From: Paul D. Cumming, PhD [paul.cumming@talmed.com]
Sent: Saturday, June 22, 2002 6:12 PM
To: fdadockets@oc.fda.gov
Cc: Edward L. Wallace Ph. D.; Louis M. Katz M. D.; Laurie Rogenski RN
Subject: Comments on Draft "Guidance for Industry: Streamlining the Donor Interview Process: Recommendations for Self-Administered Questionnaires'' 

Center for Biologics Evaluation & Research 
Food and Drug Administration 

Department of Health & Human Services

Via Email to: fdadockets@oc.fda.gov 

RE: Request for Comments

Docket No. 02D-0080

Draft ``Guidance for Industry: Streamlining the Donor Interview 

Process: Recommendations for Self-Administered Questionnaires'' 

To Whom It May Concern:

Thank you for this opportunity to comment on the above cited draft guidance.  It is noteworthy that Talisman Limited is the producer of the Quality Donor System (TM) (QDS), an audio video touch screen computer assisted self-interviewing system (AV-T-CASI) for blood donor screening.

Our concern is that there is a great deal of scientific literature supportive of the draft guidance which extends the Turner work cited in Reference 4 which is not cited in the draft guidance document. The cited Turner work (1998) showed that audio-CASI increased reporting of embarrassing, stigmatizing or illegal behaviors by up to a factor of 17 compared to paper self administered questionnaires (Table 2 adjusted odds ratios). By our count, about one-third of AABB Uniform Donor History Questions fall into these categories. A sampling of additional literature is summarized below.

In a September 1992 JAMA article, Steven Locke and others reported that in a sample of 272 blood donors using conventional keyboard/on-screen text only computer assisted self interviewing (traditional CASI), 12 donors reported HIV risk behaviors or symptoms not identified by routine Red Cross written questionnaires or face-to-face (FTF) interviewing. The problem with this mechanism is that it presumes donor literacy and literacy is a large and often hidden problem in this country. According Census 2000 over 31 million US residents are foreign born, 52% are from Latin America, and 21 million speak English less than "very well."

People go to great lengths to hide their illiteracy. The American Medical Association Web site (http://www.ama-assn.org/ama/pub/article/3216-4469.html) documents that sixty-seven percent of patients with reading difficulty admit they have never told their spouse about their reading problem. If two out of three illiterates can hide their deficiency from their spouses, how then are blood center staff to detect the problem? Isn't it better to use technology which reduces or eliminates the literacy problem?

A more recent/2001 article by a Turner group (attached - A.A. Tayyib & Others) provides the best estimate we are aware of as to how great the illiteracy problem may be among blood donors. With a sample of 1014 Baltimore MD adults aged 18 to 45 (probably from an STD treatment center) using the Rapid Assessment of Adult Literacy in Medicine (RAELM) instrument, via Table 1 the authors note that 18% of subjects with "Some college or a 2 yr degree" were reading at levels of 8th grade or below.  The "Some college or 2 yr degree" group is sometimes cited as typical of blood donors.

Also in 2001, a Turner group led by P.C. Cooley (attached) addressed "Using touch screen audio-CASI to obtain data on sensitive topics." The article begins by reviewing the advantages of audio-CASI: eliminates literacy requirements, can be multilingual, standardizes presentation and has the ability to accommodate complex skip patterns. The added potential benefits of touch screens are listed as: 1) enlarging the strike area (the on-screen button) thus reducing likelihood of keying errors; 2) specified/constrained response categories eliminating the need for range checks; and 3) pressing a labeled box reduces undetectable keying errors.  In short, audio touch screen CASI (A-T-CASI)  has the potential to "yield more accurate recording of responses" but may compromise the privacy of the interview via the large, visible strike area. To learn more the authors drew a pilot sample of 108 STD clinic patients aged 15 to 39.  For the 101 of 108 subjects that reported a mode preference, 51% preferred A-T-CASI compared with 26% for keypad audio CASI and 23% for the interviewer administered mode. When comparing just the computer procedures, 70% preferred A-T-CASI. Respondents noted that "touch screen technology was more innovative and 'fun' to use . . . ." The authors expressed surprise that touch screens were viewed as a more private. "One-half (49.1%) of respondents reported touch screens as more private, compared with 27% . ." for traditional audio CASI and 23% reporting no difference. In concluding the authors note the need for additional studies but offer that this pilot study suggests that within a clinic based population touch screens are preferred by a majority of patients and see touch screens as easier to use and more private. This "... holds promise for improving measurements of sensitive behaviors and hence improving the quality of behavioral research." 

In December of 2001 our 1999 Hoxworth pilot work on AV-T-CASI was published in TRANSFUSION with Thomas F. Zuck, MD as lead author. Results have since been corroborated and expanded via studies at the Mississippi Valley Regional Blood Center led by Louis M. Katz, MD, and Laurie Rogenski, RN, and financed largely by Small Business Innovation Research grants (HL 61111) from the National Heart Lung & Blood Institute. Study information presented below is available in greater detail on the Reference page of our Web site (http://www.talmed.com/talisman/reference/index.htm) and has been presented at forums such as the CBER Best Practices workshop this spring. 

We have found almost no donors refusing to use Talisman's QDS. Of about 400 Hoxworth Blood Center donors in the 1999 pilot test, only two refused to use the system, reportedly for lack of time. Of 20,000 donors in eight Mississippi Valley cites only one is reported by staff to have objected and that was to the question descriptive color pictures. The Hoxworth QDS utilized a customized questionnaire while the Mississippi Valley (MV) QDS employs the AABB Uniform Donor History Questionnaire. As far as we know, QDS is the only audio-CASI system to employ either touch screens or question descriptive color pictures in routine health history screening of blood donors. Comprehensive results from the first year of implementation at MV have been reduced to an abstract and submitted to AABB for presentation at next fall's annual meetings. Below is a sample of results from previous public scientific presentations. 

In addition to Hoxworth, standardized donor satisfaction surveys have been conducted at three MV sites and replicated at two of these facilities. Completed surveys total over 1000. We have discontinued surveying because results are so similar. In brief, across all surveys donors with a preference almost overwhelmingly prefer QDS to traditional FTF interviews, i.e., by a minimum of a factor of four across measures of clarity, privacy, truthfulness, time satisfaction,  understanding and likelihood of donating again. For some measures like clarity and privacy, over 90% of donors prefer QDS. (http://www.talmed.com/talisman/reference/bianco1.htm). 

With one exception, MV staff prefer QDS to manual system alternatives by a factor of three or more when measured by variables of faster for staff, faster for donor, donors more honest, answers more accurate, answers more confidential, fewer staff errors and personal satisfaction. The exception is faster for donors where staff (correctly) perceive QDS takes donors longer. In spite of the longer time, donors prefer QDS. Additionally, MV showed that QDS reduces Errors/Omissions by at least 60%. (http://www.talmed.com/talisman/reference/02182002_12.htm)

We believe the literature supports stronger guidance from CBER encouraging the use of technology that enhances understanding and honesty, eg, AV-T-CASI technology.  Two options that come to mind are implementation of AV-T-CASI as a CBE, instead of CBE-30, and a CBER/NHLBI sponsored workshop to make the industry aware of the scientific work that has been done.

We trust the research reported (above) will facilitate achievement of CBER's objectives. If we can provide help, such as electronic validation of new or proposed donor questions or technology, please let us know.


Sincerely, 


Paul D. Cumming, PhD

Talisman Ltd.

www.talismanlimited.com

Office 703-938-0300

Cell 703-283-3992