSON TALAMINI:  Okay.

            DR. DUFFELL:  I agree with Dr. Aranoff.

            MS. MOORE:  I pass.

            CHAIRPERSON TALAMINI:  Dr. Hoy, what should be in the physician labeling?

            DR. HOY:  I have to tell you I have never read a physician label on a machine or a dialyzer, for that matter.

            CHAIRPERSON TALAMINI:  Don't say that, for goodness sakes.

            DR. HOY:  I know.  I know.  I know.  Shocking.  But I think that tells you something.

            Secondly, anything that you would put in there that has to do with the actual practice of nocturnal dialysis will probably be out of date by the time it's bought off the shelf.  So I think it's not worth anything more than just doing what you would normally do for the device.

            CHAIRPERSON TALAMINI:  Okay.  The fact that it goes out of there quickly might not mean it shouldn't be done in a manner that could be updated in a timely fashion.  So I wouldn't necessarily ‑‑ you know, the one doesn't preclude the other is what I suggest.  It could be on a Web site or something.

            Dr. Lockridge?

            DR. LOCKRIDGE:  I have mixed emotions about this.  I still go back to 1,000 people in this country doing some type of home hemodialysis.  I go back to the fact that there are only two training centers in this country that train people to do nocturnal dialysis.  And I think there are some differences.  Three.  Excuse me.  There are some differences in nocturnal dialysis as far as the calcium, the phosphorus, and stuff.

            Maybe it's not in physician labeling, but I think the companies need to be responsible.  If they're getting approval for a nocturnal machine, I think that somebody has got to be responsible to know that the people who are using those machines understand the differences.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Moran?

            DR. MORAN:  I have a mixed response.  Looking at the list there of the examples, I think physicians do need to know that they do need to know some sort of full legal motion detection if that's not part of the machine.  I think they do need to know the need for a partner for remote monitoring.  The vascular access requirements I don't think are necessary.

            And I think the need for dialysate additives, as has been suggested, there are a lot of people who don't know a lot about nocturnal hemodialysis.

            CHAIRPERSON TALAMINI:  Okay.  Thank you.

            Dr. Blagg?

            DR. BLAGG:  I agree.  Nothing else to add.

            DR. GILLESPIE:  And I agree with what has previously been said, too.  I think you have to base the label on whether it has been proven to be safe.  And if that involved adding a third party item to the setup, then you have to let people know that you are doing that.

            With regard to ‑‑ what was the other thing?  I'm sorry.  That's all I have to say.

            CHAIRPERSON TALAMINI:  So it sounds like the predominant view on the panel is that these details should not be part of the physician labeling.  Is that what I'm hearing?  Dr. Lockridge, you had a comment?

            DR. LOCKRIDGE:  I was going to comment about the sleeping issue that was brought up.  Actually, there are studies now that show that the rotational sleep pattern that occurs in normal people will go away with in‑center hemodialysis three times a week.  And it actually comes back to normal terms with nocturnal.  So the sleep disorders actually improve.  So it's an enhancement or help for people to respond appropriately when they deal with it at night.

            DR. WEINGER:  Though if they're like me, if they get paged at 3:00 in the morning, their response is not going to be as sharp as if they're awake in the middle of the day.  That was the point I was making.  They may be normal, but that's still an issue because it's happening in the middle of the night.

            CHAIRPERSON TALAMINI:  Any further issues regarding labeling that the FDA would like the panel to address?  It has to be fairly brief.

            DR. MITCHELL:  This is Dr. Dianne Mitchell.

            I just want to clarify.  I think what I heard was a comment to the effect that it would be, somehow or another, helpful to make a statement about that the physicians need to have an appropriate understanding of what nocturnal home hemodialysis is in comparison to conventional hemodialysis because the concerns are a little bit different between the two.  And that is actually something that we can on some level put in the labeling.

            Now, my interpretation of the conversation was that that would be a reasonable thing to do, but I didn't hear that directly.

            CHAIRPERSON TALAMINI:  Thank you.

            So let's put that quickly to the panel, whether physician labeling for a device for NHD should contain information on the differences between standard dialysis and NHD.  Is that ‑‑ no?  Help me.  I want to make it as clear as I can.

            DR. MITCHELL:  No.  We can't say, "You must be a board‑certified nephrologist in order to use this machine," but we can say things like "You must have experience or training in nocturnal hemodialysis in order to prescribe this."

            CHAIRPERSON TALAMINI:  Okay.  So should physician labeling contain that sort of a statement?  Dr. Lockridge, what is your opinion?

            DR. LOCKRIDGE:  No.  I think FDA is going too far the other way now.  I would not have them say that unless you had some type of training because what is determined is training.  Is it going to a CME class about this?

            I think that the responsibility is to let the physician known in some manner, the manufacturer know where it's different.  And the difference is related to the base.  The difference is related to the monitoring and those issues.

            So there is a difference that needs to be spelled out somewhere.  I'm not sure it needs to be in the labeling, but it needs to get across to the physician.  But to say that you can't have somebody on home dialysis because ‑‑ I just don't think we want to get into that.

            CHAIRPERSON TALAMINI:  Is there a panel member who thinks physician labeling should contain such a statement?

            DR. SCHULMAN:  I think that a general statement like when we use cyclosporin or agents like that, that the person prescribing should have some experience with immunosuppressant agents, I mean, that said and not more than that.

            CHAIRPERSON TALAMINI:  Other opinions from the panel regarding whether some sort of a statement ‑‑ now, obviously, we're not going to wordcraft the statement, but should labeling address this issue?  Dr. Weinger?

            DR. WEINGER:  I mean, I suspect that there is some kind of statement like that on a regular hemodialysis labeling.  We're really talking about how is this different from conventional.  And I have to agree fully with Dr. Lockridge that outlining what are some of the differences or at least providing references to documents that outline those differences would be reasonable.

            CHAIRPERSON TALAMINI:  Okay.  Terrific.

            That is great work on the panel's part.  I would like you to be thinking about the summary statement.  After our public comment, I want to get each panel member to make a brief summary statement.

                OPEN PUBLIC HEARING

            CHAIRPERSON TALAMINI:  But we do need to go to the open public hearing portion.  Now that the panel has responded to the FDA questions, we will proceed with the second open public hearing of this meeting.

            Prior to the meeting, we received one request from Rod Kenley of Aksys, Limited to speak during the afternoon open public hearing.  And I need to read the disclosure statement before having that comment.  Is Mr. Kenley here?  Yes.  Okay.  Great.

            So let me read this statement, "Both the Food and Drug Administration and the public believe in a transparent process for information gathering and decision‑making.  To ensure such transparency at the open public hearing session of the advisory committee meeting, FDA believes that it is important to understand the context of an individual's presentation.

            "For this reason, FDA encourages you, the open public hearing speaker, at the beginning of your written or oral statement to advise the committee of any financial relationship that you may have with any company or group that may be affected by the topic of this meeting.

            "For example, this financial information may include a company's or group's payment of your travel, lodging, or other expenses in connection with your attendance at this meeting.

            "Likewise, FDA encourages you at the beginning of your statement to advise the committee if you do not have any such financial relationships.  If you choose not to address this issue of financial relationships at the beginning of your statement, it will not preclude you from speaking."

            So with that having been said, I will invite Mr. Kenley to make his public comment.

            MR. KENLEY:  Thank you.

            Yes.  My name is Rod Kenley.  My credentials are that I have been on the manufacturing side of the business now for about 28 years, about half that time with Baxter and the second half with Aksys Limited, the company that I founded back in 1991, with the specific intent to manufacture and design a machine for daily hemodialysis, daily home hemodialysis.

            I would like to first start off with advising both the panel and the participants from the FDA to think about the broad picture effects on the industry in general and innovation, in particular.

            When I started Aksys Limited, that was the first occasion in the business of kidney dialysis that there was a start‑up company funded by private money.  It wasn't a corporation that was preexisting in other areas that got into the dialysis business.

            When I was out trying to raise the capital to start this company, I met with a lot of potential investors.  And I got one question from every single one of them.  Will this require a clinical trial?  What will be the length of the FDA review process?  Because they always translate that back to a return on their investment.

            My answer was always the same.  There has never been a clinical trial required in the history of mankind for a hemodialysis machine.  Unfortunately, I was proved wrong.  We became the first machine that was required to do a clinical trial.

            We figured it cost us about $70 million.  It kept us off the market for an additional two years, in addition to the cost of the actual clinical trial.

            If you require clinical trials for each small step forward in a modality change, you've got to think about the effects on what we're all supposed to be here trying to do, treat these patients better, because it's only through these kinds of innovations that these patients get treated better.

            Dr. Lockridge made a comment that before 1996, almost nobody was talking about daily dialysis.  And when we did start talking about it in '95 and '96 at the annual meetings, there was universal skepticism except for a few early adopters, like the doctors who are present here.

            Now it's basically universally accepted that that is a better form of therapy.  So we've got to ask ourselves, is what we're doing creating a higher risk of keeping patients off a more beneficial therapy than it is in reducing risks that if they go on that therapy, they're going to be injured?

            Now, to that point in specific, I would like to talk about hemodialysis machines that have been around now for four years.  They are some of the safest medical devices that you can possibly imagine.

            We do 45 million treatments in this country alone every year with an extraordinarily low incidence of patient injuries related to the machine.  Every hemodialysis machine that is on this market fails safely.  Absolutely it fails safely.  There is almost no way a patient can interact with a machine that will cause injury to themselves.

            I can challenge people to describe how that can happen.  We can debate that.  But next I would like to take on some of the specific issues that have been talked about here in this afternoon's session.

            With regard to requiring a clinical trial to prove the safety ‑‑ and I think we establish that efficacy is not at issue here.  If you use a dialysis machine to treat a patient that has no kidney function, it's efficacious.  The question is, is it safe when it's used?  So any clinical trial would really be looking at whether a particular device according to its indications for use is safe.

            Blood access, the first item here is additional safeguards to prevent blood access disconnections, completely unrelated to a hemodialysis machine.

            There is no way that my machine nor any other ever built or designed can detect a venous needle disconnection.  If it could have been done, it would have been done a long time ago because you could sell a lot more machines.  The simple fact of the matter is you cannot detect the change in venous pressure if the needle comes out of the access site.

            You have a window of venous pressure in which the machine is allowed to operate in order to prevent these nuisance alarms.  By the way, there's no such thing as a false alarm.  If there's an alarm, it's because it hit a preset limit of some kind.  There are nuisance alarms.

            If you set those pressure alarm limits too tightly, you'll get them going off all the time.  You've got to operate within this range of pressure.  And it's within that same range that you could get any kind of a change in the venous pressure.

            If you expect the dialysis machine manufacturer to develop a device that identifies that blood is now dripping off the patient's arm, like these enuresis pads, I think you'll be waiting a long time.  Consider the product liability issues there.  Those things are not 100 percent effective.  What if a patient bleeds out and we manufactured that device and it didn't alarm soon enough or at all?

            Also, it's not really related to the practice of hemodialysis.  When you build a hemodialysis device, you build a device that sends blood through an artificial kidney and dialysate through an artificial kidney and assures that it's done safely.  The pressures are safe.  The temperatures are safe.  The conductivity is safe.  There is no air going to the patient.  There is no blood coming out of the dialyzer.

            That's all we can do for you.  To ask us to assure that the venous needle stays in the patient arm is not part of our deal.  I wish it were.  If it could be done, I would have done it before.

            Software incorporated in the device, allowing connection to the internet, nice to have.  It might help me sell more devices.  It's not a minimum requirement to treat a patient safely on nightly dialysis, same with central monitoring, as we have heard, I think, multiple occasions.

            User‑friendly design.  Who will design a study that quantitatively determines that your machine is now safe because its user‑friendliness is above a certain hurdle or it's not safe because it's below a certain user‑friendly hurdle?  As I said before, there's almost no way a patient can interact with one of these machines and hurt themselves, fortunately.

            The quality of water.  To the degree that that is integrated with a machine ‑‑ and it's rarely integrated.  It happens to be in our machine.  I think the issue is valid in terms of how much extra water they see.

            I agree with most of what was said around here.  The current standards are adequate.  What you should be more concerned with is the quality of the dialysate.  That is, after the water is mixed with the dialysate concentrates and what is the inflammatory stimulating quality of that solution?

            Yes, Dr. Sadler, there's not conclusive proof, a whole lot of suggestive indications, that the lower the inflammation stimulation on these patients, the longer they'll live without amyloidosis, the longer they'll live without atherosclerosis and malnutrition.

            As far as a clinical study design, we have to, as was said here, prove that we are substantially equivalent to another device.  Let's take my example.  I have now made a device that has gone through a clinical study and has been on the market for two years with essentially no patient adverse events that is cleared with an indication for daily home hemodialysis.  It's used any number of hours.

            Most people have told us that our graphic user interface represents the gold standard in terms of communicating with the patient clearly and directly, as you have all suggested you would like to see with a big flat panel display and a touch screen.

            That is all well and good, but when it comes to now going through a clinical study to prove that my same device is used safely for nocturnal use, what more do I need to prove?  Currently my device can be used for one hour, two hours, three hours, four hours, five hours, six, seven, or eight hours.  In fact, it is already.

            What is it about a hemodialysis machine that is less safe in the fifth, sixth, seventh, or eighth hour than it is in the second, third, or fourth hour?  The answer is nothing.  It cannot get less safe.  It already is as safe as it can possibly be.

            And with regard to the dialysate prescription, that has to be individualized according to the device.  Our device happens to use a batch.  All 50 liters of dialysate that you're ever going to get are there at the beginning.  That means ‑‑

            CHAIRPERSON TALAMINI:  Excuse me.  I'm sorry.  About one more minute.  One more minute.  So I just want you to have an opportunity to summarize.

            MR. KENLEY:  Thanks.  That means that the phosphate removal will not be as large as if you used a single‑pass device.  So, in fact, what we found in our nocturnal treatments is that the phosphate removal is almost optimal.  So there may not be a need for phosphate.  And that is a different device.

            The dialysate are regulated under their own 510(k)'s as their own devices.  And those are stand‑alone from the machines for the most part.

            Psychological effects are clearly not related to the machine.  Increased rate of vascular effect infection, not related to the machine.  Increased blood loss, not related to the machine.

            Training, who is going to design the experiment that says what is the cutoff for good training material for us to train the clinicians versus inadequate training material?  It's completely subjective.  And I have run into many occasions where physicians have told me, "Don't tell me how to train patients.  I know."  Clearly they do know.  What we have got to do is send our clinical nurse educators in there to train their trainers on how our machine is used, not how to do dialysis.

            So, in summary, I appreciate the opportunity to address the panel.  And I hope you would consider the suppressing effects that any form of regulation can have on the expansion of new technology and innovation in any industry but particularly in this one.

            Thank you very much.

            CHAIRPERSON TALAMINI:  Thank you, Mr. Kenley.  Is there anyone else here who would like to address the panel now?  Please raise your hand if so.

            (No response.)

            CHAIRPERSON TALAMINI:  It looks like we have nobody else who wants to make comments.  Since there are no other requests to speak in the open public hearing, I'd now like to ask for final comments from the panel and from the FDA.

            So if each panel member could give a minute or two summary on today's proceedings and the issues that we have discussed, we would be deeply appreciative.  Dr. Hoy, perhaps you would be willing to begin.

                7.  FINAL COMMENTS

            DR. HOY:  I'm actually going to be much more specific.  Oddly enough, we haven't talked about the machine issues.  And my staff and my patients were very clear that I was to bring up four issues that they were concerned about that have to do with these machines.

            One is the transducer protectors tend to leak.  They're designed for four‑hour treatments.  They tend to leak.  You have to put a second transducer protector on to get through the night.  So please design them so they fit more tightly and don't leak.

            Secondly, the machines are a little noisy, especially mixing water.  They'd like them less noisy.  They'd also like the ROs less noisy.

            Thirdly, pH and conductivity need to be able to be checked independently by external meters preferably.  We're also concerned about redundancy because in the Fresenius 2008H, there is only one conductivity meter.  In our code machines in our dialysis center, there are four in‑series conductivity machines.  So it's much more fail‑safe.  We'd like more redundancy.

            And, finally, our patients are tired of mixing bicarbonate in containers.  They would like to develop machines that use the bicart, cartridges that are designed to last for eight hours.

            I have nothing more to say.  Thank you.

            CHAIRPERSON TALAMINI:  Again, the privilege of the chair.  If I could just ask for the opinion, particularly of those of you with the most experience in this.  If you were to look down the road five years, what you might see in terms of numbers of patients with this sort of a therapy.

            DR. HOY:  I think there are two questions there.  One is how many patients can do it.  I believe that about 20 percent of our patients in our unit could do that.

            Now, I think the issue of how many are going to be doing it five years from now really depends on how realistic physicians get on how to run a business because physicians think that if you're not being reimbursed at a profitable rate for one unit, then it's not worth doing it.  In fact, you have to make a bunch of widgets before you actually start to make profits.  And although we're a not‑for‑profit, our operating margin is real.  So I think there are issues there about the way that physicians look at how one does this.

            This modality is perfect for the insane health care system we work in, which is so labor‑intensive and capital‑intensive because our staff costs are half or less of what they are in center.  And the supply costs are the one area of price elasticity that is available in the future.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Lockridge?

            DR. LOCKRIDGE:  First of all, I would like to thank the FDA to have me here to make comments.

            Second of all, I do think that more frequent daily nocturnal dialysis is the future.  If you had a disease that the incident year that you had that disease, 23 percent of people would die and thereafter every year 17 percent would die and that, in fact, is still increasing and has not improved over the last 15 years and that we were spending 8 percent of the monies that were spent on Medicare for less than .5 percent of the patients, I feel that we as nephrologists, we as the FDA, we as the government have a responsibility to reach out and to seek new ways of doing things.

            This is here.  It's acceptable.  It is a better way.  We need more, better technology.  Training people how to use machines that are in‑center machines is not acceptable.

            I think the FDA needs to be wise in trying to figure out how to approve quality of one machine versus the other and safety, but we need to get on with trying to have new technology come to the market.

            If we had new technology come to the market, I'm planning on ‑‑ about 11 or 12 percent of my population now are doing nocturnal dialysis.  My goal is to have 25 percent in 2 to 3 years.  If I have new technology, I feel that I can easily reach that.

            Can you imagine being 35 or 40 years of age, have a high school or less education, working a factory, where you're making reasonably good money and have hope you get end stage renal disease and your disability check is $600 and you are trapped to come to an in‑center unit 3 times a week to feel lousy?  That is going on across this country on a routine basis, 300,000 people.  We have got to make a difference in what our people are seeing and spend their money wisely.  It is my money.

            Thank you.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Lockridge.  It's well‑said.

            Dr. Moran?

            DR. MORAN:  I agree with a number of at least 20 percent of people should be going home.  Whether they will be going home in five years is another issue.

            I constantly hear people say that in this country to so many people that couldn't go home.  And that's absurd to say that.  There are so many people who couldn't have a transplant, but that doesn't mean we shouldn't be doing the maximum number of transplants we could do.  The more people we get home, the better for the system and especially for the patients and the taxpayer.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Blagg?

            DR. BLAGG:  Again, thank you for asking me to come.  Just a couple of brief comments.  One comment that I think has some bearing on everything we have talked about today is a comment that Dr. Scribner made in about 1965, which was basically that with any chronic disease, the more the patient knows about the disease and its treatment and the more responsibility he's able to take for that, the better rehabilitated he will be at every sense of the word.  And I think that is what we are trying to do.

            I would hope that as a result of what we have said and done today, that the FDA in its wisdom, whatever it decides to do bears in mind that we have to change the numbers of patients.

            Bob may get to 25 percent.  I agree.  I think 20 percent of more of patients could do it with the right equipment and everything else.  The problem is, though, besides providing the right equipment, we've got to make the large dialysis corporations much more enthusiastic towards this system.

            We've got to educate physicians, most of whom have never seen home dialysis.  They just think it's a strange thing that goes on in Lynchburg or in Seattle or somewhere.  And we have to educate patients that this is one of the best things for them.  And I just hope that whatever the FDA comes up with will not serve as any sort of deterrent to increasing the numbers of patients.

            Thank you very much.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Blagg.

            Dr. Gillespie?

            DR. GILLESPIE:  Yes.  I think we have covered a lot of ground today.  And I think certainly I feel like there is a lot of potential for this as being one modality among several that we really need to move forward beyond the thrice weekly paradigm, which is really more of a 1970s standard of care.

            And so whether it's nocturnal or short daily or peritoneal, there are a lot of options.  And this is one that we need to look at.  I think we certainly have the technological ability to do it with the state of our technology today.

            What we really need now is the will to transform what our patients think, the doctors, the dialysis companies, the manufacturers, everybody.  There are a lot of different pieces to that puzzle, but we certainly need to keep pressing on for that, anything we can do to try to encourage more innovation in this, whether it's adapting existing equipment to work more optimally for this or whether coming up with completely new equipment is a goal we need to strive for.

            CHAIRPERSON TALAMINI:  Thank you, Dr. Gillespie.

            Dr. Weinger?

            DR. WEINGER:  Thank you.  This is a very exciting opportunity for society to really improve quality of life and potentially outcomes for patients with a serious disease, but because we're talking about lay persons using a potentially lethal device and process in their homes with limited or more limited support, it really kicks it up a notch.

            And I feel that the FDA and the manufacturers, if for no other reason than for their own self‑protection but regardless, need to look at these devices a little more rigorously than existing in‑center devices.

            That doesn't mean that the FDA should stipulate how these things should be designed but, rather, to ensure that the manufacturer has a rigorous design process, which in this case particularly focuses on the human factors engineering principles and includes, as specified in existing national and international standards, key aspects of the design process, including a user‑centered process; that is, patients and nephrologists are actively involved in the design of the devices, that there is iterative design and evaluation, and that there is functional; that is, performance‑based testing of the user interface to assure that use errors are minimized.

            In addition, I think it's critical that both representative users and the use environment be included in the evaluation, the design considerations and evaluation, of these devices.  And it includes not just the design of the device but the training and labeling as well.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Gibson?

            DR. GIBSON:  As I have said before, I just wanted to make a plea that there be no arbitrary exclusions based on psychological criteria that centers, of course, have specific exclusions, much like transplant programs do with regard to active substance abuse and unstable mental illness.

            And even though I agree with the proposition that industry should not be charged with proving this treatment is psychologically safe, I do think that if there is a study, that it should be collected with regard to psychological effects that occur, particularly on the family.

            Whatever labeling is decided upon, again, I just want to remind people that it should take into account the cultural and ethnic diversity of the country that we live in.

            It's my unfortunate duty to say that when I read through this, I noticed that we were to talk about single needle devices, and I didn't hear that mentioned.  I suspect that could fall under physician practice and using a single needle device with this, but I'll just remind people that that was not mentioned in the discussion, even though it was mentioned in the material.

            CHAIRPERSON TALAMINI:  I would take the privilege of the chair just for a moment.  I run the home TPN service at Hopkins.  We had a young woman who lost all of her bowels dependent on home TPN and had been quite an outdoors woman.

            She called my office and asked if she could go camping on Assateague Island, wasn't sure how that was going to happen.  But, in fact, she managed to do that.

            She took her cooler with her TPN and had it in the park ranger's tent.  And every night, she would walk to the tent and get her cold bag of TPN and infuse in her tent with her husband so that she could have her camping experience.

            So I know that dialysis patients can't do that right now, but it certainly was exciting for me to study these materials and to see the possibilities of increased patient freedoms for this group of patients.

            So those are my only comments.  Dr. Aranoff?

            DR. ARANOFF:  I agree with my colleagues that dialysis is good and uremia is bad.  And although nocturnal hemodialysis shares many aspects with current in‑center and home dialysis therapies, it is fundamentally different in one way, and that is that the treatment is provided by the patient while they are unconscious.

            So I think it is incumbent on this process through the FDA to assure that the devices that are involved in this unique form of therapy have unique safety precautions to provide these treatments in a safe and effective way.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr. Afifi?

            DR. AFIFI:  I would like to make a comment about the use of retrospective data to demonstrate substantial equivalence.  In principle, there is nothing to prevent us from doing that.  And if, indeed, someone comes in with a 510(k) application and can demonstrate that there are existing data that could be used to show the substantial equivalence, then all they have to do is demonstrate that, indeed, they have the power in those data to be able to reject the null hypothesis of equivalence against a specified alternative that the new device is so much worse than the original device.  So in principle, I think we should not rule out historical data or retrospective data analysis.

            CHAIRPERSON TALAMINI:  Thank you.

            Dr.  Sadler?

            DR. SADLER:  I did not come here as an advocate for nocturnal dialysis nor for in‑center dialysis nor for daily dialysis or peritoneal dialysis.  I come as an advocate for patients.

            I think that all of these therapies are beneficial.  We know that if we don't do any of them, the patients die.  If we do one of them, the patients live.  Depending on what we do and how we do it ‑‑ and it varies from one facility and one physician to the next ‑‑ the quality of the outcomes is different.

            It is very important that we focus on the patients and we try to do the best job.  I have serious concerns about the nuisance value of dialyzing every day or five days a week.  And I think that the optimal dialysis is probably not the same for every patient and that we have not yet defined what really should be considered optimal dialysis.

            We have certainly defined adequate dialysis in terms of something that avoids many complications and keeps patients reasonably well but not really well.  And we have enthusiasts who say that more is better, and presumably there is no ceiling on that.  I don't believe that either.

            So I think we have much to learn.  And I think that support of these efforts to use every modality that we can get to is beneficial because it will give us more knowledge and we will learn where the golden mean among these therapies is.

            Having said that, I'd like to say that I would agree with Rod Kenley that our manufacturers have done a remarkable job.  We have very good machines.  They may not be quite as elegant as Charlie Willock's first product in terms of the ideas in it and the conservation of materials and energy, but they are remarkably reliable, safe, much quieter, much more accurate, and they perform a good service.

            I don't think that their performance is in question.  And because of that, I don't think that we have a lot to recommend to the FDA.  And I believe that if I were a manufacturer of one of these products, I would not come forward and ask for another certification.

            And I'd be careful how I use my advertising, but I would point out to people what my machine could do because I think all of them can do a lot.  And I think that we need to be appreciative of the information coming out of these trials so that we know what is best for patients.

            I hope that Bob and Chris are correct that 20 percent of patients can do this, but I'm skeptical.  I think the number may be closer to half that.  But I think that more people should do it.

            I do like home dialysis because of the convenience and flexibility it gives patients, whatever kind of home dialysis they do.  I don't believe that home dialysis just because it's at home is better than because it's not at home, but the convenience and flexibility are a great advantage.

            I think we have made some progress.  We have done a lot of conversation.  And we have reached a lot of agreement.  I think we have managed to demonstrate that whatever differences we had were far smaller than our agreements.  And I feel pretty well fulfilled by what we accomplished today.

            CHAIRPERSON TALAMINI:  Terrific.  Thank you, Dr. Sadler.

            Dr. Duffell?

            DR. DUFFELL:  Well, my compliments to the chair.  You ran a tight ship today, and we kept moving.

            CHAIRPERSON TALAMINI:  Thank you.

            DR. DUFFELL:  So thank you for that.

            My closing comments are really more directed at FDA.  Thank you for bringing this together.  I think what I would like to ask of FDA is just that you all quickly and efficiently pull this information together in industry.  The toughest thing is not knowing, rather than knowing, quite honestly.

            So even if we don't like what comes out of some of this today, I think at least having it in a written format to start dealing with or to prepare cogent arguments for another viewpoint is really what is important.

            So as quickly as you can summarize this and get out a guidance document, even in the draft form, I think is going to be really helpful to everyone.

            Thank you for the opportunity to participate.

            CHAIRPERSON TALAMINI:  Ms. Moore?

            MS. MOORE:  I would like to thank you professionals for tolerating and listening to a lay person.  This has been a learning experience for me.

            I believe that the nocturnal home dialysis system is a positive and that if it increases the quality of life for patients, if it increases their state of a well‑being, I think then it is worthwhile.  And I think that if we could get the technology to the point that there won't be too many questions about it and that the users will be able to handle it, then this will be a gift to society.

            So, therefore, I feel that with all that we have said here today and with FDA, in a few years perhaps there will a few more lives saved.  And that will be good.

            CHAIRPERSON TALAMINI:  Thank you, Ms. Moore.

            Ms. Brogdon, does FDA have any further comments?

            MS. BROGDON:  Yes.  I just wanted to say a couple of things.  I wanted to thank all of the panel members for your preparation time and for your experience and the energy that you have expended in this discussion today.

            This type of preliminary discussion for guidance development is typically very difficult for panels to handle.  And sometimes we come away with not much to work with.  I have to say in my many years at FDA, this is one of the best preliminary discussions that I have heard.

            So our plan is to put together a draft in the next few months and send it out to the public and to the panel members.  It would come up for further discussion at a panel meeting.

            Also, I just wanted to allay the fears as expressed earlier.  This is not just an intellectual exercise.  We are getting these sorts of requests from companies.  And we do need a guidance document, and this will help us develop that.

            I also wanted to say that we never forget that we have an obligation to regulate in a least burdensome manner.  And it's least burdensome to all parties.  And we will try to keep that in mind as we draft the guidance document.

            So thank you all for your time.

            CHAIRPERSON TALAMINI:  Thank you, Ms. Brogdon.

            I would, as the chair, like to also thank the panel.  It's really pretty incredible to me the degree of expertise, the set of pioneers that we have here, your ability to speak clearly and succinctly, your obvious preparation before the meeting to be able to come and crystalize these things quickly.  So I thank you very, very much.  It's what has made this really I think an excellent panel.

            So having said that, this concludes the report and recommendations of the Gastroenterology and Urology Devices Panel on nocturnal home hemodialysis.  And, again, on behalf of the FDA, I would like to thank the entire panel.  The meeting is adjourned.

            (Whereupon, at 4:18 p.m., the foregoing matter was adjourned.)