• Decrease font size
  • Return font size to normal
  • Increase font size
U.S. Department of Health and Human Services

Medical Devices

  • Print
  • Share
  • E-mail

Infusion Pump Workshop: Transcript for May 25th, 2010 (Part 2)

(Continued from Part 1)

(Janice Polcheck): Thank you very much. You guys all still awake? I’m the caboose bringing up the end before lunch. How many people are actually manufacturers, come from manufacturing?

Okay. How many of you are the reps that come and try to sell to the home infusion companies? Any reps? Just one? Well, we really enjoy when the reps come to our office and give us the insight on new products and tell us how wonderful the products are going to be.

And most of the time that is the case. I am the end user and I am the one whose license will be taken away if some of the products don’t work. So I’m giving you my perspective as the home health nurse that actually has to teach the patient in the home.

I have also worked as a nurse in the hospital. We had less limited information or less pump usage back in the 1970s when I worked and we also - I have worked for home doctor’s offices and home free standing suites. So we did use a lot of pumps there but I was the one that actually used the pumps and programmed them.

I wasn’t asking a patient to do that. So I was asked to just give you some pros and cons of what I think pumps are. And so this is purely my perspective and I’m sure you guys can come up with a few more things. I am a pro pump person. I do like using pumps. I think they’re the technology is wonderful.

I believe that you can make sure that the patient’s therapy is accurate and that they can get the medication prescribed in a timely manner and hopefully that that all works out for you. I love the ambulatory capacity of the pumps. Patients truly are using the pumps sometimes 24 hours a day for pain management.

They can go to the doctor’s office, their therapy does not have to be interrupted. They use the pumps to give TPN, some of them continuously but usually through the night. You have patients that have internal feeding; if it’s a child it’s usually continuous. These people go on vacation, they go on planes, they go to the beach. We’re shipping drugs and their supplies sometimes to Myrtle Beach so they are having a very active life.

And the fact that they’re sick and they need IV therapy does not stop them. Sometimes these therapies go on for their whole life. We have patients that have short gut syndrome that have had TPN for years. And usually they have the same type of pump if they’re with the same provider.

They’re very, very good with the pumps. We very rarely get a call from them unless there is some sort of a fatal error but they know if there is error line, if there is an up occlusion or down occlusion, they know how to handle it. They know how to change the cassettes. All we see them for is just a weekly lab draw or dressing change but they are very good with the pumps.

We also like to use the less technology possible for patients. So an innermate pump for almost all of our medications really works very well. It’s low tech, basically plus or minus a couple minutes gets the patient’s infusion in in the time that we ask it to be done.

And it’s just wonderful. The patients like it and we get a lot of referrals because we use low technology for the patients. A patient who has no caregiver at home or has to give a medication every four hours through the night, it’s very easy to use one of the pumps that you can set it on an intermittent setting and put the whole amount of medication in, one cassette for the day.

And maybe you only need to have the nurse or the caregiver’s wife, husband or father or sometimes it’s a daughter for an older patient, to just change the cassette once a day and it makes it very ease of use for that patient. They don’t have to worry about it unless of course the pump goes off and that person hasn’t been completely nursed on that.

Cost effectiveness of the pump - the hospitals don’t want to keep a patient that needs to have an antibiotic in the hospital when they can get a more highly acute patient in there and probably make a little bit more money or at least get them out and have a bed available for another patient.

The insurance company of course wants to get the patients out of the hospital and that’s usually the driving factor to get somebody out of the hospital is the insurance company. And then of course the patient wants to get out of the hospital so that they can get back to work.

I would say that probably in our population probably 75% of the people still work. And they don’t want the fact that they need to have IV therapy in any way have to keep them from going to work because their working pays their insurance and their insurance is paying for their therapy.

And also if they get out of the hospital they are less likely to get a hospital-acquired infection and that is certainly a plus for them. The smart technology that I hear everybody talking about, we have not used that in the home. But I see that as a very big pro to where you can hopefully people won’t override the parameters and that the less drug errors will keep patients safe.

And I think another really good thing is and I’m not sure that anybody touched on that is a lot of these pumps have a downloading data that if you have like a PCA pump and you think that the patient has been overdosed, you can actually download the data and find out did the patient somehow override the lockout and was able to give themselves more drugs?

So it’s very good in the fact that you can hit that sort of - gather that sort of information and it’s not necessarily because the pump was programmed wrong when it left our facility but maybe somebody messed around with it.

The cons on the other hand are a little more but that’s just the way life is I think. Sometimes there are more cons than there are pros.

I feel that the more options on a pump, the more likely you are to have an error. So that’s why I like to use the correct device for the therapy as opposed to buying all these pumps that have a whole lot of options. Sometimes you just need the old work horse pump and I think you’re less likely to have an error in some cases.

Sometimes the features that are put on them are useless. So if they’re not used they’re just useless to us so again if you have less options and have a good sturdy pump that’s better for us. As Mary said the clinician and sometimes the non-clinicians, these pumps sometimes are just not user friendly for how many steps you have to get to actually get the pump turned on to get the therapy.

We get an awful lot of malfunctions that for example it will say error in line and there’s not one bit of error or there will be door latched – a door unlatched or door open and it doesn’t appear that there’s any problems but the alarm just will not go off.

So you have to take those pumps out of service and get a new pump to the patient. Same with equipment breakage, we have a lot of problems with the doors, you know it’s not people being gentle with them, these are people that these pumps aren’t owned by them.

They figure if the pump breaks, no big deal if they drop them, we’ll just get them a new pump so we do have a lot of problems with you know the doors not latching or the – or buttons not pressing properly and it’s just because people aren’t really gentle with them.

And I don’t think that they should have to be – you shouldn’t have a pump that you have to be gentle with. In our facility we have about 250 patients that are active right now and we have about 175 pumps that we keep.

And only about 50% of the patients get a pump. We have to rent those pumps from a rental company, not necessarily from the manufacturer and that costs anywhere from $65 to $100 per pump.

Luckily for us because we rent them we can send them back and it doesn’t cost us anything like the other places have to send them back to the companies. It doesn’t cost us anything but if we were to buy the pumps then of course we’d have to have them PM’d and it would cost us money for that.

We do have to keep a large inventory of pumps because we have patients that are on very critical therapies like inotropic therapies and we have to give them a second pump as a backup.

So you know so that’s a pump that we can’t charge for, that we can’t use for another patient because we have two pumps in a patient’s home.

If the patient lives any distance away we always send two pumps with them, that way they have a backup pump. So that extra inventory that we have to carry every month to accommodate those patients.

You have to – another con would be that you have to buy the tubing and accessories, it’s not universal. You know that’s been forever that you have to buy the accessories.

The command language, not every pump start and stop isn’t always the same, there’s a run pause. Some of the functions are the same but the language is different.

I think that that would definitely help if some of the language was universalized. We also, going back to the pumps we have to have one dedicated person to actually take care of the pumps to make sure that we don’t lose them, which that happens a lot.

And then we end up having to buy them from the rental company, but it would be nice you know if there was a – maybe some sort of a method that the manufacturers could help for inventory maintenance.

And lastly we have to make sure that all of our nurses are competent so every year there’s an annual competency as part of INS standards you have to make sure that everybody that uses the pumps that you dispense are competent.

And that we do on an annual basis and some people end up doing that a little better than others. Anyhow if you have any questions, put them down on your card and I’ll be glad to answer them afterwards. Thank you.

Melissa Eackle: Thank you Mrs. – Miss (Polcheck). All right, I know you’re all ready, it’s lunch. Please be in your seat at 1:45, gather your question cards on the way out. Please fill them out and hand them back in to the registration desk or you can stand at the microphone.

For those on the webcast listening please do not hang up and the session had no slides that you just heard.

Thank you very much, see you at 1:45. If you need a dining guide they should also be on the registration desk.


Melissa Eakle: All right. Mary Brooks - Lieutenant Commander Mary Brooks has a clarification to make. So if you could all please be seated so we can get started. Thank you.

Mary Brooks: Well I know that a good majority of everyone in the room today was also here yesterday and this information will also be available on the home care page. But there were a few questions yesterday that was related to when - how do I get new labeling to the FDA?

And, you know, I made the little comment that I didn’t have enough stripes on my sleeves to make that decision without talking with Office of Device Evaluations. And they did agree that if you have already went through your 510(k) decision-making tree and you do not need to submit a new 510(k) to the agency for your changes in labeling you can send an add to file.

If they’re minor changes you don’t need to send anything. But if you made some major changes to your labeling you can send it to the agency as an add to file.

When we receive the add to file and also the Office of Compliance that when you - but when you make your changes and you send an add to file to the 510(k) we’ll actually look at that and confirm that you did not need to submit a new 510(k) for these changes.

Sometimes we do receive labeling changes. And in the labeling changes we know the intended use of the device has changed so we will actually look at your add to file labeling changes. And if it has tripped the limitations of when do you file a new 510(k) we’ll actually send you a letter and let you know that you need to submit a new 510(k) to support your labeling.

Okay? So does that - I told everyone I’d get back with you. I’ve had a couple of people ask me for that answer so I’m giving it to you. And also yesterday we had a question about manufacturers and their decision-making tree for MDR reporting. That question came up again earlier.

I’ve spoken with Office of Surveillance and Biometrics and the RS&B staff and those are questions that you can pose directly to the RS&B staff and that is (Sharon Cap)'s group with RS&B. So if you have questions about your manufacturer decision-making tree and when to, you know, when do you need to file a 510 - excuse a MDR report they’ll help you with that process.

(Debbie Oder): You have to have (unintelligible).

Mary Brooks: All right. And that’s (Debbie Oder). Thank you very much. All right?

(Debbie Oder): (Unintelligible).

Mary Brooks: 1-800-Deb. So we just wanted to make sure that, you know, we’re trying to be as helpful as we can to the manufacturers and the clinicians in the room.

Also another clarification that I was asked to is to hospitals. You do not have a mandatory filing requirement for an MDR for malfunctions. It’s not required. You’re required if there’s a death. You have to file a 3500(a) and you have to file an MDR report for serious injuries if you did not notify the manufacturer or the manufacturer is unknown.

However you can report malfunctions to the FDA as an - under the 3500(a) as a nice to know but still give us the information but you are not required to. So I just want to clarify a few points and give you some follow up information that was asked of yesterday. Thank you.

Melissa Eakle: All righty. I’m back. I just want to go over the afternoon so we’re all clear on what’s going to happen. Again we’re collecting questions at the registration table. And you will have an opportunity if you want to stand up at the microphone -- speak into the microphone clearly, state your name and your organization or your affiliation.

We have questions and answers after the presentations - the two presentations that remain in the afternoon. That is our international guest who will be introduced and also our AvMed and AAMI representation here for their perspective.

We have a question and answer period from 4:20 to 4:50. However I will reassure you that if the questions go over we have time tomorrow to take them back up again. We’re going to try to get you our of here by 5:00 because of the traffic and other issues but we will be able to answer questions. We have plenty of time tomorrow for questions. All right. Any questions before I go on? All righty.

I would like to introduce someone who has been very helpful at the FDA to us in this process. And this is Carole C. Cary she’s the Director of International Staff in the Division of Small Manufacturers International and Consumer Assistance belovedly known as DISMICA under the Office of Communication and Radiation Programs or another acronym, OCRP within the U.S. FDA Center for Devices and Radiological Health.

Her major responsibility is international advisor on medical device regulatory matters providing technical support and regulatory training. She coordinates the centers harmonization by doing an collaborative consultation and review pilot programs and participates in standards development.

Prior to her current position she was for many years the scientific reviewer in the cardiovascular respiratory and anesthesia device section of the Office of Device Evaluation. She was a Mansfield Fellow in Japan as an expert regulatory scientist. She earned her Bachelor Degree in Electrical Engineering at Johns Hopkins University and a Masters Degree in Engineering at Loyola College in Baltimore.

She is a senior member of the Institute of Electrical and Electronics Engineers and the Engineering and Medicine and Biological Society. Please welcome Carole Carey.

Carole C. Carey: Thank you Melissa for such kind introduction. It’s a bit much but thank you.

Good afternoon. We hope you had a good lunch and had a good break. And probably it’s time for a little nap. Right? But moreover I wanted to say that I hope that you enjoyed the presentations this morning and you found them very informative as well as provocative.

So we started off the workshop with FDA speakers first defining the problem and then you heard from several of the users groups who presented their experience and perspective with great enthusiasm. I just wish I could see the slides a little bit better. But there were - they’re going to be posted so that’s good. Thank you.

So this afternoon we are going to hear the international perspective from our regulatory colleagues outside the United States. This will be followed by presentations from the industry perspective.

So again my name’s Carole Carey and the (unintelligible), the Division of Small Manufacturers International and Consumer Assistance primarily is your resource or “go to” place for first line inquiries in terms of general regulatory support and questions.

So we have about maybe six or seven desk officers assigned different regions in the world to handle the questions. And if they’re more in depth and require more technical response then we refer you to the appropriate office or division or branch.

And so I know people don’t normally know where to go. It’s like, “Oh, how do I go to the FDA and ask questions?” We are your first line of inquiry. So the general address is dsmica@fda.hhs.gov or you can also just, you know, contact me if you don’t remember that address.

Okay. So the other thing I just wanted to mention is we are engaged with our international stakeholders and regulatory counterparts in the world. But however when it comes to involving sharing nonpublic information with foreign governments and international organizations we work very closely with our Office of International Programs which is in the commissioner’s office.

So with infusion pumps group here for example we work with David Kelly and I don’t know if (David’s) here. I thought he said he was coming. Hi David. And David is the - an Associate Director of the FDA European office stationed here in the U.S.

So the director and I believe deputy director are stationed in Europe. The main office is in Brussels. So as many of you know we have established several FDA offices throughout the world in China, India, Middle East, Latin America. And I don’t know if we’re going to have more maybe in the future.

So in my present capacity as the Director of International Staff and former Senior Scientific Reviewer in the Office of Device Evaluation it is my honor and privilege to present to you this afternoon our two international speakers from Canada and from Sweden.

They have traveled long ways and taken time away from their busy schedules to participate in this workshop and we appreciate that very much. I think all of you share the same appreciation as well. Thank you.

So unable to participate is a third person which is our colleague from UK’s MHRA or Medicines and Healthcare Regulatory Agency. So after the two presentations we will run the slides to you for viewing but it will not have any audio because we didn’t get any text with it but again the presentation will be posted on Tuesday.

So without further ado our first speaker this afternoon is (BC Lowey). (BC) is Senior Scientific Evaluator in the Device Evaluation Division of the Medical Devices Bureau of Health Canada.

She works in the General and Regulatory Devices section where she is part of a team that evaluates safety and effectiveness of an array of devices such as general surgical devices, extra (corporeal) treatment devices, software, (physiological) monitors and imaging.

She has worked in the bureau for three years now and holds a Bachelor’s Degree in Biochemistry and a second Bachelor’s Degree in Chemical (Engineering) and a Masters Degree in Biomedical Engineering in the area of Biotechnology.

Based on her knowledge and experience (BC) will speak to us today about device classification in Canada, problem reporting - the importance of problem reporting, pre-market and post-market requirements for infusion pumps as well as challenges and next steps. Ladies and gentlemen please welcome (BC Lowey).

(BC Lowey): Good afternoon everyone. First I’d like to thank the FDA for inviting Health Canada to be present today and participate in this workshop and to speak on our regulatory experience with infusion pumps.

In particular I’m going to share on our pre-market requirements and highlight some of the key aspects that we look at in order to do our work in the Device Evaluation Division. Next slide please. So just to first introduce the Bureau, the Medical Devices Bureau is under the Health Products and Food branch in the Therapeutic Products Directorate. We consist of four divisions and we work together to ensure the safety, the effectiveness, the quality of devices prior to receiving authorization to be on the market.

The device evaluation in particular our key activities involve pre-market evaluations, evaluating requests for investigational testing, i.e., clinical testing as well as well special access requests, that’s for requesting unlicensed devices for emergency use.

As well we provide technical support to the bureaus who are responsible for post-market surveillance and post-market evaluations, risk communications as well.

Two other Directorates I’m highlighting here which are important are the Inspectorate, they’re responsible for the compliance and enforcement actions - so the recalls, investigations and responses to mandatory - to problem reports, excuse me.

As well as the marketed Health Products Directorate. They’re the Directorate who’s involved in the post-market surveillance and the risk communication issues. Next slide please.

So the foundations of our regulatory approach are the tools we use to ensure our work include the food and drug (vacs) as well as the medical devices regulations. We use these to enforce the sale, the importation and the advertising of medical devices on the market in Canada.

We do not regulate the use of devices. That’s to say it’s just the sale, importation and advertising of devices. Further, as well in order to have these opportunities manufacturers must obtain a medical device license from us which we issue.

The regulations are also based fundamentally on a risk-based approach. So that’s to say that it balances pre-market scrutiny with post-market surveillance in order to ensure the safety of devices.

And the degree of this pre-market scrutiny and intense - in turn the pre-market requirements that we require for a specific device is based upon the potential risk that the device may pose and then hence it’s risk classification.

Devices are classified from one to four - one being the lowest, four being the highest. And in particular external infusion pumps are categorized as class three devices.

One other aspect I’d like to note. In our pre-market regulatory process we do not have the equivalent of a 510(k). So in that to say for all medical device license applications whether they be for a new technology or an existing one or a new device or even amendments to a new device we require that manufacturers meet the full set of pre-market requirements in terms of the information they provide and the sufficient data to support the results - must meet the full set. Next slide please.

So further in terms of our pre-market requirements for infusion pumps key part is background information. So background information includes device description as well as the design philosophy. And in this information we’re looking at for instance the physical properties, the modes of operation, the features of the device. And one key area that we highlight are the extent of the different safety features that are incorporated to the device. So we’re looking to make sure that there are a comprehensive set of measures in place, methods in place to detect, to warn users and to prevent the common and recognized hazards associated with pumps.

So I’ve given examples here of alarms. We’re looking at the alarms for instance for the battery, the alarms for - and warning signs on the displays for occlusion for instance. We’re looking at as well the function of the sensors, whether - what are the extent and the comprehensiveness of the sensors in terms of error in line in terms of occlusion for instance.

We’re also looking at what other unique features the devices incorporate. Examples include free flow prevention or dosing limits that the device incorporates - that the pump in particular incorporates.

As well we look at the accessories and compatible devices that are recommended for use whether, one, they have been licensed in Canada, that is a requirement if they’re going to be referenced for the device as well that they have been validated for the particular pump that is the subject of the application.

In many cases manufactures often make reference to an older model of predicate device whether it be their own or one from another manufacturer. And this predicate they use often to support the design philosophy of their pump.

And in such cases as well as in amended applications when they’re looking to add some new features to a new existing model that’s already licensed manufacturers will leverage off this predicate. And in such cases we’re always looking for a side by side comparison of the devices looking at the properties, the features. And we look to make sure that the areas that differ are highlighted and if the similarities are highlighted.

And again this does not necessarily exempt manufacturers from providing the full set of requirements and information to meet the requirements, it’s more so to support the safety and design of the - their new models.

In terms of marketing history this is another area where we leverage a lot of this information in terms of determining the safety and how to go about and whether to go about licensing pumps. We’re looking at information on whether the pump has been licensed for investigational testing, clinical testing or special access. And we’re looking for the outcome of those testings.

We’re looking for information on whether the pumps have been licensed in other counties or regions as well as the number of units that have been sold. And we’re looking for a breakdown of the types and the number of reported problems and incidents as well as any recalls that have been issued in those markets.

In terms of the case where manufacturers make reference to a predicate, an older version of their device where maybe they just added some new features we specifically ask for the marketing history of that predicate.

And we can obtain quite a bit of information - we don’t leverage it too much. Mind you there’s a great advantage in the information that comes from the marketing history. We’re looking to see whether on the other markets and jurisdictions where the pump has already been licensed if there’s been any early signs that we can identify problem reports, human errors or human factors that have been incorporated or other issues. So we’ll look to ensure that if there are early detection signs that the manufacturer is investigating it and looking to resolve the issues.

And as well particularly when their new models leverage are based on an older model that’s already licensed we’re looking for information to see if the marketing history identifies signs of component ware for instance or other issues. And before licensing we would look to make sure that the manufacturers have resolved this issue prior and to prevent it from being incorporated into these new models that they’re coming out with.

And further a clean marketing history will provide us some indication, some assurance that there is a reduced likelihood that the issues - reduced likelihood of risk with the pump. Next slide please.

Among the pre-market requirements as well we’re looking at safety and effectiveness studies. One key area we’re looking at the list of standards that the manufacturers comply with in their device design, their manufacturing, their safety, their effectiveness testing.

And Health Canada in particular has a decoration of conformity policy. It is optional and we invite manufacturers and recommend manufacturers to fill out our Health Canada specific forms on our Web site. And they’re allowed to attest to that they comply with specific test standards that are recognized in our list. And this is in view of providing all of the reports to demonstrate that they comply with the methods.

Bench testing, that’s one of the critical verification methods that we look at. Areas include physical mechanical testing of hardware components, we’re looking at for instance the sensors or the pumping mechanism. We’re looking at electrical testing as well, examples include the battery, the internal communication of the subcomponents. We’re looking at accuracy testing of the dosing particularly for the different fluid types that are recommended and prescribed for the pump.

Environmental conditioning. Looking at the pump performance in response to different - within the operating range of temperature, pressure and humidity particularly important for portable pumps. As well as testing - strength integrity testing referring to how the pump responds to bumps, to impacts, to vibrations. And again that’s very important for ambulatory pumps.

Software verification and validation testing another very critical aspect and key part that needs to be verified and validated. We’re looking not only at the acceptance criteria and the results of testing but the scope of testing that’s been performed in terms of how does it compare to the intended software functions of the device? And making sure that it’s comprehensive and that it has - and that it’s in line and that there aren’t areas that have weaker test cases in scope the testing.

We’re looking at the outstanding anomalies and software bugs and assessing them not just based on the manufacturers deemed risk level but we’re assessing the accessibility of these bugs in terms of how it affects the devices safety and effectiveness.

Usability testing and risk assessment. These two requirement areas are not standardly required for class three pumps - or class three devices, pardon me. However for pumps we impose this due to the history of the safety concerns that surround it.

Usability testing as we’ve heard mentioned today and as known is a very valuable validation area. It looks to support the intended use of the device in terms of how the manu- how the user, pardon me actually use it. Whether the full set of modes and features are up to par. Trying to identify for instance the labeling - whether it’s sufficient, training program - whether it’s sufficient. And identifying whether there are any human factors such as the user interface of heretics that may have impact on the use of device which may not be detected in the other bench testing and verification testing methods.

And in terms of deficiencies in past we’ve found deficiencies in labeling and required - requested for medications be made. We found deficient either areas in the training that may require modifications and made recommendations as well as it does highlight areas that are more challenging to users as well as areas that may be more user error prone.

And from the outcome of the usability testing we use that information and we ensure that it’s been incorporated in the risk assessment. We use it in our critiquing or valuation of the risk assessment ensuring that mitigation methods have been put in place to address any issues that have arose in the usability testing.

As for labeling we do evaluate the labeling. We look at package inserts, we look at the manuals, we look at the brochures. And in particular we’re looking for the indications - the contraindications, the warnings and limitations and the instructions on how to operate the device.

We’re looking to ensure that they are accurately and adequately described for whether it be the lay user or the healthcare professionals who will be reading it. Next slide please.

So as I mentioned the regulatory approach we take is - the basis of the regulatory approach is based upon a balance between pre-market scrutiny and post-market surveillance. And a very important part of post-market surveillance as been mentioned today is problem reporting. It’s mandatory for manufacturers and importers however it is voluntary for the users and healthcare facilities. But it is recommended.

Reporting is made to the Inspectorate Directorate of Health Canada and for manufacturers is required within ten days if the incident led to death or serious deterioration of health. Within 30 days if it were to reoccur could lead to the death or serious deterioration of health.

This is a side we don’t have representation here from Health Canada who are responsible for overseeing the problem reports. But in general the types of problems that we’re seeing with infusion pumps parallels what was described this morning and what the FDA is seeing. It’s similar software error issues, it’s human factors and use error, it’s battery errors issues.

And in terms of percentages we’re seeing fairly similar where 1% has led to death. Approximately the highest category would be those where if it were to reoccur could lead to death or serious deterioration . And then there’s also a large category that are unspecified or undefined. Next slide please.

So problem reporting - the importance of it really cannot be over stressed. A lot of - and it was mentioned this morning many of the problems and issues that arise with devices but particularly with pumps are not anticipated. They’re unanticipated and they cannot be detected always and as easily in the pre-market stages.

Whether they be serious issues or just minor flaws and minor problems they have the potential to lead to safety concerns and serious safety problems. And so often in such cases they are best detected in the actual clinical use scenarios.

So when the pump is being exposed to the typical environmental conditions of day to day as described this morning with water dripping or liquids dripping on it, when they’re exposed to the normal day to day use, long hour use, intermittent use and over duration of the lifetime of the pumps.

So healthcare practitioners, the users they have a unique experience and they have a wealth of knowledge by being able to see exactly and identify problems and flaws or design issues that just don’t sit right with them and think may lead to some issues.

And so we really encourage and stress that they should provide this information to the manufacturer as well as the Inspectorate in order to support investigations and do prompt investigations.

Problem reports as I mentioned they often are what lead to root cause analysis when provided to manufacturers and instigate the correct preventative actions. So that’s the importance that they need to be reported.

As well problem reports being sent to the actual Inspectorate is what - it guides us in our approach to how to deal with the risk. It informs us to what the risks are that maybe we’re not - they’re not aware of first hand.

And as well it supports their investigative action, it improve the efficiency of their investigative action and it’s also a mechanism and method by which they pursue the manufacturer to ensure that the investigative actions are being taken seriously, not downplayed and actually actions are being implemented.

Challenges that they face in this area include problem reporting, reliability and accuracy of the information that’s being provided. So in terms of the details how well and sufficiently it’s being described, what happened, why, what went wrong, what were the events that led up to the issues -- very important. And sometimes there’s a bit of a deficiency and it’s difficult for them to efficiently and accurately drive their investigative actions.

As well a concern they have is under reporting. Whether the reports and the issues are being delivered to them as often as they actually occur whether it be because manufac- or sorry the user or the healthcare facilities have reported it once and feel that’s enough. Or whether because the standard for pumps in particular is being lowered and maybe the issues are being accepted as the status quo.

Another being the implementation of corrective actions. Challenges in terms of ensuring that the investigative actions by manufacturers are effective, that they’re done in a timely manner and that they’re not being downplayed.

That the corrective actions that are actually being developed are effective and not going to elicit other - and result in other problems in other areas that were not a problem before. As well as timeliness and efficiency of deployment of the corrective actions to the users. Next slide please.

So next step. Our key focus right now is on collaboration. We are working with our international counterparts. We are going to be together looking at and listening and waiting to hear from the partners and the stakeholders and R&D’s together to investigate and determine what are the issues and the problems associated with these pumps? What are the contributing factors? What are the best resolution and methods that we can take.

And then from there looking at an approach to move forward to improve the reliability and the safety of the pumps that are on the market and protecting users.

As well in the meantime we continue on our pre-market evaluation and taking the information that we’re learning a long the way and carrying it back into our pre-market scrutiny.

We’re looking at continued post-market surveillance and compliance and enforcement by our colleagues in those bureaus who are responsible for that. And keeping vigilant for incidents that - types of incidents that are occurring in a higher number and particular types of pumps maybe that are producing higher numbers of issues. Next slide.

And just to finish off I left you with some of the contacts for the different bureaus. We don’t have representation from all of those here today so I may be limited in the information I can provide to you on those sides. But I left you with the contacts should you have further questions. Thank you.

Carole C. Carey: Thank you so much (BC) for that very informative presentation. Our next speaker is (Matt Olson). We are delighted to have him with us with him today, another speaker outside of the United States to give us a presentation.

(Matt) is a Senior Administrative Officer and this year he was promoted to Senior Expert with Medical Products Agency or MPA of Sweden. Congratulations.

His responsibilities include inspections, vigilance, market surveillance and auditing of notified bodies. He is a certified lead auditor and holds a Master of Science Degree in Biomedical Engineering. He is very active in medical device standards development such as the IEC infusion pump standards, the IEC TC 62 and the ISO quality management.

Currently he is also the Chair of (Senelec) Task Force to prepare standards proposals for medical device manufacturers post-market surveillance. We are delighted to have (Matt Olson) with us today to talk about MPA’s concerns on the current market situation from a Swedish EU perspective. Please help me welcome (Matt Olson).

(Matt Olson): Thank you. I’m very honored to be invited to this occasion. It’s very great value for us to participate. We are a small country of course and the European market it’s a big market, it’s as big as the U.S. market. But anyhow it’s a great things to see most people listening and all the effort you have made to do this. We really appreciate that.

So I’m going to bring you some thoughts about our concerns and much of that has been said before today. But I just want to say it from our perspective and I could say also that it’s from a European perspective because we have similar - or we have the same legislation around Europe. So next slide please.

Yes. The reason for our commitment in this project is of course mainly patient safety concerns. The concerns are raised by a lot of adverse event reports.

We also can see the reason behind why it’s a little complicated with the infusion pumps and we really want to see it from another perspective, namely the infusion therapy perspective where the medicines have developed through the years and have created a situation that wasn’t there say 20 years ago when we started.

Our concerns about infusion pumps and standardizations on the situation was another. You didn’t use the pumps, you didn’t use the infusion therapy in the same way as we do today.

Sorry to say that we also have seen a lot of very poor actions from manufacturers and representatives as well as in Swedish as in the Euro markets. We are not very impressed.

And to put more force behind those efforts and to get the solution we see a need for international harmonization both rules and also surveillance. And cooperation between authorities and other stakeholders to get the situation better. Next picture please.

Typical adverse events we have heard from them early on. I didn’t copy from FDA really although it looks like that. But this is what we also have seen in - for ourselves. The alarms - the alarm problems, the software problems, agronomic factors, usability factors it’s a big problem.

One thing, the fourth dot here is something that I think perhaps we haven’t talked about so much and that’s about disposable and we have a lot of problems with the disposable. They are not treated in the same way as let’s say the main products, the pumps and they are coming in behind.

And the pumps they are - well at the hospitals, the biomeds they take care of the pumps and everything with a power wire on it. But the disposable they are - no one really is caring for.

Then we have of course the problem with failing power supplies, batteries, we heard about - and the poor mechanical integrity. Things break, we heard very interesting descriptions before lunch about different failures that could arise. I don’t know if you’ve got that but I have a former experience, a type tester, a test house so I’ve been testing those things several years ago and also was involved in much manufacturer development procedure. So we have seen those things in the very early stage and tried to help out. Now it’s not that - I’m not working with that today but I can see that we still have the same problems. Yes. Next picture please.

This is copied from the FDA. Now I have (unintelligible) I think it will be approved.

Woman: Right here.

(Matt Olson): Yes. It’s from the draft guidance total products lifecycle infusion pump premarket notification. I thought it was very interesting to read because I recognized most of it and really appreciated the paper. Some things to do but mainly I think a very good thing.

And I took the risks from that paper because I - again I see that they match. They match well what we can see. These are the outcome of the flaws of the device when it doesn’t work as it should. We have those problems.

Now the under dose problem - I don’t know how many of you who were working with the infusion pump standardization for let’s say 20 years ago - was anyone here? Yes?

The philosophy at that time was that under infusion wasn’t an issue. It wasn’t such a big issue at least because the medication and the methods well they weren’t done that way that under infusion could be so dangerous.

And the manufacturer said we couldn’t protect you from under infusion anyway because anything could break. So that issue was taken away, it was not a very important question.

So it took some year before the standard was adopted by IEC. It was in Toronto 1998 (unintelligible) meeting. And I think you weren’t there at that time. No?

A very good meeting. But at that time the standard had already been a little outdated really and we had some thoughts about that. We tried to - yet the main tenants were started directly and (unintelligible) people say from also from the committee that was the IEC TC 62.

That (unintelligible) said that the maintenance work would start immediately. Unfortunately it didn’t. It didn’t start until well quite recently - a few years ago. And now it’s much to do, very much to do with that standard. Yet the more modern type of infusion therapy (unintelligible) into it. Okay. Next slide please.

So this is a little - well statistics you could say. This is reported events it was in Sweden to the Medical Products Agency. We have in the lower you have the report, the manufacturer has to report the same criteria as you have, the same as they have in Canada I believe.

But we also have requirements for the user to report. So if you see at the bottom line - if you just think of that the Swedish population is 9 million and you can multiply it to your numbers and it ends up even so that we have fewer reports. Compared with for example UK it’s about the same compared with Germany we have more -- more reports. And I have no reason to believe that’s any difference in real life, the background is the same.

But you can see that we have about 70, 80 reports per year and now in 2009, ’10 those nearly 60 were user reports. So of course it’s a little better than in U.S. And of course we also realize there are lots of incidents that are not reported.

We had a meeting the other day, we had the Swedish representatives from Swedish Healthcare dealing with infusion therapy mostly from the biomeds department but also nurses and people from healthcare.

And they were very afraid because if they were to report everything that happened they would let us drown in reports. But we said, “Try us.” We had heard that threat before and yet no one has succeeded yet to drown us with reports. So let’s see what happens. I think next year I hope there will be other numbers.

But where are the missing numbers? Are they on the user side? Yes, we know they are on the user side. But how much are we meeting the manufacturer side? Lot’s I think, very much.

And of those reports we have received many of them have been - let’s say we have urged the manufacturer to do reports because we haven’t had information from the healthcare so they’re not always spontaneous reports.

I don’t know about you but you have a lot of reports from manufacturers I don’t know if they are spontaneous or voluntarily done. And you can see the examples as cited they are the same examples that you have, the same things are happening. Okay. Next slide.

So - but user - not user error, use error. There are two situations I want to address. The first situation is where the pump was used or the device as such, system was used according to local clinical guideline but outside manufacturers intended use. So the guidelines, the method, the rules for the healthcare didn’t fit so well with what the intended use was - so from the manufacturer. That is common really.

But examples you use the pump for wrong kind of drugs, the pump is powered by batteries when not intended for transportation - for transport use, for human transport. It’s very common after such an incident when the report comes from the user or the manufacturer directly. It wasn’t intended to be operated on battery for so (unintelligible) only for inside hospital short break - power breaks or such things.

The batteries weren’t charged according to the directions for use, there were other routines at the hospital. The management of disposables, how - what kind of disposals were suitable for this device. The cleaning routines were not suitable.

We heard early on that the infusion pumps they are supposed to be wettened, drowned. No not drowned but at least a nice rain. And the rain not just from the (unintelligible) but the rain from the side. And it’s important to remember that this is the way they’re going to be treated. And the manufacturer who doesn’t realize that well he has problems.

The pumps are also wrongly used as kind of well monitoring device and not supervised often enough. So you start the infusion then you go away, you came back within one hour and something has happened. We also have heard about those kind of incidents. And of course you can find somewhere in the instructions for use that you cannot do it. It says you have to supervise it all the time, you cannot trust it.

I learned a new word this morning it was wrong clinical use model. And I like that because that describes just what it is about. To use into a method that wasn’t appreciated by the manufacturer.

So - but the second example the pump is used according to the manufacturer intended use but outside clinical guidelines. It’s easy when you’re dealing with software if you put in different features. It’s very easy.

That’s a standard or technical report from IEC. It’s a guidance to the IEC 62304 standard for software lifecycle management something and much about risk management. But the technical report it has another number it’s 80,002 and it was published what was it, some time ago. Half a year? Someone knows. Have you seen it? It’s very good.

If you are a manufacturer you should show your programmer’s software engineers. They have to read it because it’s really good reading. It’s very educational and it’s easy to read too, it’s not too complicated. So that’s also a reason for the incidents. Many of those incidents that we have have reports about - they end up user - end up in those categories.

In talking about this we can also talk about the use for children that the ordinary pumps are used for - in children. Children they are small and moving patients and they should be moving of course. And they are crawling around with a pump.

Does the pump get batteries charging as intended from the manufacturer? No it doesn’t. Deep down it’s charged and sometimes the batteries don’t - can’t take that, they are destroyed. Discussion about air bubbles, which size is suitable for grown ups and which is enough for small children. It’s not the same. Flow rates. If you use a grown up pump, a big pump and you have to set it for children of course you have to use often very low settings and then you end up at the end of the range where the settings could be made. And everyone knows that the end of settings, you should use them because the pump is more inaccurate that way.

And of course if you use a very low settings and something goes wrong what is the worst case? Well that is that the pump goes at the maximum rate and the difference between the very low rates and the maximum rate is very big so you get a very large (activation). It’s a great risk. Next slide please.

Well as I said before the method for infusion therapy methods have developed. And I want to stress that although we are talking about infusion pumps here I think we should write the view to see an infusion therapy. Because it’s a system with different devices that should match together.

And it’s not only the pumps it’s different devices around. There are of course the infusion sets, we have the syringes, you have the bags, you have extension sets, you have those valves and you have also computer interfaces and so on. We have lots of devices together and let’s don’t forget that it’s a system.

So anyhow we think that the infusion therapy has been increasingly sophisticated and demanding more confidence in the product since the drugs are changing and expectations are changing.

So the former criteria that over infusion is dangerous but under infusion is not, is not true anymore. It should be understood. And the infusion therapy, the methods are dependent on the existence of devices. You can’t just switch it to not use any device. (Unintelligible) is not sufficient often. We have more potent drugs now than before and also leads to more serious consequence. And if a infusion pump fails it ends up in a problem because we have to have another pump. So the availability of the infusion pump has become more critical. It’s not like going back to any old methods anymore. We have to have them. Next slide please.

So back to what has happened on the market. We can talk about manufacturers but I think we also have to think about the distributors, the local representatives or vendors and so on. And that’s a problem with the communication, who should do what?

After an incident who should take care of the investigation, who should collect the things, who should send in the report. Many times this (unintelligible) the time share. It’s a very common problem.

And so we also have the subcontractors to the infusion pump manufacturer. For example the subcontractor who are making the - manufacture the infusion set and they do a change or something.

Clearly what we could call (a field) safety corrective action. You know, the expression? Is it used here to? You do a corrective action it’s about safety and you’re doing - the market, they are used on the market, the devices the (field) part of it.

Very often some changes are made by a subcontractor but no reaction from the manufacturer, nothing done on the market. And they say, “That’s their problem that’s not our responsibility.” Of course we tell them that it’s wrong and we see it but it’s not always that we are informed about it.

And then we have a serious problem. We have seen very poor investigations after serious accidents. I don’t know if you have - you should have, I believe you have in U.S. requirements for the manufacturers to have a decent routine to investigate accidents. I mean not only that they should do it but also that you have to go through are they effective? Are they really going to the bottom?

Because we have seen in example that an investigation after an accident is merely a routine check like you do before delivery and nothing more. That’s not enough.

And if you have some description of how you do I think you should put it forward to let them international use because I think it should be put into standards that we can point at that then certification institute is notified (unintelligible) and so on all could share. Because that’s one thing we are thinking about.

Okay. Deficiency in communication with the relevant authorities from corrective actions in the market. Yes, you do it in one market, you don’t do it in the other market. You do it in Sweden, you don’t do it in Belgium. You do it in U.S. but you don’t do it in Prague and so on. There’s several examples of that.

And the last thing low parameters only for a client - one client without declaring it’s other uses. You do a fix by someone and then he’s satisfied and no one else is bothered by it although they should be because they have potentially the same problem. Not okay. Next slide please.

So I have put forward one question we got from the FDA. We have all got questions I believe. And one question, how can authorities academy use as patients and industry work together to improve the safety and efficiency of infusion pumps? And one thing we would put forward is focusing in the requirements for usability and risk management. And of course it has already been said and I just want to emphasize that we believe that as well is the most important part. And that has to be done in association with and have requirements for manufacturers to supply the intended purpose or intended use.

Because if we have intended purpose with the device, that is what it’s good for, what should it do then we - it hangs - we have something to put it together with. We cannot say that usability is good if you haven’t said enough specifications. So what should be do with the problem? So this is very important part. And if you look at the instructions for use and specifications and like market materials this is very poor. Really poor.

And also connected to the usability issue manufacturers must evaluate usability aspect with what actually applies to all perspective users. That is, I mean, an infusion pump - a manufacture for an infusion pump does a clinical evaluation in one hospital in let’s say Germany or the U.S. - done it for the whole world.

I don’t think it is because the methods - the personal different companies, they’re different. You have to look into what is different. Responsibility for nurses for example is different in different countries. But you have to do what you can do. What lies in the basic local method or routines or the local health care organization? You have to look into that.

If you don’t know it from the beginning of course you have to take in all the time. It’s the process of post-market surveillance to have an effective system to collect experience from the market. Next slide please.

We also need to address to the people who are buying, purchasing this devices that they are more - have a better process just to find the intended use and purpose. What is the device intended for when they are buying it? In order to match the local circumstances to the manufacturer specifications. We have to shake hands. The intended purpose has to match with the intended use from the manufacture - from the user. Sorry.

And the last thing is little sensible, maybe a little delicate. But the status of the management for infusion devices must be increased. It’s handled on the low level where risks are not appreciated.

I think when we are looking at the MHRA presentation later on there are a little bout that at the end that we have no respect - the user have no respect for these devices. And - because management, clinicians, doctors leave it to, well, it’s a simple task to asses an infusion. It isn’t. Next slide please.

And then the reporting and we have talked about reporting problems. And of course we have problems. We have problems to address that because people are afraid of reporting. Next please.

So at the end is the situation in Sweden and EU comparable to that in USA? Do w have the same manufactures? Yes, I think so. Same devices? Yes, approximately? Same users? I don’t know but I feel that there’s not a big difference, what I heard.

Same kind of patient’s? I believe that. Same risk? Should be then. Same kind of adverse events? Yes, we recognize what we are seeing. Number of reports? Well you have a little more. Of course. But it’s the level the same.

The outcome, the injuries - perhaps I can see a little difference because we cannot see so many injuries although the potential has been very severe. We have some deaths, yes we have that but some way we have - we are happy with that. I can’t explain that. Next slide.

So thank you again. It’s an honor to be here and thank you for the cooperation I see in front of us. Thank you.

Melissa Eakle: All right. We’re going to be going for a break in about five minutes. But before we do we’re going to actually put the MHRA presentation up on the screen and Margaret and (Tori) are going to toggle through it for you.

As (Matt) said we do not have a script but MHRA wanted to participate so much that they sent it to us and it does have some very valuable information. So please give it your attention. Then we will have a break and we will come back and we will be hearing from AvMed and (Amy). So thank you.

Woman: (Unintelligible).

Melissa Eakle: I forgot about the people on the phone. All right. These are the number of incident reports through a ten year period. Next. Then they do by type of device. You’ll see where pumps we often (unintelligible). Next slide. Fatalities and you can see where pumps are. Next slide. This is their analysis of how their adverse incidents arrive. God bless the English -- shortcomings in the device itself. Inadequate instructions for use, inadequate servicing and maintenance modifications or adjustments.

They call it user error, we will call it human factors including inadequate training, inappropriate management procedures and the environment in which it is used or stored and incorrect prescriptions. In other words, the wrong device for the wrong purpose. The intended use that (Motts) was talking about. Next slide please.

And this is their analysis of the causes. Again use error - it happened before the patient actually had the delivery of the drug, or after the delivery of the drug. Next slide please. They also have an issue with administration sets. Next slide please.

And again some more analysis. Next slide. Major findings, these are the percentages. Next slide. And this is an interesting slide because it says causes of adverse events, incidents for infusion pumps from 2005 to 2010. And you’ll notice that there is a very large red area on this pie chart where no cause was established. Next slide please.

This is also - this is you, (Sarah) do you see where pumps are? For human factors? Even in England - up there. Next slide please.

All right examples of types of incidents. They gave us - and again this will be up on our Web site next week. These are some of the causes of the - and some of the symptoms they had of their adverse events. Next slide please.

Typical things that a user can do. Next slide please. Now the interesting question that the UK came up with is are there equally complex devices? And they choose an anesthetic machine simulators and dialysis machines. And they - you’ll see that they looked at these very complex, very critical other devices in relationship to infusion pumps. Next slide please.

And they talk about how they are similar. Next slide please. And that was it? Okay, I thought there was another one. Can we go back to 9 - or the very last slide? My fault, I thought there was another slide. (Unintelligible) the last one? No, the very last slide. Number 9 - whatever the last slide is.

Okay. But all those devices can monitor flow in and out. But pumps only monitor what is going in. I thought that was interesting.

The ventilators and the dialysis equipment have built in safety devices such as pressures that can be monitored. Pumps only have it inclusion at pressure alarms. Ventilators and dialysis equipment are used by staff specifically trained to use them in a specific areas.

Pumps can be used anywhere and the question always comes up are the pump users as trained as other people on other device that are as critical. Patients may be connected to many pumps but only one ventilator or dialysis machine. And there are fewer makes and types of ventilators and dialysis services thus there is less to learn.

The message from the UK is pumps are critical devices. They are complex, they require a good understanding of the environment of use. The intended user, the patient population, et cetera. These are not your father’s automobiles so to speak. So that is the take home message from the UK.

Please have a nice break we’ll see you back in 15 minutes. Please for those on the line do not hang up.


All right we’re going to be getting started. If you all could assume your seats. Thank you.

We have heard from the FDA, we’ve heard from the user community, we’ve heard from the international community and now we’re going to get some perspective from the industry community. And the first speaker will be from AdvaMed.

The Advanced Medical Technology Association or AdvaMed represents manufacturers of medical devices, diagnostic products and health information systems that are transforming healthcare through earlier disease detection less invasive procedures and more effective treatments.

The AdvaMed members produce nearly 90% of the healthcare technology purchased annually in the United States. And more than 50% of such technology purchased annually around the world.

Numbers range from the smallest to the largest medical technology innovators and companies. And nearly 70% of AdvaMed members have less than 30 million in sales annually.

Representing AdvaMed and I want to make sure since he has already told me how to pronounce his last name, and I want to do it justice so (Krisnop Upakanorie) who is Vice President of Quality and Regulatory Affairs for Infusion Systems at CareFusion Corporation.

He has more than 19 years experience and operations quality and regulatory functions in the medical device industry. His experience spans various medical device technologies including infusion systems, electro surgery devices, neurovascular catheters and implants. And implants for cardiovascular on orthopedic application.

He has a MBA from the California State University in Hayward, California. A Masters Degree in Industrial Engineering from Wichita State. And a Bachelors Degree in Mechanical Engineering from Napier University in India.

Please welcome (Krisnop). Thank you.

(Krisnop Upakanorie): Thank you. And on behalf of the Infusion Pump Working Group and AdvaMed I want to thank FDA for providing us an opportunity to provide the industries prospective on the draft infusion pump guidance document.

Ultimately it’s the objective I believe of the FDA and other stakeholders like healthcare providers and clinicians to make infusion pumps safer to provide therapies to patients.

And to that end industry and infusion pump working group is committed to working with agency and other stakeholders to provide infusions systems that deliver therapy in a safe manner. Next slide please.

As most of you may know clinical requirements for infusions systems that we’ve logged over time from the first implementation of infusion pumps to what they are today.

And other the last 15 years or so the infusion pump have evolved in their utilization as the challenges they pose to users and manufacturers.

The industry has stepped up to some of these challenges and has implemented technologies to keep up with the challenges. For example, pre-flow prevention - okay just catching up with the slides, so catching up. All right, so for example some of the improvements that have been made are technologies to prevent pre-flow, rug libraries, implementation of hard stops.

But the point of this is not to boast on the improvements that have been made, but to acknowledge that there is a lot more to do. In fact there is a lot more to do than what has been done so far to make infusion pumps safer. Next slide please.

Infusion pump working group at AdvaMed (unintelligible) the draft guidance document and that is the right step - step in the right direction to enhance safety, and improve the robustness of infusion pumps.

I also want to point that currently infusion pumps are developed, modified and monitored to some regulations and various elements of the quality system regulation including design controls, post mark surveillance, risk assessments and so on.

And while we have these controls today and they provide adequate means to monitor infusion pump safety, we believe that the new guidance document adds additional robustness to make the infusion pumps safer and to enhance overall safety of infusion pump - infusion system (unintelligible).

We will be seeking clarity from the agency to help understand some of these new requirements. And to seek and to look forward to raise the implement these new requirements into the pump systems.

Some of the new requirements that I refer to are insurance case reports for example and are collecting chemical valuation. And these are some of the new requirements that are being added to the infusion pump industry today.

We will be as we mentioned working to collaborate with our stakeholders and industry and FDA to understand and adapt these requirements to enhance the safety of infusion systems.

We heard this morning from healthcare providers about the time it takes sometimes for recalls be implemented.

With that in mind I wanted to just mention that for safety related modifications whether it’s a manufacturer that discovers the issue or a healthcare facility that informs us of an issue, and such an issue comes up we all have one objective on all of our minds. And that is how do we make the modification or the improvement in a safe manner and deploy it a expedient manner so that the user is (unintelligible) benefit and we make them safe as they can be.

Our preliminary assessment is that the (unintelligible) documents have additional requirements and they can perhaps add significant time to obtain in clearance even for such modifications.

So we will be working proactively with the agency to seek for the guidance on how to implement such modifications in a safe and expedient manner. Next slide please.

When I touched on a few aspects of the new guidance documents there were many details in the draft guidance that will apply. And AdvaMed’s members will be working with the agency and other users to incorporate as appropriate these guidelines into the quality facility requirements if you will.

We are also reviewing the details of the guidelines at this time and will be providing comments and questions on the draft guidance document in writing. Next slide.

We have talked about throughout the day on primarily the user related issues, software related issues, validations and so on. And the draft guidance documents provides various tools and solutions to address these issues that were brought up.

And AdvaMed is willing to work with the agency and other stakeholders to better understand the requirements. So all of us as a combined unit can understand how to better deploy the new requirements so that we can make products safer and in a timely manner.

Example would be, how can we use the insurance case reports so - that cause industry that we all know how we are going to be deploying these out for our products.

When it comes to clinical evaluations obviously you know, they might have heard these questions. What does it mean when we talk about clinical evaluations how close are they to perhaps a clinical trial, or an implantable device versus you know, some other type of a device - approval process.

So, those are the areas that we would be seeking to understand upon the agency how to best deploy them as the last perhaps collaborate on holding joint education programs for our workshops.

We also have other stakeholders to consider standard commodities, clinical organizations and so on. And there is a lot to learn from those organizations in understanding workflow, clinical application and the clinical needs.

And we will do so to do - to better understand clinical needs to make the infusion pumps and delivery of infusion therapy safer. Next slide please.

Again, I want to thank FDA for providing us an opportunity to provide a prospective on the guidance document. Thank you.

Melissa Eakle: Thank you very much we appreciate you coming and sharing that information with us. I have to get my notes I haven’t memorized all eight pages yet and everyone’s bio.

All right, our next two presenters are from AAMI which is the Association for the Advancement of Medical Instrumentation.

It’s a non profit organization that was founded in 1967 and is a unique alliance of nearly 6,000 members from around the world who are united by one mission.

To increase the understanding and beneficial use of medical instrumentation through effective standards and educational programs and publication.

AAMI provides a unique and critical form for members who cover a complete range of interest from clinical and biomedical engineers and technician, physicians, nurses and hospital administrators to educators and researchers, manufacturers, distributors, government representatives and other healthcare professionals with an interest in medical devices.

These diverse groups have been instrumental in making AAMI the leading source of essential information on medical devices and equipment since 1967.

We actually have two people from AAMI. And I will be reading both of your bios first.

We have Dr. Nathaniel Sims who is a Clinical Teacher, Cardiac Anesthesiologist and Medical Advisor to Biomedical Engineering at Massachusetts General Hospital.

He is also an Assistant Professor of Anesthesia at Harvard Medical School. He is a strategic and hands on innovator who has developed numerous technologies that make patient care safer and more efficient. Working in interdisciplinary teams involving biomedical engineering, nursing and various other hospital departments.

Dr. Sims and his colleagues have pioneered improvements in patient monitoring, patient transport, and error free intravenous drug delivery systems.

The overall focus is developing advanced system technologies to improve safety in patient care while reducing cost.

Dr. Sims holds numerous US patents. He is working now through other organizations on product checks that move innovative technology into patient care.

Dr. Sims is the 2006 Winner of the AAMI Foundation Laufman/Greatbatch Prize. The highest honor given by the association. Honoring an individual that has made the unique and significant contribution to the advancement of medical instrumentation.

His development of smart drug infusion pump technology and flexible monitoring systems are contributions of patient safety technology recognized in the awarding of the prize.

At AAMI Dr. Sims serves as Vice Chair of the Medical Device Research and is Academic Co-Chair of the Infusion Device Committee. He is also a member of the AAMI Board of Directors. Please welcome Dr. Sims.

Well you were standing up I’m sorry. Introduced the wrong person. Just remember what I told you.

Man: I’m sorry.

?(Melissa Eackle)?: Okay. Thank you. I’m going to be introducing Mary Logan who has 20 years of Senior Management experience most recently serving as the Chief Operating Officer of the American Dental Association in Chicago.

As ADA COO she was responsible for the operational management of the organization. Most of its divisions reported to her including the ADA’s well respected standards area and it’s large science division.

During her tenure the ADA was named one of the most remarkable associations in the United States in seven measures of success. A report issued by the American Society of Association Executive.

Dr. Logan also served as the ADA’s General Council and is General Council to the Global Finance and Administrative Arm of the United Methodist Church.

Please welcome Mary Logan and then Dr. Sim.

Mary Logan: Thank you. That was a nice dance. I’ve been AAMI’s President for one year now and what inspired me to join the organization continues to inspire me today and I know that is why we’re all here today.

And that’s the passion that people in healthcare have toward better patient outcomes. All the staff at AAMI, all the volunteers at AAMI and all of you I know are inspired by the same thing.

So, thank you all for the work that you do to help me be inspired and happy to go to work everyday. Next slide please.

The theme for both Nat and me this afternoon is kind of a roundness idea of it takes a village. And when I first joined AAMI the first three months I as there was a time to listen and learn. And I came across this quote and used it in organizational assessment that I wrote for the board.

“Never doubt that a small group of full committed citizens can change the world. Indeed it is the only thing that ever has.” And that’s what we’re all here for today as well.

So I’ll be putting in a plug for standards, and also a plug toward a multidisciplinary approach to solve the current problems with pumps and to set a consistent bar for the future below which no pump will fall.

I also want to thank the speakers this morning who were part of the clinical engineering and biomedical engineering and technician community.

I was in the back doing a rah-rah for that community which is often in the basement of our hospitals around the country. And they are the ones who spoke so precisely and logically about real problems in the field. And they can translate it in a way that all of you engineers can understand.

I would be the one saying that thingy broke and I can’t fix it. Can you help me? Next slide please.

We’ve come a long way in 40 years. There is several slides here that show photos from AAMI’s 40 year history. And I find these fun to look at. I’m not sure what the little animal was in the hand of that guy. You can advance again to the next slide. Just have three.

And then the next one - one more, this one, to me shows we still have a long way to go. My imagination wants this to be the next home health care device that infuses new brain cells into the baby boomer brain. Is that possible, can anyone out here invent that?

And then the next slide my imagination says, this is where I’m going to go if I say anything offensive this afternoon.

Okay one more fun one. And then the next slide one more, this one is a flow chart from AAMI’s early days. And I’m not exactly sure I could even explain it, but I love this because to me it really illustrates the one theme from everyone’s remarks today about complexity.

And how complex the world of medical devices has become from these early days when we had those stand alone devices that people tinkered with. Next slide please.

You all should have a handout from AAMI that is really what my speech was supposed to be. And it is all about the AAMI Standards Program. And I hope you’ll take it home and read it. If you’re familiar with the Standard’s Program if you participate on our standards committee’s you won’t need to read it.

But I want to highlight just a couple of things from that handout for you. It includes and overview of the Standard’s Program for people who aren’t familiar with that.

It provides some background about the infusion device committee of AAMI and also the companies and experts and the clinicians who serve on that committee. And it also includes on Page 5 a project list.

The committee has - no pun intended, infused new enthusiasm for a very ambitious project list that includes perhaps most importantly so I’ll mention it first, they have a vision for an infusion pump safety council that would be a multidisciplinary group - ongoing group to address safety issues around infusion pumps.

The vision is that no patient should be harmed by an infusion pump it’s an extraordinary vision, sounds very simple, it will take a village to work through that.

We need time, we need expertise and we could use some resources and some sponsors to help us get this off the ground.

Second project that is important to mention is the summit that AAMI is in the final stages of working on an agreement with the FDA to co-host a summit October 5th and 6th. There was a handout you should have that, the dates are set, they’ll be more information forthcoming about that.

And the hope is that we will build from what happens here today and tomorrow and be able to take some of these discussions to the next level.

And then a third project that I want to just mention I think it’s so important the committee in the process of really looking at where we are with infusion pumps and standards has noted that there are 296 different standards out there from different organization addressing different aspects of pumps.

There are obviously some voids sand there is obviously some duplication and perhaps over standardization. So there are going to look at the gap and the overlaps between those standards and try to take a more systems approach to that.

The last thing I want to mention before we move to the next slide from this handout is please, take note of AAMI’s standard philosophy because I think it’s really important in the global environment that we’re in today.

Our philosophy is that there should be one - we should be able to work towards one standard worldwide for any given product. One standard worldwide.

We’re far from that, but I thought it was important to mention that in light of the prior speakers who spoke from a more global prospective. Next slide.

Back to this slide again. Okay, next slide - sorry. Next slide. Okay, I think we missed one but that is okay.

So why are standards important. First of all I think standards are important because of the multidisciplinary approach to solving problems. Everyone today has said something that you use the word systems in it.

And standards allow for standard setting if it’s done well allows for all the different stakeholders to come together in a multidisciplinary approach. In a way that no one stakeholder can do.

AAMI offers that opportunity because we have in our standards participation clinical engineers, biomedical engineering technicians, nurses, physicians other clinicians, experts on particular devices and industry and the FDA all working together around the table.

It’s also a cooperative approach to solving problems which in a market based system which the United States is always preferable to regulations when possible.

Having the FDA participate in the standard setting arena also helps to ensure that the regulatory environment is going - the regulators are going to be comfortable with the standards that are set. There is great input and output - conversation going in both directions.

Systems, a lot of the speakers today - all the speakers I think have used the word system in one context or another. And I think it is fair to say that in today’s world we have multiple systems and systems upon system.

So it’s easy to see why we have a problem with infusion pumps. Many patient care devices are complex systems connected to another complex system. So the next few slides I want to elaborate a little bit more and layer on to on this system of systems idea. Next slide please.

All right, back to that chart. This is what it used to be like. Next slide. This may be a simplistic version of what is really like today. Next slide.

Hopefully most operating rooms look a little neater than this lovely simulation lab but it’s still a nice simulation lab. Next slide.

So let’s take hospitals. A hospital is collection of systems all in one system. And when you have a system wide hospital with multiple hospitals it gets even more complicated.

So you have clinical workflow as a system, you have operations workflow as a system. The patient and the patients family is a system. IT is another system we won’t talk about that one. And then clinical technology and you could layer on more - there are more. Next slide please.

Clinical technology is a system. Many people want everything to connect with everything else. It does at our home, so why not in a hospital. Device to device, device to network, network to network, network to the EHR and the EHR to my iPhone? Next slide.

Industry is also a system of systems. Each device has a unique proprietary design that is what innovation is all about in this country. Design for a single device impacts many other systems and vice versa. The alarm examples that have been given today are just one small thing where one device does has an impact on what happens with another device.

Other system considerations, software, human factors, components, suppliers, quality system, standard system. Big companies also have multiple layers of systems where often the folks who work on the design of a product are in a completely different facility from the folks who work on human factors from the folks who work on software. How you get all of those systems to interact with one another. Next slide.

Government, another huge system just the FDA alone is a system of systems. With thousands of people. And then you layer on the other governmental agencies that have a stake here, CMS. Joint commission, not government but Quasi Government. I could keep going with that. Next slide.

So no one organization, no one industry can do this alone, it’s going to take a multidisciplinary approach.

When the FDA acts it impacts everyone else. When the industry designs a new product, it impacts everyone else. When AAMI sets a standard or some other organization sets a standard, it impacts everyone else.

If we don’t all work together, we’re going to continue to have problems down the road.

So my plug for standards is it allows voluntary multidisciplinary participation with regulatory input. And it is taking a village to solve a problem together. So imagine what we can do together that we can’t do alone. Next slide please.

AAMI needs to step it up. Some might ask well what about those infusion pump standards, they need to be better than they are. They need to be stepped up a lot. They don’t anticipate like all of us haven’t anticipated the impact of these devices and other devices on this complex system upon system, upon system in a world where everything connects wirelessly, or with wires.

Standards are an answer to some of the complex challenges, certainly not all. We can’t - no one of us can take that burden, it’s not fair for the FDA, it’s not fair for industry, it’s not fair for standard setting organizations or hospitals to take all the burden, we have to do it together.

It starts with a risk analysis which there have been a lot of great presentations here today that start with that. Identifying what risk can be managed through standards, what other projects are needed, like a safety council, like filtering through whether there are too many standards and which ones are the ones that really count and then developing a work plan. Last slide.

So my pledge to you is that AAMI will step it up. Working from the common ground that we all share for better patient outcome, we’re committed to bringing industry, clinicians, experts and the government together to create better standards.

We hope that our reputation will continue to stay here. That we’re an honest broker through our neutrality. Our stake is for better standards.

And we hope to bring best practices and ideas forward through our summit on October 5th and 6th and we welcome your participation.

(Matt) asked me what I would hope for at the end of the day today. And I said I’d be ecstatic if people came to me at the end of the day with their business cards to say, I want to help, how can I help. I have time, I have money, I have expertise, I want to play. Thank you.

Dr. Nathaniel Sims: How and good afternoon everybody. I notice that I am the caboose speaker here but at least we’re not behind schedule.

I won’t take all the time that is allocated because we go into questions and answers afterwards.

So just to recap the bio, obviously I have many roles I’m a physician at Mass General Hospital I’ve been actually there for about 30 years. I probably have 10,000 hours of infusion pump programming time as the cardiac anesthesiologist.

And in fact the motivation for everything that I’m going to talk about which is to some extent a walk down a memory lane. And look into the future is based in that hands on boots on the ground experience. The frustrations and the joys of delivering drugs to critically ill patients.

But I’m non only speaking as someone from Mass General and from the biomedical organization but also in the AAMI role which is with (Pat Beard) who is sitting in the third row acting as the Coach here of the AAMI Infusion Device Standards Committee.

A role which I have been I think off and on kind of for the last 10 years. And I agree absolutely with Mary Logan much that needs to be done to reinvigorate the standards process not only just in the US but in the organization in a sense.

And so what I wanted to focus on here today actually was to recap some the discussions that we had with the FDA with Al Taylor who spoke earlier today with Paul Jones, with Raoul Jetley from the FDA’s Office of Science and Engineering Labs - to begin to flush out what would be the sort of three or four major sort of flags we would put on top of the mountain of a new standard, what would be those core elements.

And a paper - white paper was written called Environment Factors Affecting Drug Infusion Safety which is not published yet. And I think really the time wasn’t right to do it until today occurred.

Today is an incredibly important milestone and I think it feels like being given a flashlight with which to look into the fog of the past, the confusion of the present and to see a little bit toward into the future.

So on the next slide we - it’s not exactly centered on the screen but that’s okay. (Unintelligible) not quite the whole slide there.

We - the focus of our conversation in the summer of 2007 was around the process of developing a safety assurance case. Really focusing on the development of a safety assurance case for a particular - blue screen here...

Man: (Unintelligible).

Dr. Nathaniel Sims: No I hope it does. That’s cool. Ah-ha right, good job. So the steps that Al and Paul Raoul wanted to try to teach us in our hospital was the notion of a orderly process that begins with a hazard analysis that has a conceptual safety model that’s at a very high level.

And then introducing some more specific issues, the potential for a standardized user interface. The potential for a model for a generic infusion pump that could be used to analyze software for a deviceless market and then rolling it all up into a safety assurance case.

But that thought of a safety assurance case on the next slide reminds me to point out to all of you that everyday we have to run a hospital. A hospital where we promise to provide world class tier to whoever who walks in the door, no matter how sick they are.

And this slide is a picture of our - one of our ICU cubicles in our Cardiac Surgical ICU. And it shows the physiological monitor, a bed that can move back and forth between the ICU and the operating room. A defibrillator sitting on the foot of the bed. And a transport monitor whose screen is there on the left. And a sort of a drug infusion pump rack with let me see what, 12 drug infusion pumps there.

The notion that that is often the number of drug infusion pumps that are on a cardiac surgical - a neurosurgical, a transplant patient. And that any one might possibly fail and you don’t have but really maybe 10 seconds to recognize that a pump has failed and to use the redundancy of yet another pump that is already there and it can be swamped out.

So I bring this to you to remind us that everyday - today, and tonight at 2am one of these cardiac surgical patients could need to return to the operating room in three or four minutes for a bleeding problem. And that’s the philosophy we which we as a hospital have to everyday operate a safety assurance system that is resilient to all of the challenges that we’re discussed on today.

I don’t know whether to - some in the room this picture looks messy, or it looks very tidy and neat.

But the next picture is what I saw when I came to work in sort of ’84, ’85 at Mass General. There is a terribly sick cardiac surgical patient whose head is in the lower right hand part of the screen covered by a towel, she is very, very ill, has an intra aortic balloon pump, has post cardiac surgery there is a physiological monitor on the left. A couple of large volume infusion pumps with coil cords going up to drop counters.

And then these great big heavy metal syringe pumps with adhesive tape labels on the front saying which of the drugs because she was so sick that everything had to be infused at very slow rates because of kidney failure, lung failure and heart failure.

That was our contact this environment of care was completely impossible to work with if that patient needed to return to the operating room. There is no batteries in any of those pumps. You just couldn’t do it. So we have come a long way. Next slide.

And in our conversations with Al and Paul Jones and Raoul. We thought back about this old metal infusion pump and recognized that the system boundary for the standard with which that might have been evaluated in 1960 as to its pre-market suitability had to assume - it’s the diagram on the left that said SLC that the front line caregiver is perfect. The frontline caregiver has read the manual. The frontline caregiver understands this pump has a gearshift, no display one red light when it’s on, no occlusion detection.

So that the frontline caregiver has to completely understand everything about that device and prevent any of hundreds of failures modes of that device, including electrocution if there is a ground fault et cetera from occurring.

So this diagram that Paul made that points out that to date the (unintelligible) process and the standard process have tended to assume a frontline caregiver who was perfect in their knowledge in order and however the reality of life as it is doesn’t make that a sensible concept to stick with.

So, beginning on the next slide ah, and this was - is a phrase, that paper one could argue that the frontline caregiver were competent innovation is a rather weak primary risk that is all measured. Great observation.

Next slide, shows that by 1987 or ‘88 we begin to have small Intel microprocessors that could store a memory of maybe a 36 or 64K have a small program in them. And we had LCD displays and we could begin to move guidance from the device to the frontline caregiver into the systems boundary space.

So this is a pump that was designed - I’ve covered up the logo. In the late 1980s which was one of the first dose rate calculator pumps. You could take a syringe of lidocaine, you could enter into the middle display the concentration milligrams per mill of the drug.

And then, you could enter the patients weight and it would help you make the complex calculation of micrograms, per kilogram per minute which (unintelligible) and with those old metal devices, the caregiver had to do on the back of an envelope and often did unsuccessfully.

So in the next slide illustrates (Paul’s) drawings that as the devices could have software microprocessors and displays the human interface and begin to play some part in the role of the medical practice of delivering a drug. We move the front line caregiver into the system boundary space for the evaluation and for the purpose of defining a standard. Next slide.

Begins to show other devices of that era which begin to even be specific about helping you give a particular drug. This is a wonderful device of which God knows a 150 - 200,000 may have made or used primarily in operating rooms.

In actually has a user interface with four knobs and magnetic plates - there were 24 magnetic plates made one for each of commonly given anesthetic drugs, muscle relaxants, (unintelligible) and so on.

And it’s almost inherently it’s actually quite a beautiful human interface because if you put the plate on the pump, you choose the patients body weight, you choose the infusion rate, and you choose anything about bolus. But you sort of you - you’ve got lower limits, upper limits, a middle range and you can adjust in a sort of perceptual way.

And I just point out this is an interesting early device that began to provide drug specific dose rate guidance in a sort of mechanical form.

So going on to the next slide we see by 1995 we began to have colored displays, we began to have micro processes that are big enough to hold substantial memories that have IL ports that allow you to load a software about drug names, drug practices and so on into the pump.

And this I believe is the first pump that was ever implemented in the sort of thousands throughout numerous hospitals where you began to see dose error reduction systems.

So it’s a two channel syringe pump. The right hand channel has (unintelligible) it says the concentration of the drug is such and such and the dose rate is two (mics) a minute I guess.

The left hand one lidocaine the drug name had been chosen, the user has used that trim knob at the top to dial in a proposed drug dose rate of 4.3 micrograms - milligrams per minute. And it says that’s greater than the maximum recommended infusion dose rate. Do you want to proceed or not.

And I think that is the sort of origin historically and technically from the point of view of the devices capability of bringing the front line caregiver into the device designed boundary space by allowing a medical practice to be sort of tossed up at the point of care.

On the next slide again would - it is just representing yet another form of modern device. This is the PCA pump it’s got a drug name up there it says, epidural infusion obstetrics bupivacaine 0.1% continuous rate of something or other.

And it has a user interface that has been very successful. And I would attribute that largely to having been tested extraordinarily intensely in a simulation lab in Toronto.

And I may be a little bit wrong about that story but I think one needs support for the notion of testing device design in the “The Environmental Care” I think sophisticated realistic simulation center can be a phenomenal place to refine user interface in this pre-market period. And that point was made rather sternly AAMI FDA Meetings in early March. Next slide.

So, however, the hardware’s made with so called smarter drug infusion pumps have still be regarded by a clinician is not sufficient to reach the ultimate goal.

And the way we try to phrase it in our organization is that there needs to be a seamless digital pathway between the thin apps of the ordering provider and the patient blood screen.

And that really implies network communication between the systems - the systems that Mary talked about.

And I’m not going to go through this I chart. But its simply says - this is (Paul Jones’s) representation, the system boundary for the evaluated process on the 510K side and also for the standard size has to take into account many more players.

Listed here are the provider organization, the pharmacy, the pharmacist, the frontline caregiver a label that has some positive identification or data transfer characteristics to it.

Some means of flow rate assurance monitoring beyond what the pump thinks it is doing the ability to put dose error reduction systems in a user interface that works alarms and so on.

But I think most important is this growing perception is that the devices need to be on the network in order to close that seamless digital loop.

So I’m going to now try to articulate the outcome of what was in the paper, but also in some meetings that our hospitals had with other healthcare organizations that are trying to push the boundary and getting some successes here into four or five bullet points for what we think might pop out as the forward trajectory for the standard to address this much larger system of (unintelligible).

So if we can go on - oh, and then I - making a pitch here for the involvement of the village. We believe that the provider organization particularly those that have huge amount of resources to invest in provider order entry, electronic health record, electronic medicine (unintelligible) systems and the future documented patient, these provider organizations have terrific teams who understand the complexity are constantly at work on each of the elements in this space, have a lot to say, want to say. And that those folks that are deep into sort of Level 7, Level 8 electronic health records are prototyping some amazing really cool stuff.

So, next slide. This is just a picture of any drug infusion pump but what I want to focus on I’m sorry it’s at the lower part of the slide is this is a Massachusetts General Hospital field syringe of some drug. It’s got a barcode on it a little tiny one of the left that is the sort of tracking bar code in the pharmacy.

But on the right hand there is three concatenated Aztec or data matrix barcodes that contain sufficient information according to a new standard that was written in Partners Health Care and adopted by the Emit Group to completely program a pump.

It has the drugs name, the pharmacy tracking number, it’s got the concentration, it may have the default dose rate and we created this in the hope and expectation that pumps would emerge that could read those bar codes.

The cool thing about Aztec and I don’t have any (unintelligible) right here, is that some of these bar codes can actually be concatenated together. So the constraints of a narrow label on a long syringe, you can pack more information by putting a bunch of bar codes side by side but they are actually all one bar code.

And you can get maybe you know, 180 characters of information. That is a lot of information and may be able to eliminate a lot of program errors. So positive identification that is kind of number 1 on my list. Next slide.

So two sides of positive identification. One is label on provider prepared ID drug containers that can transmit information to a recipient scanning device.

And the second one is an infusion pump that can read that. This is an example of prototyping at our hospital we had maybe 8 or 9 years ago a catastrophic incident involving chemotherapy administration.

A two look alike ambulatory pump one of which was programmable in milliliters per hour, and the other one in milliliters per day. You can imagine what happened, the pump for milliliters per hour was picked up and the drug put in it, patient sent home, came back in a couple of hours devastatingly ill et cetera.

And the hospital asked our team to immediately create a drug infusion system that allowed direct programming of a pump from a barcode that contained in chemotherapy it’s usually just the dose rate and some amount of time. So it’s not a very complex program.

And it turned out that there was a pump, the one of the right that was designed - and I’m going to step away from the mic. It’s a little barcode reader in the side that you were supposed to sort of rub over a label and it could read a barcode and program itself.

And what we decided to do was take a more sophisticated imaging bar code scanner and create a little electronic circuit that would flash that red light to emulate the scanning of a bar code so that you could really - a carryover could pick up a pharmacy container, scan it with that it would shoot the information to the pump. The pump would be actively programmed and we used that for I think eight years or so before a replacement technology was found.

But again, this level of energy putting into prototyping solutions that are sort of fused together but are better than the catastrophe we just had is the kind of energy that I hope can pull forward into the standard process as insight.

And so this simply says Number 2 on our list, automated drug identification, automated programming from information which bar code. Next slide.

This is going to be a little controversial to say. But this is a picture of a small drug infusion pump I guess you could call it or controller. I found that (unintelligible) I think it’s made in Indonesia.

I don’t know who the manufacturer was, I’m not too exactly sure how I got it I think somebody from (cache). But it basically is a small electronic device with a battery and it goes back to 30 years ago we had drop rate counters.

The problem with drip rate counters for controlling infusion pumps has always been that every drop is different, whether it is saline or dextrose, every drop size is different and at every temperature every solution has a drop size that is different.

But, if you know what the solution is, and you know what the temperature is you ought to be able to have in modern devices a look up paper that allows you to sort of bring back the notion of some sort of flow rate assurance secondary process.

And I just throw that up there because in our discussion with (unintelligible) group on that diagram that I showed is the notion of better flow rate monitoring.

And this is particularly - I don’t have time to get into an issue with secondary infusion to piggyback where nobody knows exactly which bag it’s going in, did the caregiver remember to raise the one bag higher than the other, did the check valve work all of these issues about knowing whether something that was supposed to happen actually did happen are still perplexing.

The next slide is really just a reminder about the challenge of secure wireless connectivity. It's a multidimensional challenge. I don't have time to go into all of the problems we are having with so-call WiFi networks as a means of data communications from medical devices into a server or an electronic health record.

It would literally take probably an hour to go through them all. But I will say the bottom line for our organization is that we have felt it is vitally important for us to protect our entire information system from weak authentication security because when you are a person in the hospital with a wireless laptop, it may have a fairly weak automatic authentication security process.

But you log on with your user name and your password and that changes every six months so there can be reasonable assurance that that laptop is not a rogue device trying to burrow its way into HIPAA protected information.

For infusion pumps, which you don't log into for obvious reasons something has to replace it and our organization has written very defined standards to say, and this is gobbledygook.

But WPA2 ETLS, what that really means is we have a certificate issuance server and we can assign and provision a medical device with a certificate and replace it every day or every six months or every two years so that when that device, the headless device is on the network, it is notably not a device that shouldn't be on the network with all the potential for harm.

This picture is a picture of a, I don't know, 12 infusion pumps with their little radio (kind of) showing there. We are deep in the process of validating reliable secure wireless connectivity. Because it's one thing to have a server post a new drug library to 2300 infusion pumps and have it take a week or two for them all to be turned on and have it loaded up.

But it is another thing to have a particular drug infusion pump on a particular patient posting its settings of the moment to a server to then compare whether where the order - that's much more real time requirement of connectivity and this is a huge - a huge, huge effort for us.

So the big four, positive identification of pharmacy provided drug containers; Number 2, pumps that are capable of reading information rich data on label; Number 3, thinking about flow rate assurance in a surveillance (tent); and Number 4, secure wireless connectivity.

So those are my headlines. The next two slides are a little bit more detail and this is the section of the paper that we talked about that's called tactical considerations. Because the first part of the paper was the strategic considerations which says the environment's more complex.

And I'm just going to read them quickly, not to talk about them in any detail and keeping a close eye on the clock. And this is not verbatim from this paper. It's tweaked a little bit.

Number 1, requirement to validate software with software assurance tools; that will be discussed a lot more tomorrow. Data networking and device connectivity assurance. Fluid flow rate assurance with improved sensors. Standardized workflow. Positive identification of drug, device and caregiver.

Standardized startup infusion pump display. Tieback of device setting or cross check to electronic health record physician order. And then finally and very pertinent to some of the catastrophes or issues that were discussed this morning, cross check with vital signs physiological status of the patient particularly when we're giving opioid infusions to relieve pain and opioid infusions cause you to breath less.

This is - I think it's a problem that was underrepresented in a data that came to the FDA is that the issues with opioid infusions and respiratory depression are probably more common than somehow is popping up in what is coming to you. So that was a list of tactical considerations that came out of this (OSEL) paper.

The next slide is the result of a meeting held between our hospital team and another hospital in the Western part of the United States whose name I'm not going to mention because they need to celebrate what they've achieved in their own way. And it's quite extraordinary.

And at the end of a meeting to see some amazing things that I'll tell you about in a minute that they've done, we decided that the list of the next generation features that Massachusetts General Hospital wants and that this other luminary organization wants are actually the same list.

That suggests it's possibly a good list. And what would happen if we then went and found six or eight large healthcare systems that were deep into progress with electronic health records and data connectivity. And wouldn't it be amazing if all of our lists of what we wanted was more or less the same? I think it would be very helpful.

What was the list? The list that we had was Number 1, strengthen pump wireless connectivity; 2, wireless software upgrades. And that - what that refers to is just like I'm sure every week when you turn on your iPhone there's a new operating system that may come down and you may not even know about it in the night. The notion of versioning firmware over the network is we believe achievable and also incredibly a desirable. I shouldn't have to reiterate why.

The third one is a little bit of a mouthful. It says gateway and database monitoring and reporting tools. Well if I've got 3000 infusion pumps talking to a server every couple of minutes or whenever there's a rate change or a action done on the pump, that's a lot of data. And we need to be able to pull out reports. Of my 3000 pumps, how many have been on the network in the last day, week, whatever? How many of them are reporting a problem?

I think that - and all those sort of inventory management biomedical issues. But the power of the data coming from pump to some place where it's aggregated and sliced and diced across the manufacturers, across a given hospital, across the country looking for best practices, it's a huge upside opportunity. But we have to build the tools to do that effectively.

Now the third - the fourth one that we've already discussed, medication - really pump setting verification backend crosscheck. So that means the patient is ordered for 800 units of Heparin an hour, we ought to know that the pump was set to Heparin in the correct concentration at 800 units per hour. And we don't know that now.

The next - no, I'm sorry, I didn't mean next slide. The next item was lower on the list. It's - there is a lot of talk about the notion of pumps should simply be programmed directly from the pharmacy and caregivers should accept that programming. We see that as a very, very distant goal for a sort of gigantic list of reasons. But it's on the list.

And then the last one down there, improve rapidity of drug library version wireless distribution. A lot of the times hospitals have trouble getting drug libraries out to a population of 3000 pumps simply because the pumps - half the pumps are on a patient. And you don't like to version a library when the pump's actually pumping but there are probably some ways to deal with that so that the process can be faster.

So I think I am about done. The next slide just, you know, reminds us again we are talking about changes to drug infusion, you know, that may take five years to roll out. But we have to run a hospital today and every day. Our hospital is about to celebrate its 200th anniversary.

And in a early discussion of some of the things that we're going to be talking about in our 200th year was a presentation about 750 people talking about six extraordinary cases.

One case of a man who developed an aortic dissection drove himself to the hospital in blinding pain, collapsed in front of the emergency ward desk, was immediately taken to the cath lab where there were infusion pumps supporting his life, studied.

Immediately taken to the operating room for an almost hopeless aortic replacement. Survived that, many days in the ICU, and then developed heart failure and then had a transplant. So everybody in the audience is hearing this amazing story and then the guy kind of steps up out of the audience onto the podium.

The next patient was a young woman who had a kidney transplant at Age 8 or so and very terrible complications from immunosuppressive therapy and she became the first person in the world to receive simultaneous replacement kidney transplant and a bone marrow transplant and she's off all immunosuppressive therapy at the end of - six months after that.

All of that depended - each one of these 150 beds with 12 pumps attached to them in the ICU ready for instant movement back to the operating room and emergency or to or from the cath lab is what we have to do everyday. And I think the ways in which we can deal with the imperfections of the (unintelligible) and still provide a world-class healthcare system is the joy of what we do.

So I'd see if there is a last slide there. So the summary in close basically FDA, AAMI, academics working since at least 2007 to define a conceptual framework for the scope and the look and feel of new infusion pumps (unintelligible).

At the moment we would see the SDO, Standards Development Organization responsibility as being vested in AAMI as has traditionally been the case and see no reason to change that.

We see the potential for various kind of user participation, not just the Safety Advisory Council, but organizations that are pushing the envelope clumping together to see whether their lists of the future are the same. They want to participate in defining forward trajectory.

And then the last remark here in the - well, we've got a couple minutes remaining. Always like to look for historical model for some uncomfortable space that we're in. And we're in an uncomfortable space today. Any time we have 700 deaths, 56,000 incidents and 84 (recall), that's got to be an uncomfortable space for everybody in this room.

We were in a terribly uncomfortable space just about at the time I chose to do cardiac anesthesia to join an anesthesia residency because 60 Minutes had a expo day on the fact that there were at that time approximately 2000 deaths due to anesthesia in the course of surgery in the United States every year. I mean that's outrageous. And it became a public outcry.

And so how do we get out of that mess? How did we go from 2000 deaths a year that were preventable and unnecessary to 10 or 20? How did that happen? Well, a group of - the Harvard Hospital had actually been interviewing anesthesiologists for four or five years to talk about near misses.

What are the things that happen or just happened or almost happened but I caught it and codify hundreds and hundreds of injuries and stories and published a paper called Human Error in Anesthesia and a bunch of - or I'm sorry, major errors in anesthesia and analysis of human factors and equipment failures.

And they analyzed that collection of stories and actually came up with five things which if anesthesiologists would do them, it could get better. Monitor the ECG. Wasn't done all the time in those days. They invented the pulse oximeter before it existed. They said we need to - we need to monitor continuously that blood is getting to the corners of the body every time the heart beats because an ECG alone does not mean there is (triculation).

Monitoring respiration was the most critical one of all because ventilator disconnect in the operating lowered (unintelligible). And so these sort of end-tidal capnograph and then several other things. Monitoring temperature and certain kinds of standardized equipment design and position of controls on devices and certain user training. Those were the core of the five principles that became the success story.

But most interestingly, those rules were published as Harvard anesthesia standards. The day those standards were adopted, the insurer for anesthesia in Massachusetts promised to drop the malpractice insurance rate in that state by one half the minute every operating room in the given hospital would implement pulse oximeter and end-tidal CO2 and the other rule.

And so I think that often - but, you know, until that clarification five things got articulated, people were in just as much of a fog and confused state about the crisis in anesthesia as we are today in the case of drug infusion therapy.

And I feel in my heart that we are close to articulating four or five big things which once we agree on what those four or five big things are, the standards process, the work of manufacturers, the work of the FDA, the analysis of incidents can - will suddenly - and that's why I started this talk by saying this meeting today is like giving a man a flashlight to see through the fog of (fast) in the future. So thank you very much.

Melissa Eackle: Well that was very exciting for me. I actually was alive during those times. It was a great stroll down memory lane. I'm - first of all I want to thank AAMI and AvMed. Your perspective is very valuable and we really welcome your participation.

Now the next thing is we're going to have a question and answer session. It's a short one today but if we don't get to the questions, do not worry, we have time tomorrow. We've built in a lot of time for questions. So I would recall all of the FDA staff from this morning to sit over here because I was just told that all the questions are for the FDA.

Speakers I'd stay just in case someone wants - oh we have one while we're waiting. Yes ma'am.

Bona Benjamin: Hello. I don't really - is this on? Can you hear me? I don't really have a question. I asked permission of the FDA to let me present the results of a survey. I'm from the American Society of Health System Pharmacists. And we surveyed our members on the questions that were asked by the FDA for this workshop. And I just would like to very briefly give you the results and we'll submit the full results after the meeting is over. Would that be okay?

Melissa Eackle: All rightie. We can give you a minute or two and FDA...

Bona Benjamin: Two minutes.

Melissa Eackle: ...two minutes. And ask...

Bona Benjamin: Two or three minutes.

Melissa Eackle: ...if you could quietly get up to the dais on this side. Thank you very much. Please proceed.

Bona Benjamin: Thank you. I'm Bona Benjamin. I'm the Director of Medication-Use Quality Improvement at the American Society of Health System Pharmacists. My organization represents 35,000 pharmacists working in hospitals and health systems and ambulatory care clinics, hospital outpatient pharmacies, home care and long term care.

In response to FDA's call for comment, we did a survey of 1400 of our members and in one week we received 185 responses with 91 respondents stating that they and their organizations were highly concerned about the safety of infusion pumps, five on a likely scale of one to five.

So I just want to very quickly ask the indulgence of this group to let me present these results to you. At the risk of being repetitive, at least they'll confirm a lot of what it seems like you've already come to the conclusions about in this (unintelligible).

The question was what problems have you observed in the clinical or home study. In contrast to FDA's findings that were reported on the Web site, instead of error messages, our members reported that there was no warning when incorrect information was entered.

Now obviously that assumes that the pump knows what the correct information is but there are some cases where this can be forced function to be correct. Other priority concerns were unclear instructions or prompts, key (bounce) - key (bouncers) and false alarms or failure to alarm.

There were numerous comments addressing malfunction where the pump didn't perform as expected and the lack of functionality. Some of them have been mentioned several times this morning. The only ones that I didn't hear was the fact that there's no standard nomenclature among pumps so that one pump for instance an (Elasto) (unintelligible) pump doesn't match a smart pump doesn't match a regular infusion pump.

So the nurse can't - so you have one pump stays while there's another pump continuous. And a nurse can't match an order with a drug bag and pump that is - as part of a crosscheck.

So these - just the brief conclusions and then I'm going to sit down is that we had some organizations who reported that they were very satisfied with the functionality of their pumps and that obviously some industry designers put a lot more thought and testing into pumps than others.

But the bottom line is with some of the problems that they've had, they continue to experience their problems with design and functionality and these are of a consistent significant concern to safety.

So I will give this little results - the full survey results to you when we submit our written comments. Thank you.

Melissa Eackle: Thank you for sharing those with us. We appreciate it. All rightie. I'm going to turn this session over to Mark Barnett. I'm not going to reread all the bios. You heard them this morning and you also have them in paper form. I want to go right to the questions. Mark.

Mark Barnett: Okay. Thanks very much. Oops, wait a minute. Let me (see) this thing. All right. We're going to have to pass this mike back and forth. That's fine. Let me get my papers straightened out. I have a lot of papers because we have a huge pile of questions from you guys. So thank you very much for that.

We'll get to as many of them as we can and then we'll pick up the rest tomorrow. Some of the questions were about the FDA guidance. And I purposely put those aside and we'll do those tomorrow.

My compliments to you. You heeded my advice about penmanship this morning and this afternoon's questions are much neater than this morning's questions. So that's great. Okay.

Woman: (Unintelligible).

Mark Barnett: Yes. Okay. Let's start. This is kind of in random order. This one says you refer to user errors as human factors problems. Are there any errors that are the fault of the user or is pump design always the root cause? If there are true user errors, do you have any breakdown of how many of these there are versus human factors and design problems?

That's a question that comes up over and over with human factors. And that is well, you know, somebody just had a bad night and it's just their fault and that's all there is, you know, that's all there is to it. You hear that a lot. I wonder if you as the panelists have any comments on that? Yes.

Man: I'll start and see if anyone else has any comments. Can you hear me back there? Okay. I have a colleague, Ron Kaye. Some of you know Ron. He would tell me very clearly the word is use error, human factors. User error is probably not the word we would prefer.

It's multi factorial. It doesn't necessarily mean that - we tend to look at how the device is designed that contributes to use error. People make mistakes. That does happen. And we're not saying people do not make mistakes. We're saying is there a contributor that can be ameliorated through the design.

And that's sort of what - where we want to go. It's not about blame. Again, I want to get the blame part out of it. But what we do from our side is we regulate the devices and we're trying to make sure that the device design gets to as good a design as possible so that the user doesn't have to deal with some of the issues that we talked about today.

Sure, users make mistakes. No doubt about it. We're humans. People make mistakes. But can we improve the design of the products to reduce those? I think that's where we're going. Do you have a comment on that Al?

Al Taylor: I guess the only thing I'd want to add is that someone asked this morning - is this - this is coming through right? Yes. So someone asked this morning when are we - when can we expect to see better pumps; but I think the pictures that (Nat) showed you just a few minutes ago, we already have better pumps.

You know, the Institute of Medicine Study a couple of years talking about use error. We've already made advances. And there's just a ways to go.

Mark Barnett: (Unintelligible). This is another one that's sort of related to the user. It says since many of the examples of problems with pumps are at least partially due to the -- it says there, I'll have to read it -- clinician error, are the joint commission or any professional organizations participating in this initiative?

Al Taylor: I guess it's me again. Did you - oh I thought you had - I thought you were going to say something. No. Well I guess the answer is that - you did - oh that's great. Mary has an answer for you.

Mary Brooks: It's not much of an answer. We have tried to reach out to a lot of people with the stakeholder calls and so forth. And this is, you know, we definitely know that we need to be working - sorry, trying to pull it away from my mouth. You know, we definitely have more opportunity to work with other agencies and DOD and VA and definitely with the joint commission.

We're just not there yet. We definitely do want to have a better relationship with our hospital systems out there. We're taking it one step at a time. And this is one of many steps that we're going to be taking in the future.

Mark Barnett: from two different people about the U.S. approach to evaluating and regulating these devices versus apparently those elsewhere in the world. So let me read those too. Can you comment on FDA's requirement for pre-market clinical evaluation versus another approach of enhanced surveillance? After the first market introduction, more devices in Europe seemed to follow the latter approach.

Now let me read the other question because I think it's probably a combined answer. The other one says based on the new FDA requirements including clinical evaluation prior to market introduction, is it likely that new modified pump technology will be introduced first outside the U.S.? Will the FDA consider outside U.S. experience and form as part of their 510(k) decision or in lieu of a clinical evaluation in the U.S.?

Woman: (Unintelligible).

Man: I like this one better.

Woman: (Unintelligible).

Al Taylor: That was a test. Okay, so the first question was why are we doing clinical evaluations? Essentially why are we doing clinical evaluations in the U.S. when the other markets don't use more of a...


Mark Barnett: ...both questions really.


Mary Brooks: ...information from overseas.

Al Taylor: So let me take the second question first. We will take valid information from other countries because we do that in our pre-market process today. As long as there's a good relationship and it can be shown to be appropriate, I think that's a valid way to go.

I did hear some comments when there was a question about whether or not they are the same that maybe they aren't. I really am not in a position to say that myself. I would rely on my clinical colleagues and other folks out there who know the environment better than myself to comment on that.

As far as the clinical evaluations as opposed to what other countries are doing, I think it's fair to say that, you know, our system is in some cases different than the way the other countries operate their systems. But they have, as they pointed out, similar problems.

Our approach has been to take a little bit more course approach to it to look at the actual use of the product in its environment. And as we said, the assurances cases which is, you know, one way recording information, could make a case why other clinical environments are appropriate in data in other countries work. So we are willing to accept arguments in that area.

But we have decided that the clinical evaluation in this country with the - under the criteria that we defined in the guidance is the best way for us to get the information and make us feel like the actual environment of use has been appropriately addressed in the way these devices in the United States are marketed.

Woman: (Unintelligible).

Melissa Eackle: Let's ask some of our international folks here.

Al Taylor: Hopefully that won't blow up on you.


Al Taylor: I feel like I'm holding a grenade or something.

Swedish man: This is working. Okay. Well just to comment on the different principles, the pre-market situation and post-market situation. And it's a good question because it's discussed right now in the EU. If the pre-market part is maybe too weak but at the same time their regulations say that the manufacturer has to with pre-market evaluation, clinical evaluation of the device, clinical trial if that is motivated and so on.

So the requirements, they are there. The difference is perhaps that the different approach is how third party that is authority and so are involved in that. The post-market part I believe is the same here than as in Europe.

The important part in the post-market is when we have clinical information that is not quite sufficient and we have a term post-market clinical follow up - I think you have maybe heard about that. That you talk about it very intense market survey with a small amount of devices put on the market directly after so you can monitor them especially - make a special surveillance of those.

I think to get product on the market you have to have a make of pre-market, post-market and that situation in between when the product is new in market. I don't know if that's enough on the question.

Mark Barnett: This is a question from someone who is concerned about blaming everything on the user. It says I have been a nurse for 30 years and each product issue rarely is the response from the supplier ever anything but user error. Is there a mechanism or process to ask FDA to get involved should there be disagreement between the facility and the supplier about whether this really was user error? And then she says, or he says at the bottom supplier need to be held accountable.

Woman: (Unintelligible).

Mary Brooks: Well I would definitely ask that clinician who is having difficulty to make sure that they file a voluntary report form with MedWatch and share that information with the agency and to make sure that information is thorough to the best of their ability. And if they are working with a particular facility, then they need to make sure that they report that finding also to the facility.

And if it warrants a mandatory report under the 3500A so if was a user facility that involved a death and - or a serious injury and they were not aware of who the manufacturer was, they can report that to the FDA. But by all means, everyone has the right to file a voluntary report with the MedWatch.


Melissa Eackle: And also through the Office of Compliance. And we are processing recalls. You heard in my earlier statement (the terms) of submitting your supporting documents (for it to be used). And usually when (you receive this error), it asks for the data that you (unintelligible) out. What (unintelligible) to show that (it really is an error).

Mark Barnett: Okay. We have anything more down there? That's it? Okay. We have a live questioner.

Pat Patterson: Yes. I am live. I'm Pat Patterson. I'm with Agilis Consulting and we provide human factors engineering to the medical device industry. And to tell you it's scarier with the mike over here than it was yesterday with the mike up there. And yesterday and today we're hearing a lot about human factors and use error.

And use error I think if folks aren't familiar with it is a confusing term. I appreciate Mr. (Watson)'s response to that question. I'm just wondering if the agency might consider that in its data gathering forms, electronic data gathering forms, maybe there is use error there. There is a bit of instruction about what use error is.

Maybe there's an opportunity to drill down just another level or two to help people who are not familiar with the term because it's not about blame but it is about what are those interactions that the manufacturers didn't expect, the users didn't expect and maybe if we can help people collect that data a little bit easier at least initially, it might facilitate exactly how great the anesthesiologist helps, you know, with their professional thing.

The other thing I would offer to people is AAMI has a - I believe it's a three day workshop on human factors that my partner Bob North has been heavily involved with and it's been getting good reviews. So for people that might be something to consider. Thanks.

Mark Barnett: Thank you. Lets use this, this seems to be safer. Safer on our eardrums.

Mary Brooks: I just want to thank you very much. This is the type of information that we want from the workshop. That's why we're having a two day workshop is to hear from the clinical community and also the manufacturers.

So I applaud Ms. Patterson I believe it was for coming forward and bringing us a very good statement about data gathering and whether or not the forms need to be changed so that we can drill down a little bit more for the use error, user error/human factors all the different labels that it has.

So maybe we could improve our forms a little bit better and I'll bring this information back to RSMB and Office of Surveillance to see how we can improve.

Mark Barnett: Let's go to another live one out there. Go ahead.

Louise Buelow-Smith: Hi. I'm Louise Buelow-Smith. I'm a nurse from St-Boniface Hospital in Winnipeg, Manitoba. And for those that don't know where Manitoba is, it's also known as Northern North Dakota. So I have a few observations, a question and a challenge.

The first thing I work in equipment (unintelligible) position. I have a role in infection control and it also extends into sterile processing. But as someone who gets the marketing pitches from vendors, one of the challenges that we're facing now is that the equipment that's being presented as so wonderful and safe and perhaps looking to the future when we have all the WiFi technologies, et cetera, that perhaps we're missing the point of that basic functioning that we need today.

And so I would like to make the comment and the challenge to the manufacturers is that the future is not now. As we move towards a more seamless medication administration and the use of these infusion devices and other IT and computer technology that there are going to be many facilities who cannot get there at the same time as others.

And certainly from where I'm coming from in my region my facility is well, the only one that's got any WiFi capability at the moment. And that's one of seven within the capital city.

A second observation, the person, clinical engineer who presented from Brigham and Women's this morning, I have to say it took everything in my person to not stand up and say amen at the end of the presentation. The clinical engineering folks are spending an awful lot of time and resources trying to maintain equipment for us in nursing. And the challenge is not always possibly captured. So thank you very much for that presentation.

The question I have actually is around the term human factors testing. So I'm really glad that the lady before me got up and gave some information. Now when pumps are being presented to us, we're being told that human factors testing has taken place.

Now as beginners in the understanding of what human factors testing is, we need some help understanding what quality of human factors testing really constitutes a good human factors test or set of tests. So I'm wondering if that's going to be looked at. You cannot put the claim of human factors testing in your marketing literature unless it actually meets the definition of human - good human factors testing.

I'm thinking back to the time when it became very popular to put calorie reduced or light on foods and it took a while for us to understand and appreciate that that needed to be defined. So I'm just wondering if that might take place.

And then I'll just finish off with the challenge and it has to do with the issue of under reporting. I cannot tell you the number of times I've picked up the phone to speak with the manufacturer to hear the statement I don't know but you're the only one having the problem.

Now, to the manufacturers, vendors, distributors, unless you can be 100% assured that that person that you're speaking to is the only one that you've sold that product to in the world, you cannot say that you are the only ones having the problem.

And it's very, very destructive to trust when after a while the hospitals and different facilities and people talk and they find out you're having the same problem. Because now when you say something to us about a fix, we may not be so quick to believe you.

Also when we tie it to (unintelligible) an issue, there's nothing that's more of a conversation stopper than to say to somebody you're the only one having the problem. What's the inference? You must be the problem. It's user error. If we want to engage both industry and the users in reporting, there's a expression in lean and it's called just do it.

And I think today if we wanted to talk about getting the better pumps to the market faster, we could say let's stop putting each other down when we express a problem and that it would be listened to and more probing questions would be asked so that better information would be obtained for the manufacturer and the regulatory agencies. Thank you.

Mark Barnett: Thank you. Thank you. Thank you very much. Does anyone want to respond (unintelligible).

Man: The only thing I'm going to say about human factors is that I'm not an expert in that area. And, you know, I think there's probably a bunch of different definitions for human factors out there.

And I don't know if there is a - like when the government came out and FDA came out and defined what light means or reduced calorie and all those different things, they put some boundaries on it. I don't think there's going to - I don't think we have a definition today that defines when something has been human factors tested.

What we do know is we ask for certain information and we give boundaries. We have a human factors guidance at FDA and there's a human factors standard out there that people can follow. And those two combined produce a product that's unique for one device, maybe not the same for another and we say that there has been some testing done.

Is it perfect? Probably not. I mean we're still working on that part. But one of the things that we're - one of the reasons why we asked for the clinical evaluation is to cover the gap between possibly what is done and the - maybe a simulation lab or in some small group of people to what is in the clinical environment. So hopefully we can get it - we can do that better.

Mark Barnett: (Guess) that will add to the levity of the situation if we dump the water pitcher over while we're moving the mike around. Okay. So I'm struck that back when FDA was getting ready to publish the quality system regulation, we had a significant educational outreach concerning design controls and human factors engineering.

And that was back in 1996, 1998 timeframe. And as we've been preparing for this workshop, we kind of looked back at what we did. So one quick answer is that FDA does have a couple of guidance documents that are on their Web site that will give you a tutorial in human factors.

And it jus seems like some of the same problems that we were having in the 1990s, it's not 15 years later and we're still seeing those problems. And I don't think that means that we failed because a lot of companies have moved way up the mountain. What I think we need to do is rededicate our efforts.

Man: (Unintelligible).

Man: No. No. You...

Man: (Unintelligible) besides in forums like this, where will the findings for (unintelligible) get published? (Unintelligible).

Man: So where will the information that I presented earlier be published? Haven't thought...

Man: (Unintelligible).

Man: Well, certainly the slides will be on our Web site on Tuesday of next week. I think this is the beginning of a dialog and we're going to be talking about this a lot and with you, not to you.

Mark Barnett: I shuffled the questions around a little bit because this next one really applies to the last comment that was made on the floor here about under reporting. It says - this one says the under reporting of adverse events involving infusion pumps has been mentioned several times.

What factors lead to a decision by a user facility not to report and how can reporting be improved? And that goes back to the point that was made about calling and being told that you're the only one in the whole world that has this problem. But anyone want to comment on that?

Woman: (Unintelligible).

Woman: Melissa would like to start on that.

Melissa Eakle: From my days as an analyst and talking to user facilities, one, they may not know that they have regulatory reporting responsibilities with death. And so that's one reason that they don't report. I will tell you as a clinician, I never knew that those problems that I was having with any number of devices actually some of them wanted to hear about them because they basically just got in the way and we worked around them.

Even when they made some really bad problems, I can remember defibrillating someone who got a fairly significant burn from a bad pad. You didn't think of this as being one in many. It was just a very - something you were used to and as Al Taylor always tells me we've got to raise the bar of our expectations.

We have very low expectations of pumps so when bad things happen we tend not to report. There's also that we're very, very busy. I don't know a clinician today who is not overwhelmed most of the time by the workload and the environment.

And unless we make it as easy as possible for them to describe the failure and to be able to report, it's going to continue to be very difficult. Now over to Mary.

Mary Brooks: Thanks Melissa. I think you said quite a bit there. We did look at, and this is available on the FDA infusion pump initiative Web site. We put a message out to clinicians to report and to hospital administrators and managers to make sure that there are protocols in place for reporting. You know, there is a requirement report and I think a lot of - a lot of hospitals just don't realize it.

So I think getting the word out, having workshops such as this, having RSMB, they also go out to several workshops and trade shows and they spread the word as much as possible.

So if one person in this room tells another person in this room and then they go forth and tell another person, we'll eventually get the word out. But it's just not the quantity of the report; it has to be the quality of the reports as well. It doesn't do me any good if you don't identify the pump or the model. You just tell me pumped failed but you don't tell me the make or the model.

So it also needs to be quality reporting. I'd rather receive less reports of high quality. I'll get more information out of that. Tomorrow we're going to explain a little bit more about the 3500A key elements that need to absolutely be in there that are required.

And then also, you know, how we physically use the 3500A and how we are able to get information from that to give back to manufacturers so that they can, you know, take that information and build a better pump.

Man: In out litigious world, liability is always a huge issue. And if you think that your error might have contributed to the problem either as an individual or as a institution, it's a disincentive to tell the FDA about it. This is a problem that the aviation community had. Air traffic controllers can go to jail if they make a mistake however innocent that mistake might be.

And so the FAA actually set up an anonymous hotline that any pilot or air traffic controller could report a near miss or a problem. And this is something that has been discussed within CDRH and I think within the medical device community. It's something we might be able to do something more about.

There were echoes all this morning, you know, in the bad old days surveillance or vigilance was passive. We would put products on the market and then we would have operators standing by to take your call. And what we're finding now is that that's not enough.

Some manufacturers when they are deploying new products, they're sending those products out with their people to see how they're actually being used until some experience is gained.

Our MedSun program is active surveillance. Instead of waiting for the reports to come in, we're going out and looking or problems. So I think that active surveillance, active vigilance is another wave of the future.

Mark Barnett: I'm going to ask our moderator now - I've got about nine more questions and it's getting late. Do you want to call a halt now and we'll pick them up tomorrow?

Melissa Eackle: Yes, thank you. Thank you Mark. Oh, I'm sorry. We have a live question. Let's take one more.

Mark Barnett: Oh yes.

(Steve Preggleman): I'll make it quick. My name is (Steve Preggleman). I'm with (Hostera) and my group has responsibility for human factors testing. And the guidance document that you referred to is extremely helpful but as you also mentioned, it's about ten years old and has some gaps.

And my question is whether the agency either is planning or would consider issuing an updated guidance document for us.

Man: I think the answer to that is yes. I don't know when. I do think it's in the works.

(Steve Preggleman): Thank you.

Man: That's all I can say about that.

Woman: I'll put a plug in for standards again. There's a brand new human factors standard that just came out of AAMI that - I'm not an expert. I understand from experts it's fabulous and much improved and greatly needed. It's enormous.

So I would encourage the collaboration that we have among industry clinicians, users, FDA and AAMI to take a look at that standard. That seems like great guidance. It has a testing chapter in it someone just said.

Woman: Thank you very much. All rightie. We have one last speaker. That's Mr. Tony Watson. I'm going to ask him to come over and join us. We start again tomorrow at 8:00 am. And for many of you, the guidance, you're all looking forward to actually having that explained because it's the first of a kind.

So we invite you back tomorrow. You have the agendas. Registration begins at 7:00 and we begin promptly at 8:00. And now I turn it back over to Mr. Anthony Watson.

Anthony Watson: I think it is fair to say - can you hear me back there in the back? Okay. All right. I think it's fair to say that a considerable amount of work has gone into the events that bring us here together today.

This morning I asked you to listen carefully to the story in each presentation. We learned of the large number of NDRs for infusion pumps and other signals that indicated to FDA that some action was necessary. Yes, we know that this is only but a fraction of what should be reported.

In addition, the data although voluminous are not particularly rich. A significant amount of detective work was necessary to develop a cogent analysis.

Another data source that proved useful for us was the recall database. The recalls allowed us to assess the root causes of device failures. These root causes are analyses provided by the manufacturer. Our three big buckets or root causes are derived from these data.

By scouring the data we were able to discern patterns in the data pointing to the most likely causes of failures. Human factors, software and designer manufacturing, those are the three buckets in the guidance document.

Al Taylor from our Office of Science and Engineering Laboratories has shown us how these problems are preventable. This is the foundation of FDA's story of infusion pumps.

We also know that FDA is not the only group with a story to tell. We heard from the user community. They told us that sometimes simpler is better. We also learned that manufacturers and vendors do a fair job of talking to users when a problem occurs but perhaps could do a better job of being transparent about similar problems occurring in other facilities.

In fusion pumps also we know today consume a disproportionate amount of service department budget. And last, we also heard that pumps need to be tested in the environment of use. I think we heard that in several different venues including from out foreign regulators.

The manufacturers in AAMI offered a compelling perspective as well. Specifically the industry is committed to working with FDA and other stakeholders to improve infusion pumps. That's good news.

We at FDA will have to be clear about how we're going to implement some of the requirements of the infusion pump guidance such as assurance reports and clinical evaluations and perhaps even offer workshops. I also heard that which would be - which is a great idea. I wrote that down. Hopefully we can make something happen in that area.

There's also a concern - I think this is a big one and it's a valid one about delaying clearance of products. That is a big one. We understand that. But as I think we also heard earlier today as much as we can get upfront to save us in the backend, we want to do that.

AAMI is looking forward as well to wide participation in the standards development. So I think I'll put another plug in that - for that, so.

Woman: Thank you.

Anthony Watson: I heard participation but I also heard money in there too, so.

Woman: Yes.


Anthony Watson: We don't worry about the money but I understand money is important there. And I think also we heard about the complexity of the infusion system.

Our regulatory colleagues from other countries provided us with a complimentary window into their experiences with infusion pumps. So thank you for that.

We learned that the - particularly in Canada have similar pre-market requirements to what we have in the U.S. They are also seeing similar problems in their reporting systems.

They also are in alignment with FDA that better collaboration is necessary to correct the problems throughout - really throughout the world. The root cause investigations by manufacturers also could be better in their countries as well. That is the feeling.

Each group provides an important piece of our puzzle by either reaffirming what we already know or by adding new pieces of information. The questions were insightful and provide clues into what our stakeholders consider important.

I would say that this is a successful first day. I also want to bring out that we heard, as I mentioned earlier, we heard a lot of positive feedback on this initiative and not so much in the area of negativity.

However there is one thing that I wanted to bring to everyone's attention because I think it's important to note. If we had heard that because of the perhaps extra burden provided by the guidance document that this might be too hard of a country to get a product through and that perhaps, you know, some manufacturers might consider going overseas.

I think that, you know, those decisions are made - those are business decisions they have to be made. But we haven't done an economic analysis. We're not in a position to say whether that's a good idea or not.

But I would just say that based on what I heard from our colleagues across the sea, I don't think there's a safe haven there either. Just a thought. So we think that the best thing to do is just let's get it done. Let's just go ahead and approach the problem and fix it.

We're willing to step up. I've heard that every - you know, I've heard a lot of people stop me as I'm actually trying to make my way to the bathroom and, you know, tell me how enthused they are.

So I think I really do believe that people want to fix this problem. And we can do that. We want to make infusion pumps. We want to change the mindset of infusion pumps from unnecessary evil because they are necessary and they're critical to the care of patients to an actual ally of the healthcare worker.

Tomorrow we will explore the infusion pump guidance in depth. I want to remind you that this workshop is not intended to replace the regulatory process. So anything that - any questions that you ask us today are there to - we're going to provide answers to help and form any comments that you want to provide to the docket but we as that you provide the comments to the docket.

We hope the questions that you bring to the discussion tomorrow will be as insightful as the ones today. So I'll just close by saying please join us tomorrow and have a great evening.

Melissa Eackle: And once again as you leave, I just want to thank everyone who participated and especially all of the people who sat up on the dais and spoke. Thank you.