Skip Navigation
  • Text Size: A A A
  • Print
  • Email
  • Facebook
  • Tweet
  • Share

Blood Safety Transcripts

NATIONAL INSTITUTES OF HEALTH PUBLIC HEALTH DEPARTMENT
ADVISORY COMMITTEE ON
BLOOD SAFETY AND AVAILABILITY

Monday, August 11, 1997
Lister Hill Auditorium
National Library of Health
National Institutes of Health
9000 Rockville Pike
Bethesda, Maryland

Transcribed From Provided Tapes By:
CASET Associates, Ltd.
10201 Lee Highway, Suite 160
Fairfax, VA 22030
(703) 352-0091


PARTICIPANTS:

Committee:

Arthur Caplan, Ph.D., Chair
Janice K. Albrecht, Ph.D.
Larry Allen
James P. AuBuchon, M.D.
Michael P. Busch, M.D., Ph.D.
Ronald Gilcher, M.D.
Edward D. Gomperts, M.D.
Fernando Guerra, M.D.
Paul F. Haas, Ph.D.
William Hoots, M.D.
Carolyn D. Jones, J.D., M.P.H.
Dana Kuhn, Ph.D.
Tricia O'Connor
John Penner, M.D.
Jane A. Piliavin, Ph.D.
Eugene R. Schiff, M.D.
Marian Gray Secundy, Ph.D.
John Walsh
Kristine MacDonald, M.D., M.P.H., Consultant

Ex Officio Representatives:

Eric Goosby, M.D., Office of the Secretary
Paul R. McCurdy, M.D., NIH
Mary Pendergast, J.D., FDA
Mary E. Chamberland, M.D., CDC
David Snyder, R.Ph., D.D.S., HRSA
CAPT Bruce Rutherford, MSC, USN, DOD

TABLE OF CONTENTS

Call to Order - Dr. Caplan 1
Opening Remarks - Dr. Goosby 2
Open Public Hearing 4
Agency Discussion, FDA - Dr. Mied 11
Agency Discussion, CDC - Dr. Margolis 23
Cost-Benefit Analysis of Lookback - Dr. AuBuchon 45
Legal Risks of Lookback - Ms. Ungerer 77
Effect of Lookback on Hospitals - Dr. Kaplan 100
Course and Outcome of Hepatitis C - Dr. Gordon 113
Course and Outcome of Hepatitis C - Dr. Seeff 125
Lookback Experience in U.K. - Dr. Robinson 148
Lookback Experience in Canada - Dr. Gill 173
Open Committee Discussion 185

P R O C E E D I N G S

Agenda Item: Call to Order - Dr. Arthur Caplan

DR. CAPLAN: I'd like to call to order the Advisory Committee on Blood Safety and Availability. In so doing, I'm going to give a little heads up to Eric. I'm going to ask him to stand in for John Eisenberg, who is not here yet.

This is the second meeting of the committee. We are going to be looking over the next two days at the issue of lookback for hepatitis C, which we will hear much about throughout this morning's program. We are going to start today with some public testimony in an open hearing, and then move into some presentations which the committee asked for at its last meeting when it was looking for more information on cost/benefit, some of the legal issues raised, some of the impact of lookbacks on hospitals and blood banks, and some of experience of other nations that have wrestled with the lookback question.

What I would like to do is just run through a couple of brief housekeeping announcements, that are important, particularly if you are out in the audience.

[Administrative remarks.]

Why don't I let Eric say a word at this point, just about where we are at in terms of task and mission for today.

Agenda Item: Opening Remarks - Dr. Eric Goosby

DR. GOOSBY: Thanks, Art.

I really am very happy to welcome you here. Dr. Eisenberg has a conflict that he was given last evening by the secretary, and sends his apologies for not being able to be here. He will try to come in either late or with your conclusions tomorrow.

This is the second meeting of the committee, and you are focusing on a very difficult issue that has concerned the Public Health Service really for a number of years, and we really look forward to the focused discussion that you are going to embark on today and tomorrow, focusing on hepatitis C lookback, and analyzing and considering the two proposals that were presented at the end of the last meeting, worked on a little bit between meetings, and looking at cost/benefit, legal issues, et cetera around the hepatitis C lookback is going to be a very productive and interesting discussion. We look forward to the conclusions that this committee will offer the secretary.

I want to deal with a little business and announce to the committee that we have identified an executive secretariat who will also serve in some other functions within the Office of HIV/AIDS Policy in the Office of the Public Health and Science in the secretary's office. Dr. Stephen Nightingale -- Stephen, if you could just stand up for a second -- will be orchestrating, coordinating this committee and the work of it, and will serve as a focus point for committee members. We'll give out his phone numbers and fax tomorrow, which are the same as mine.

Dr. Nightingale graduated from Harvard College and the University of Pennsylvania School of Medicine. He trained in internal medicine and nephrology at Johns Hopkins, and is board certified in both areas. He has taught at Hopkins, Case Western Reserve, the University of Chicago, the University Center for Health Sciences in Yaounde, Cameroon before becoming the chief of general medicine at Cook County Hospital in 1982, and the medical director of the Parklawn Memorial Hospital AIDS Clinic in 1988.

For the past year, Dr. Nightingale has been a medical officer at the FDA Center for Devices and Radiologic Health, and attending physician at D.C. General Hospital, where he works in the AIDS clinic on a volunteer basis.

He is the author of 50 publications, has had research interests that have included opportunistic infections in HIV, and psychological influences on medical decision-making. He is married to a Dr. Elizabeth Wiley, who is associate chief of surgical pathology at Northwestern in Chicago, and they have two boys, aged 14 and 20. He lists his hobby as commuting.

So we are very happy to have someone of his stature join our team, and hopefully can serve the needs of this committee faithfully and effectively in the future.

So thanks, Art. I look forward to the deliberation.

DR. CAPLAN: Thanks, Eric.

[Administrative remarks.]

We have two people who have requested time to present to the committee, Jim MacPherson from America's Blood Centers, and then Fay Lamb(?) from the Cooley's Anemia Foundation. Is she here too?

Agenda Item: Open Public Hearing

MR. MAC PHERSON: Thank you, Dr. Caplan. Thank you to the committee.

For those of you who have a copy of the statement, I did not had a chance to see the draft until this morning of the recommendations, so I have revised my remarks accordingly to coincide with the draft of the statement.

America's Blood Centers, whose not-for-profit members, community-based, provide nearly half the nation's volunteer donor blood supply, thanks the committee for this opportunity to once again address the issue of lookback procedures for hepatitis C.

At the committee's April meeting our vice president, Dr. Sela Bianco(?) provided a comprehensive outline of the controversies and options surrounding this issue, so today my comments will be short, and hopefully to the point.

First, so-called donor lookback will not identify the majority of patients who have infected by a blood donor's potential silent infectious blood donation, nevertheless, it is difficult to argue that blood centers in possession of information about a recently acquired infectious state of a donor should not perform lookback for HCV as they already do for HIV and HTLV.

ABC only asks that a limit be placed on the time of an infected donation, and lookbacks on previous donations. Given that there are no public health issues that would compel a longer period, we would support the one year time limit recommended in the Drs. Hoots-Penner draft as a reasonable balance between a patient's right to know about the infectious status of his or her blood donation, and the practicability of a hospital to search its computerized patient records.

In addition -- and this comment may not be effective any more if the draft is accepted by the committee, but of course that is the point of today's and tomorrow's discussion -- any such recommendation should only be applied to this day forward; that is, the donor lookback. To do so otherwise could stymie the current operations of blood centers and hospital transfusion services for months in exchange for very little public health gain, especially given the high degree of false positive test results from the first generation of HCV screening tests.

If the federal government feels compelled now to require retrospective donor lookback for HCV after years of rejecting that option, the government needs to provide funding to support the hundreds of thousands of work hours that would be required for blood centers and hospitals to effectively perform this daunting activity.

Second, regarding past transfusion recipients -- and this is consistent with the Penner-Hoots draft -- as it has for many years, ABC recommends two options to inform patients about their risk and choices, about determining whether they need to be evaluated for a transfusion transmitted disease.

The first option is for the Public Health Service to send a letter to all practicing physicians, one of its "Dear Doctor" letters, recommending that a history of blood transfusions be incorporated into routine medical health histories. PHS should include guidelines as to when a doctor and patient might need to consider testing for HCV, HIV, et cetera.

For example, when a transfusion took place before 1990, for HCV for example, although I notice the draft is not even recommending that. Ultimately, such health decisions should remain between the patient and his or her physician.

The second option is for the PHS to join with organizations such as the American Liver Foundation to develop and disseminate public service announcements making patients aware of past transfusion risks, that can stimulate a meaningful dialogue between the patient and the physician.

Finally, we remain proud that over the last seven years we have been a leading voice in the blood community to advocate for some reasonable approaches to informing blood recipients about their past risk -- sometimes a lonely voice -- even the face of public health uncertainty. We continue to offer the committee any help or insights we can on this, and other issues under consideration.

Thank you.

DR. CAPLAN: Questions? We should open the floor.

All right, let me ask Ms. Lynn to come down from the Cooley's Anemia Foundation.

MS. LAMB: Mr. Chairman and members of the advisory committee, my name is Fay Lamb and I am a thalassemia patient. Today I represent the Cooley's Anemia Foundation and its patient support network, Thalassemia Action Group. For four years the Cooley's Anemia Foundation has funded thalassemia research, patient care, public and physician education, and has worked to assure a safe blood supply.

Now some of you probably know to remain alive, thalassemia patients receive an average of two units of red blood cells every two weeks. Thalassemia patients are the single largest group of chronically transfused patients. We deserve a blood supply that is free from deadly viruses and contaminants.

In the brief time that I have available this morning, I intend to address the important issue related to hepatitis C lookback, particularly as it relates to patient notification that is on your agenda today. It is critically important, however, that you understand how the world of blood supply and blood safety looks to the people who have to rely on the system to live.

Every time we go for a blood transfusion we know we place our lives in the hands of large numbers of people we will never meet: regulators from the FDA, researchers at NIH, lab techs working for private for-profit companies. The lives of thousands of persons receiving blood transfusions every day rests on the faith that everyone is doing his or her job. You have an awesome responsibility, and to the extent that America has the safest blood supply in the world, I am here to thank you.

I am also here to challenge you. In the midst of criminal prosecutions of persons in the largest blood center in our nation, think about the child, teenager or young adult who has to receive blood regularly. They read a newspaper or see a TV news report about unsafe blood, or about lab techs receiving incentive payments to approve more blood on a shift. They know people who have contracted AIDS or hepatitis. They are frightened, and so are their families.

This week, employees of the New York Blood Center are on trial in federal court for breaching their duty and obligation to properly test blood. At the same, this week thousands of thalassemia patients in New York are being tested for HIV. This action, as a result of the recall stemming from the New York Blood Center.

It is within this context that Cooley's Anemia Foundation views the hepatitis C lookback issue. Our patients deal with so much, it would be tempting to say that they are better off not knowing. It would in some ways, seem easier not to make the effort to identify individual blood recipients and notify them of HCV contamination.

One might even argue that on a cost/benefit basis, it not worth going through all this effort, but would that be right? I do not think so.

In spite of all the difficulties involved, it is essential that this committee support the establishment of a system that will result in the notification of patients, not just suppliers, that there is a reasonable probability that they have received tainted blood.

In addition, there should be a system comparable to that which occurs in genetic counseling in which the issues involved are fully explained to the patients. They need to have the opportunity to know what their exposure is, what the risks are, and what steps, if any, they can take to address them.

This system, which would be a kind of ex post facto informed consent, would be of tremendous benefit to patients and their families. It would treat them fairly and provide them with the opportunity to make individual judgments about their medical status. I would suggest that it should be paid for by an assessment against for-profit blood centers.

I realize that much of what I have addressed is beyond the scope of this advisory committee, but I do not believe that you can adequately address the singular issue that you are currently studying without taking into consideration the larger universe within which patients operate.

I thank you for the opportunity to present the views of the Cooley's Anemia Foundation.

DR. CAPLAN: Questions? Thank you for those statements.

I think what we will do is move, unless there are others who might want to address the committee in the open session? All right, then what I would like to do is move to some input from the agencies.

Is Dr. Mied here? I didn't see him before. Oh, he's hiding out there. Why don't you come forward, and we'll maybe do these next two, and the move to a break after that. I know there are some caffeine-dependent souls, but I think we can do these two.

Dr. Mied, from the FDA.

Agenda Item: Agency Discussion, FDA - Dr. Paul Mied

DR. MIED: Thank you, Dr. Caplan.

This morning I am going to summarize the FDA policies related to lookback for transfusion transmitted infectious agents. Then present some issues for consideration in addressing the potential utility and optimal design of a program of transfusion recipient notification for HCV.

The issue at hand is whether and how to focus a program aimed at the identification, notification, testing and counseling of persons who may have been infected with HCV through transfusion. Assuming that such an effort is appropriate, what would be the most efficient way of identifying the largest number of HCV infected transfusion recipients?

Lookback refers to two procedures: product retrieval, the tracing and retrieval of previous viral marker negative collections from a donor who now tests repeatedly reactive; and recipient notification, the tracing, notification and counseling of recipients of these prior negative units when the donor now tests positive.

There are two reasons for performing lookback. When a donor now tests positive on a viral marker test, product retrieval can be performed to interdict earlier potentially infectious viral marker negative donations, which may have been made when the donor was in the infectious window period.

As you know, the window period is that time interval between infection with the virus, and the appearance in the blood of a viral marker such as an antigen or antibody which is detectable in a serologic test for that marker.

Secondly, to notify transfusion recipients so that they may be tested and counseled regarding the risk of disease, the availability of treatment and prevention of secondary transmission. The current lookback policy for HIV consists of both lookback procedures.

Recommendations for product retrieval. When a donor now tests repeatedly reactive, and for recipient notification related to prior collections from a donor currently testing positive for antibody to HIV were put in place in April 1992, in an FDA recommendation to blood establishments. FDA also published a final rule in September 1996, to require certain lookback procedures for HIV.

The HIV lookback rule states that if a donor tests repeatedly reactive on a screening test for antibody to HIV, the blood establishment should promptly quarantine previous collections intended for transfusion from that donor from the past five years, or back to the time 12 months before the last negative antibody screening test. If the data of the last negative test is known, units collected more than 12 months prior to this date need not be quarantined. If the data is not known, all units collected previously should be quarantined.

Blood establishments should also quarantine previous collections of plasma from the past six months if intended for manufacture into injectable products. Now this refers to unpooled units of source plasma or recovered plasma, and this time period is limited to six months, because that represents the practical limit of time that plasma for further manufacturing generally remains in inventory prior to pooling.

Now in addition to quarantining prior collections, the blood establishment should also notify consignees such as hospital transfusion services and manufacturers of plasma derivatives of the repeatedly reactive HIV screening test result so they can promptly quarantine prior collections intended for transfusion, or unpooled units of plasma from that donor until the donor status is clarified through further testing.

The blood establish is also required to promptly test the current donation using a licensed, additional, more specific test for anti-HIV-1, and to notify consignees of the result. Consignees should be notified of the results so that they can release those quarantined units if the test is negative, or destroy those units if the result is positive or indeterminant. Notification of consignees regarding the results will also enable those consignees to carry out notification of transfusion recipients if the supplement test is positive.

These transfusion recipients shall receive notification for the purpose of testing for evidence of HIV infection, early treatment if indicated, and counseling to take appropriate precautions to prevent the further spread of the virus such as to sexual partners. The transfusion service shall notify the recipient's attending physician, and ask him or her to inform the recipient of the need for HIV testing and counseling. If the physician is unavailable or declines to notify the recipient, the transfusion service shall do so.

So that is a summary of FDA's policy for product retrieval and recipient notification for HIV.

Now the usefulness of recommending similar actions for hepatitis B surface antigen, anti-hepatitis B core, anti-HCV and anti-HTLV-1 has been discussed at several public meetings of the FDA Blood Products Advisory Committee over the last five years. For each of these viral markers, it has been necessary to examine the probability of units from prior collections being infectious based on the significance of the subsequent test information.

Now although FDA believes that currently there is a very low risk of transfusion transmitted HBV, because of the sensitivity of screening tests for HBV infections, and although there is a continued absence of licensed supplemental tests for anti-hepatitis B core and anti-HTLV-1, FDA has determined that quarantine of in-date prior collections from sero converting donors will provide an added safeguard to the blood supply.

As a result, in July 1996, FDA issued a memorandum to blood and plasma establishments which recommended product retrieval for HBV, HCV and HTLV-1. The FDA recommendations concerned the quarantine and disposition of prior collections extending back five years from a donor who now tests repeatedly reactive. If there is a record available of the donor's last negative test result, then quarantine of prior collections need only extend back to 12 months before such a test.

Blood establishments should also notify consignees of such products, and request the quarantine of those products. This recommendation is directed at in-date, unpooled units that are in the blood establishments' and consignees' inventories, not products which have already been pooled or further processed.

FDA also recommended additional testing of the donor's current repeatedly reactive sample. For units previously distributed, consignees should be notified of the results of the additional testing so that they may either release or destroy those quarantined products as appropriate.

However, in regard to products already transfused, FDA did not recommend consignee notification for the purpose of recipient notification for HBV, HCV or HTLV-1 in this memorandum.

In September 1991, the Blood Products Advisory Committee recommended that additional information should be obtained before recipient notification could be considered for HBV.

Additionally, the committee decided that the data available at that time did not justify recommending recipient notification for HTLV-1, and advised FDA to consider such recommendation for HTLV-1 when additional, more specific licensed tests become available. As you know, there are still no licensed supplemental tests for anti-HTLV-1.

I should point out that when screening for HTLV-1 antibody was implemented in 1989, blood establishments voluntarily initiated notification and counseling of recipients of blood or cellular components from donors who have a confirmed, positive test for HTLV-1, -2 antibodies. Since the number of infected donors identified were small, a sequential, targeted lookback program was recommended by the American Association of Blood Banks, and implemented without much discussion or controversy.

In summary then, lookback policies currently in place for these transfusion transmitted agents include both product retrieval and recipient notification for HIV, but only product retrieval at the present time for HBV, HCV and HTLV. The question of whether recipient notification should be recommended for these agents awaits further public discussion of the legal, social and economic issues involved.

Today, the Public Health Service brings before this committee the issue of the usefulness of recipient notification for HCV in the context of public health. The screening of blood donors for antibody to HCV was implemented in the United States in May 1990, when the FDA licensed the first enzyme immunoassay for the detection of anti-HCV.

Reflecting earlier discussions of the Blood Products Advisory Committee, an NMWR issued in April 1991, did not recommend at that time, product retrieval for HCV, or targeted lookback, or general screening programs for populations at increased risk of HCV infection for the following reasons.

First of all, the lack of availability of a licensed, additional, more specific test for anti-HCV could lead to a high rate of following up of false positives on the screening test. As you know, a specific supplemental test for anti-HCV is now available.

Secondly, the anti-HCV screening tests were not able to distinguish between ongoing infection and recovery, thus the meaning of a repeated reactive result for any one individual would not be clear. It is now clear that most anti-HCV infected individuals have a chronic transmissible disease.

Also at that time, 1991, many HCV infected donors would have been deterred as a consequence of earlier screening policies. For example, donor exclusion due to risk factors of HIV, and implementation of testing for surrogate markers for non-A, non-B hepatitis; that is, anti-hepatitis B core, and ALT.

This deferral would make redonation and therefore detection of HCV infection less likely, and this is still true today. The available knowledge on routes of transmission for HCV other than parenteral was limited at that time, reducing the usefulness of medical counseling, and this is still the case today.

Because of labeling limitations of approved indications for alpha interferon therapy, only selected persons were eligible for treatment for HCV infection, and potential long-term benefits of such therapy were not known. This is also still true today. Labeling limitations still exist, and the long-term benefits of such therapy are still not fully understood, although we know that some patients do benefit from this therapy.

In March 1992, the FDA approved a multi-antigen screening test which had increased sensitivity for antibodies to additional HCV antigens. In June 1993, FDA licensed an anti-HCV supplemental test, the REBA-2(?), and the FDA Blood Products Advisory Committee unanimously endorsed product retrieval from previous collections from donors who test repeatedly reactive for anti-HCV, and who do not have a negative result on a licensed supplemental test.

By a vote of five to four the committee marginally endorsed consignee notification for purposes of recipient notification, but reiterated many of the reservations described earlier related to the lack of an established public health benefit in performing this activity.

Now in deciding whether to recommend recipient notification for a particular transfusion transmissible infectious agent it may be helpful to address an antecedent set of issues. These broadly applicable questions merit consideration in addressing the potential utility and the optimal design of recipient notification such as for HCV.

First of all, is the infection a significant medical condition? We know that HCV infection is certainly a very serious condition.

What is the incidence and prevalence of the disease in the overall population? Is infection at an epidemic level? If so, is it a new epidemic, or is it an endemic condition?

What is the risk of transmission of the infection by transfusion? What percent of the overall disease frequency in the population is caused by transfusion? More specifically, is transmission by blood transfusion a large enough part of the total community transmission to merit special public health measures? What is the risk of disease following transmission of the infection?

What is the accuracy or specificity of the tests for markers of the infection? Is there a supplemental or additional test of greater specificity than the initial screening test that is commercially available? Do the tests distinguish between a chronic carrier state and being in a state of recovery from the infection?

Are there opportunities for intervention sufficient to warrant recipient notification? Are interventions available? Is the disease treatable? If so, is early treatment more effective than later treatment?

Can avoidance of co-factors such as alcohol be recommended to the recipient?

Is there significant risk of secondary transmission and disease? What are the roots, if any, of secondary transmission of the infection? Is it transmitted parenterally? Is it transmitted sexually? By household contact such as sharing toothbrushes and eating utensils? By other types of contact? What is the efficiency of transmission by these different routes?

Is there opportunity for prevention of secondary infection? Can intervention in lifestyle or activities prevent secondary transmission?

What ethical considerations are involved in a decision to notify or to not notify recipients? Should recipient notification, if recommended, be prospective, that is, subsequent to promulgation of a policy? Or should it also be retrospective? For example, should the recipient notification program be intended to capture all prior collection before screening was implemented, or to capture sero converters after screening was started?

The cost effectiveness of recipient notification needs to be examined. Who should bear the cost, the patient, the blood center, the government? A discussion of these factors may suggest whether recipient notification for HCV is justified.

Thank you.

DR. CAPLAN: Questions?

DR. PENNER: Do you have information on the incidence and population? You referred to some figures [inaudible].

DR. MIED: I have one figure. That is, that the U.S. overall prevalence is about 2 percent. It is much higher in certain areas for sure.

DR. PENNER: Would you agree that while all of the information on positive tests from 1990-91, that information would be available as a result of the Freedom of Information Act if someone wanted to pursue that?

DR. MIED: I'm not sure I can answer that question.

DR. CAPLAN: Other questions, comments?

I wonder if I could ask, on your slide that you had up about the BPAC decision not to do consignee notification, that was prior to some of the evidence about alcohol use and so on wasn't it? In other words, the situation has shifted somewhat about recommendations on avoiding alcohol?

DR. MIED: That was back to 1991.

DR. CAPLAN: Yes.

DR. MIED: I don't think some more information has come to light since then.

DR. CAPLAN: All right, if there are no more questions, thank you.

Why don't I ask Dr. Margolis to come down from CDC, who is also going to give us an agency perspective on lookback. We'll do the break after we have had a chance to hear from him.

Agenda Item: Agency Discussion, CDC - Dr. Hal Margolis

DR. MARGOLIS: Thank you.

What I'm going to do this morning is to give you kind of an update, since actually at the last meeting I presented lots of data, and I think various segments of that are going to be readdressed in the presentations today.

To kind of review a bit as I did at the last meeting, it was this statement from the report the Government Committee on Reform and Oversight that really has prompted this agency look at what we are going to do in terms of the issue of persons who are previously infected by HCV that may have been transmitted by blood or blood products.

This, as you recall, were what at least we were framing as two questions that we felt needed to be addressed in terms of real strategies that the department should take to identify those persons who may have acquired their infection from transfusion, to determine the infection status, provide appropriate counseling in terms of medical management, and then ultimately make recommendations in terms of preventing future infections or transmission. Then how should this effort be defined in a public health perspective?

What I would like to do is really maybe frame a few -- because we have pointed out the objectives were the early identification of those persons with chronic infection, and ultimately to reduce transmission. I think if you put this in the context of a public health prevention strategy, there are other pieces and I think it is important for the committee to realize that at least we at CDC and in other agencies within HHS are looking at the whole issue of HCV prevention, which includes both acute disease surveillance to prevent and assess what is going on with prevention of new infections.

This is the issue which we are spending most of our time discussing here, and which we feel is probably one of the most important issues, is to educate the medical, drug prevention and general community about HCV infection. In terms of just some update information -- no hard data, no slides -- but since we met the last time, we have had several focus groups with physicians in terms of trying to -- and this is primary care physicians -- get to the issue of what is it that they either know about HCV infection.

What would they like to see from an agency such as us, and from those of us that are trying to educate? It turns out that quite often a patient going to a primary care physician, the physician isn't very knowledgeable about diagnostic issues, about medical management issues, and as you will see as I finish with this presentation, those are some of the first things we feel are most important, is that we have an informed medical community.

As I pointed out in the presentation last time, it doesn't do us a lot of good to either generally educate the public to go see their physician if in fact their physicians don't know what to do.

Other parts of the prevention strategy include implementing programs to prevent infection among drug users, and to ultimately identify persons with chronic infection to prevent transmission.

When one looks at the options for identifying individuals at risk of infection, as was pointed out earlier, one approach that has been suggested is that of targeted lookback of transfusion recipients. I'm not going to represent at least our model of what proportion of the estimated 4 million infected individuals in the country we feel are due to transfusion, but actually no matter how far you push those figures, it is actually a very small portion of the total HCV infected individuals.

The other approaches are that of educating the public in general, and then as I said, educating health care providers.

Probably to put this in a slide which I have now updated from the last time was we are taking the approach that if one looks at those individuals who may have been infected by transfusion prior to the initiation of testing in 1990, that it is really public and provider education that is going to identify and appropriately handle those individuals who may have been infected via that route of transmission.

Again, it become physician and provider education for having a high index of suspicious for that very small number of individuals who have been infected during this period between 1990 and the present. Then I think as you heard, the sense is that for new sero conversions within the donor population, the recipient notification is probably the approach to take as we move forward, and has been proposed in the draft by the committee.

The public issues are ones that quite frankly we don't have necessarily the best approaches in terms of how best to do this. There has been a lot of effort that has been handled by voluntary agencies such as the American Liver Foundation, who have been targeting a number of public information activities.

I think also other voluntary health agencies that deal with patient populations that may have been at risk for specifically transfusion transmitted HCV infection have also been carrying on very intensive and probably most adequate educational campaigns.

Quite frankly, there are no data in terms of surveys within the communities, though I think at least from a conference held several years ago on the risk of chronic liver disease among the hemophiliacs, and sponsored by the American Hemophilia Foundation, at least there was some survey data that suggested that essentially all hemophiliacs under medical care had been in fact screened, and their current HCV status adequately determined.

So again, in the very high risk populations, and if one wants to look at that as kind of a combined medical and public information issue, we feel that that has been handled, but again, there are really no good survey data.

In terms of the remainder of the general population, quite frankly at this point we feel there need to be surveys, but really there hasn't been anything in terms of determining what proportion of at risk individuals have been identified.

Then again as I pointed out earlier, the need for educating health care professionals.

[Tape change, text lost.]

. . . I'll put a slide up here in a minute.

I also have some handouts. There are also I guess, extra copies out in the front. Let me also give you these here.

What I'm passing out is the first document really takes all the pieces of what I just put up in slides, and tries to put some operational activities around it. This is a working document that has been used at an Agency Heads' Blood Safety Committee meeting just to review where the various agencies are going in terms of these educational activities.

I think at your leisure you can look through these, but they basically just summarize some of the things that I have mentioned, and include things such as the NIH consensus conference, which was a major issue in terms of providing a framework for physicians in terms of guidelines as to where we are going, and how diagnosis, treatment and management ought to be effective.

Another thing that we are doing, at least from the CDC perspective is that -- I mentioned this again in the presentation at the first meeting -- on November 22nd there is going to be live, interactive satellite teleconference entitled, "HCV Diagnosis, Clinical Management and Prevention." In a two hour period, with case histories and panel discussions and some of the experts being a part of this, and again one of those handouts was an update on that, we are trying to provide primary care physicians with take home messages which include both what they are going to get on the TV screen, as well as what they are going to get in physical handouts.

Another thing that is going on besides what CDC -- this is being done by CDC and the Hepatitis Foundation International, which is another liver disease voluntary health agency -- is the American Digestive Health Foundation's Hepatitis-Digestive Health Initiative, which is again trying to provide take home message materials. In other words, the easy to use guide to diagnosis, treatment, medical management and prevention information that is also going to be targeted to primary care providers.

So I guess what I want to do is just really let the committee know that there are a lot of things going on out there. In fact, I just came from a meeting yesterday with the American Correctional Association, and talking with all of the commissioners of corrections from all states in the U.S.

Let me tell you that if one looks in prisons, anywhere from probably 40-70 percent of prison populations are infected with HCV; that is, at least, small surveys that have been done. That really tracks with drug use, and this is a major issue now within the correctional setting. Some of those have acquired infections by transfusion, again prior to 1990. What to do with that population is something that is currently facing these people in terms of a public health perspective.

We have also recently had a meeting with the Council of State and Territorial Epidemiologists, a small working meeting in terms of again, these public health issues and how to get the message out, and how to deal with such things as alternative test sites for HCV infection.

Nobody here in the blood banking community would like to see people coming in to get HCV testing because somebody comes in as a blood donor. We would much prefer that there be in fact a place for testing to be done, for those who can't afford testing because their current medical coverage may not pay for that.

So these are some of the issues that are in fact now beginning to come to the forefront, issues that are no different when we saw with HIV, and which I think this committee needs to be aware of as we move forward with some of these prevention efforts.

At that point I'll stop I guess, and answer questions.

The other thing I have, and which actually I will do for everybody on this committee is that you will get a little packet, so that you can publicize the teleconference in your institution. As I pointed out, there are about 600 institutions that have signed up, and we definitely want to make sure that physicians and providers at all levels take advantage of this. It's free.

The production is free. The actual materials with it, there is a charge, because there is going to be a syllabus and materials that again, people can take home, which will have many of the messages that we have been discussing here.

DR. CAPLAN: Bill?

DR. HOOTS: Could you just refresh our memories just in terms of the prevalence in non-high risk groups like just a person off the street, male or female?

DR. MARGOLIS: Well, the data -- and I didn't go back and put that up again -- but the data from the National Health and Nutrition Examination Survey, which is a statistical sample of the non-institutionalized population in the U.S., the overall prevalence is 1.8 percent. That translates to about 3.8 million individuals.

Unfortunately, in that data set, and when that was done, and the midpoint of that is 1990, the question was not asked as to whether you were an injection drug user. So really those kind of risk factors that you would like to have, are not available from that, but there are very significant differences by race and ethnicity, with the highest prevalences being in African Americans.

Again, these all tend to be younger individuals, with kind of a mid age of around 40. So unfortunately they are also predicting that there is potentially an epidemic of chronic liver disease coming down the line.

I think I showed the last time also if one looks in injection drug users, at least data from inner cities, prevalences of 70 plus to almost 100 percent in almost any population that has been looked at in the U.S., has been found. So again, so the average person I think you can use the approximately 2 percent prevalence.

PARTICIPANT: Dr. Margolis, I believe that there are some CDC sentinel surveillance studies being conducted by the CDC, along with I guess approximately 109 hemophilia treatment centers, looking into the prevalence of hepatitis in the population. Do you have any data on the morbidity and mortality statistics there?

DR. MARGOLIS: Again -- and I didn't realize I needed to go back and review -- I had shown that the sentinel surveillance or the sentinel county of viral hepatitis actually gives us our really only precise measure of incidence, of new cases. In fact, the incidence of hepatitis C has dropped by almost 90 percent since around 1990, most of that really attributable to apparent decreases in incidence among injection drug users.

So that gives us new cases. New cases are down dramatically. Again, some of the issues, as I put on that slide, and one of the first things is actually the issue of surveillance, our sentinel counties study, of which there are four -- we now have another two counties we are just adding -- actually because there are so few cases, we are having a hard time accurately measuring new cases. So that really means that every state in the United States needs to start looking at hepatitis C. That is something that we are actively working with state epidemiologists on.

The second question had to do with the hemophilia treatment center surveillance, and actually I don't think there is any summarized data at this point from that study. That has really just, at least organizationally, gotten off the ground here within the last year, and at least I'm not aware of any data on incident cases of hepatitis C.

PARTICIPANT: The reason I asked that is I know I was part and parcel of filling out the minimal data set. It was requesting hepatitis C cases, you know, having done studies for it. I think this is its second year. I was just wondering if you had any kind of preliminary data on that information, because again, it would ask us what are the cases in which we have found that persons were deceased due to liver problems.

DR. MARGOLIS: I'm not aware of, as I say, data on incidence. I do know that there is a lot of data on prevalent chronic infection in that population.

PARTICIPANT: I know that Dana is talking about actually two different things. One is a data set which is kind of retrospective and based on available data, and then there is a prospective study, which I guess Dr. Margolis is speaking of, in which the CDC will actually collect the data on the hemophilia population prospectively. Pilots are beginning this month for that.

In terms of the retrospective in hemophilia, the problem, as you well know, comes from the complication of multivariate risks, particular HIV plus HCV in terms of the impact on death. I think the strongest data comes from Jim Getter(?) and Elaine Isters(?) data that suggest that the mortality rate in combination is very high in a relatively shorter period of time than one would expect just with HCV alone.

I don't think the natural history follow-up has been long enough to get a good numerator over denominator.

DR. CAPLAN: Dr. Guerra?

DR. GUERRA: One of the very short-term observations that we have made in San Antonio is I think we an increasing use of the antibody testing to better define the cause of liver disease, especially chronic liver disease is that in 1996, and this was just something we looked at very quickly without doing any specific analysis, there were 66 deaths we could directly attribute to hepatitis C that was coded out on the death certificate, which was quite interesting for us to have access to that information.

I wonder if there is some observation that is taking place nationally from death certificate data, where they are being signed out as hepatitis C?

DR. MARGOLIS: We have made estimates again, as I had summarized at the first meeting, that there are about 8,000-12,000 HCV attributable liver disease deaths that occur annually. That is based on applying prevalence data from a single community study of all chronic liver disease, to a national death certificate, multiple cause mortality data.

We are, I hope, by the end of the next month here in this fiscal year funding our first chronic liver disease surveillance site, with the anticipation of at least having four, maybe as many as six of these sites, to be able to get national chronic liver disease surveillance, which will include obviously HCV infection.

There is no such surveillance system at this point. We have been struggling for several years to get funding up, and we will fund the first one, and plan to fund more. So I think we can then have national data, which will also include information about potential source of infection, realizing that you are going back pretty far retrospectively. Again, I think it is going to be data forthcoming. We don't have anything ongoing at this point.

DR. CAPLAN: Dr. Penner?

DR. PENNER: [Inaudible].

DR. MARGOLIS: We're hoping in the 10,000 range minimum. I think with 600 sites, even if you have a few at each site, we are already fairly well on our way. That is our goal. Again, the materials that are being developed by the Digestive Health Initiative are looking at probably a similar number. Some is going to be overlap. We realize you can't get all probably 40,000-60,000 primary care physicians that are at least in some of the various professional societies.

PARTICIPANT: [Inaudible].

DR. MARGOLIS: Well taken, and as pointed out, when you have time, probably when you get back home and look through that kind of implementation guide, the reality is that I think everybody at CDC, NIH, FDA have been on the talk circuit. That is from the agencies, and there are plenty of other people in this audience who have been attending and doing many of those meetings.

It is obviously a multi-pronged approach. We felt that at least in this day and age, the input we are getting is that the teleconference will capture -- and with CME credit; this does have CME credit with it -- will capture one segment of the primary care population so to speak. We have been working with many other groups, and I think everybody in this room who is at least dealing with chronic liver disease or the blood bank community has been giving those talks also. So it is not just single focused, just to assure you that we are trying to look at those.

DR. CAPLAN: Mary Jane?

PARTICIPANT: Hal, I think you are probably aware one of the proposals that was put forward by the committee at the conclusion of its last meeting, and which I think will be actively discussed in the next couple of days was to include as part of the package of notification would be some element of general education to physicians, providers and the public.

I guess I wanted to ask you a sort of follow-up to Dr. Penner's thoughts that if in fact the committee were ultimately as part of its final recommendation to include some sort of general education as part of the recipient notification lookback strategy, do you feel that that would fit with what you are doing now? Would any changes have to be made, or further embellishment to the PHS education program that is ongoing, that you are midst of right now to accomplish that?

For example, your messages I'm sure include transfusion before testing and et cetera. Could you just kind of give that sense of --

DR. MARGOLIS: I think all of the messages that we are discussing here in terms of risk factors are part of the messages we are trying to get out. We have not gotten past that bounds so to speak.

I think the intensity of it, and we have been very honest in that in that update flow sheet I put down there, there are some resource issues. All of this has been done at this point, and again, let me just point out a few things that are happening right now. There is a cooperative agreement that is going to be awarded by the end of this fiscal year, hopefully with at least two of the liver disease voluntary agencies to assist in both trying to get current information to the public and to health care providers, as well as to develop new messages in areas where there are gaps.

We know there are gaps out there. So getting back to your comments, Mary, I think in terms of just the magnitude of this, we need help, and we are trying to put some money out there in that respect. CDC just initiated a 1-800 -- well, we're out of 1-800 numbers -- it's now a 1-88 numbers which are your toll free numbers for chronic liver disease information. That just started, and again, that is listed in that summary. That just started this month.

As I say, there is a lot of information that needs to be out there for the public in various languages, in various reading levels. I can assure you that has not been encompassed with the same breadth that we have for let's say hepatitis B vaccination, which we're a lot farther along with. We and others, again, working with a lot of people outside of the government, or trying to get these as quickly as possible and as accurately as possible.

One of the other important areas is the broader prevention issue, and we view that somewhere in 1998, we, at least CDC, probably leading, but with all the PHS agencies, need to really develop a conference on prevention issues for HCV infection which bring in some of these other issues that I brought up: getting this message into the substance abuse, the drug use prevention community out there, the drug treatment community, the public health community in general.

Frankly, nobody has been dealing with those at this point. They are getting a barrage now, and they realize we need some approach to this. I think recommendations that come out of this committee will help to at least set part of that in context.

I don't know if that answered your question. I think the major risk factors are in there. Most of the messages are there, and yes, we need to get different iterations of the message out. I think that is the bottom line.

DR. CAPLAN: Hal, do you have any success in shifting knowledge about hepatitis C in preparing the groundwork if people were to come with questions based on a lookback, and dealing with the managed care industry at the level above the individual doctor. This is on my mind, because I was just out in Minnesota, and saw them trying very hard to emphasize health maintenance in some of the organizations that are out there.

So I'm just curious, what is the focus beyond reaching to that individual practitioner, since so much of our health care is now getting delivered at a higher organizational level?

DR. MARGOLIS: We haven't opened those discussions or those forums at this point. It is something on a long list, and again, I can give you an example with hepatitis B, where again we are already 10 years down the line with some good prevention efforts. There has been in fact, a kind of re-opening of that discussion; but, no, at least we haven't. There may be others in other agencies that have, and it is obviously something that has to happen.

PARTICIPANT: [Inaudible] on the lookback situation and counseling that individual, because physicians for one, won't. We know that there is the possibility of cirrhosis. There is also the possibility of the infection being trivial and of posing no impact on that individual's lifestyle and how that individual is going to live and ultimately die.

What is not clear is what are the risks. When a physician presents this information to an individual patient, he or she will want to know how do I counsel this individual?

DR. MARGOLIS: I'm going to defer that question as Dr. Seeff will present some of that data later. In fact, Dr. Seeff is going to one of the participants in the teleconference. I am going to let him deal with that in a later presentation, because it is very important.

Our view is -- let me just put the overview on it -- is that the primary care physician has to know the front end risks. They also have to know what the hepatologist, who they may be ultimately referring a patient to is going to say in order to again, support those issues.

Yes, most people are not going to die from HCV infection, and we have heard this again in some of these small focus groups, is this is what the primary care physician wants to hear in terms of how to counsel. I'll let Leonard deal with that later.

DR. CAPLAN: One last question just on outreach to different communities. This is from my sample since our last meeting of asking some people with hepatitis C that knew what to ask their doctors what they should do. None of them could answer that question, so there is room. In the three cases that I checked out in my ample survey design, none of the doctors knew quite how to respond.

It did occur to me that these people said, well, I'll ask my dentist, or I might see a nurse and I could ask there too. What are we doing again, in terms of not just for content, but where people are likely to get information to reach out to those audiences? Any systematic efforts that you are aware that go beyond the primary physician audience?

DR. MARGOLIS: Again, in terms of targeting we are putting a lot of our resources, personnel, time and efforts at least into the front end of this teleconference, which also helps us to get a message and materials that we can use in a number of venues. We are sending information to try and get nursing personnel as they would see appropriate.

We haven't targeted dentists, but we have gone fairly wide in terms of providers outside the traditional medical or physician community, so that's a start. Obviously there are other groups that need to be targeted.

DR. CAPLAN: All right, if there are no more questions? One out in the audience here.

PARTICIPANT: I just wanted to clarify a number on prevalence. The 3.9 million estimate is based on the NHANES data, correct? That is of the civilian, non-institutional populations. It occurred to me when you spoke about going to the corrections community this weekend, that you have to add another 500,000 to that, since there are a million institutionalized people. Then that doesn't include any infection that there might in the military population. So that's really a very conservative number, correct?

DR. MARGOLIS: Correct. That is how we state it, that it is conservative. So it's really probably a minimum.

PARTICIPANT: Are you going to revise it so that we would have an idea when we think about the numbers, that really reflects the entire country?

DR. MARGOLIS: Well, we can add those on as data become available. I think that right now the NHANES is available. I think that gives us our best estimate to grapple with, which points out that it is epidemic. It is a very large number; so some of the criteria that were put up here early on. Yes, we will try and add the others in as best estimates become available.

DR. CAPLAN: All right, thank you.

Why don't we take our break at this point. So be back here at say 11:15 a.m., and we will resume then.

[Brief recess.]

DR. CAPLAN: I would like to get underway. We've got a number of other presentations. I wanted to thank Paul McCurdy for helping me track down a number of people both who you've heard from already, who we're going to hear from for the rest of the day, really trying to respond to some of the issues and requests for more information that came up at the last meeting.

So we tried pretty hard to look for expertise, both here and as you will see later on this afternoon, from outside the United States, to get more information on some of the themes and questions that you all had asked to hear more about. One of the places we looked was down the aisle and saw that Jim AuBuchon to actually talk about some of the lookback issues. He has been studying some of these things, and so I'm going to ask him to lead us off in this session.

Agenda Item: Cost-Benefit Analysis of Lookback - Dr. James AuBuchon

DR. AU BUCHON: Thank you very much.

With the help of my colleagues as shown here, I would like to present some data this morning on the cost effectiveness of HCV lookback. In this presentation I am focusing on the type of lookback that involves recipient notifications. We are not talking about quarantining of units. We're talking specifically about recipient notification.

First to share with you my biases, I do speak before you as a blood banker. There clearly are certain roles which have come to be assumed by blood bankers over the last several decades beyond just provision of a safe and efficacious blood supply and management of donors and provision of consultation, but also the concept of notifying recipients of potential defects of those units that have been provided to them.

The purpose of lookback, as you have hear discussed very eloquently this morning, is to inform recipients who have potentially been exposed, before they discover for themselves that they have been exposed. The concept would then allow for earlier therapeutic intervention, possible interdiction of secondary transmission to other individuals, and also information to be provided back to the blood supply complex in order that we could better address the issue of lookback in other recipients who may have exposed from that same donor.

Now there are of course many problems with lookback inherent in the system. It can be difficult to determine who got a particular unit of blood. As time goes on, more hospitals use a computerized recordkeeping system, this is less of problem, but it remains a problem, particularly as you go back in records beyond several years.

The location and contact of the recipients of a particular unit of blood can be very difficult to ascertain, particularly in a mobile society such as the American one. It is important to have good clinical physician cooperation in getting the information to the recipient, and that can also be difficult as physicians tend to move around.

Follow-up care needs to be provided for the recipients, so that the promise of therapeutic intervention that may be beneficial can be fulfilled. If there is potential secondary exposure the parties involved would also need to be counseled, and that always raises an issue of what these people should be told, and just how much is legitimate to be shared with them.

It is very difficult to get information back to the blood collecting agency after the lookback has been completed, because the information has to flow usually through several clinical care physicians back to a hospital transfusion medicine specialist, and from there to a blood center before any additional follow-up can be done. So it is a tortuous pathway for data, to say the least.

With HCV lookback there are some specific issues that cause more problems. Again as we have discussed some of these already, a high proportion of HCV cases not related to transfusion; not related to any known exposure route for that matter. The number of transfusion associated cases is declining even beyond that which can be explained by testing. Secondary sexual transmission of HCV is infrequent, so that is less of a reason to consider HCV lookback and notification.

There clearly is a high rate of chronic infection with HCV, but apparently a relatively low mortality associated with that. We know that there is ineffective treatment for late stages of HCV today, but we really don't know if there is any better result of treating these patients earlier on. We presume so. We would like to think so, but we don't know that providing interferon therapy early in the course of HCV infection really makes a difference in who will go on ultimately to have cirrhosis or other late stage problems.

We have problems indeed with the test itself that we would be hanging our hat on in this notification system, because of false positivity, and the fact that we are really not determining infectivity, we are just determining prior exposure of a donor to the agent.

Also, we have to acknowledge that today the majority of the lookback cases are from transfusions that occurred over five years ago, which makes tracing even more difficult.

Because of all of these problems, and because of the potential expense involved, about a year and a half ago we began to look at the HCV lookback question with an idea of applying decision analysis. Decision analysis is just a mathematical model to project the effects of a particular clinical decision or a public health decision. The model then allows you to project the effect of a particular choice, use of an intervention, and the amount of resources consumed in the delivery of that intervention and the downstream consequences.

Now decision analysis techniques can be used in two basic modes. They certainly can be used on an individual patient basis to try to quantify for a patient who is choosing one of two different treatments, the likelihood of them reaching the desired goal of better health versus the chance that they may come into a side effect.

The option here obviously is an attempt to maximize the patient's health, and decision analysis can provide some data for the physician and for the patient in choosing the best approach.

Today we'll be looking at another application of decision analysis, and that is more of a population-based approach or macro economic approach where we are trying to determine the relative benefit for resources expended to a group of patients, or to the population as a whole. So the concept here is if one spends a certain amount of money, what is the benefit in terms of increased health for the population?

Now all of us would like to have better health for ourselves and for the population as a whole, and we know that like most things, if we put more money into it, you will get more benefit out of it. Unfortunately, the law of diminishing returns certainly applies here. As you put more resources into health care, you will still get better population health, but the curve does flatten out.

Now what everyone would like us to do including Congress, would like those in the health care system to spend less money and get more improvement in population health. That is a great idea. That is extremely difficult to do.

What no one wants us to do is to spend more money and not get any benefit out of it. That just doesn't make any sense. Usually the question is, there is a new intervention, a new way of doing something that provides an improvement of health, but at an increased cost. What is the angle on this yellow arrow here? Or in the vernacular, what is the bang for the buck for this additional investment in health care?

So this is a situation where decision analysis can be helpful when there are multiple interventions that can be applied, and their outcomes are not the same, or their side effects are different, or when the resources they consume are different, either direct resource consumption or some indirect costs, and particularly when resources are limited. When we can't spend all money on all projects for all people, where can we get the most benefit out of directing our resources?

There are many forms of economic evaluation that can be applied in circumstances such as this. Cost-benefit analysis is mentioned frequently, but actually done relatively infrequently, because the end result of a cost-benefit analysis really usually only ends up being a restatement of the problem, but possibly with some quantitation.

Instead, I would like to discuss today the use of cost effectiveness analysis in this situation. Cost effectiveness analysis is a methodology for evaluating outcomes in costs that usually ends up in determining what is called the cost effectiveness ratio.

It is a mathematical formula where you try to gather all the information about the health effects of an intervention, and put that in the denominator, and all the changes in resource use, one intervention compared to another, are captured in the numerator. So you end up with a mathematical ratio.

Now just a comment on including not only the length of life, but the quality of life, because in these analyses that is important. We all want a long life, and we would like this to be as long as possible. We would also like a high quality life, that is, being able to do whatever we would like to do.

If there were two interventions that both allowed 10 years of additional life, we would want the one that provided the best quality of life, and that is usually captured mathematically, or it could explained graphically as well, you are trying to get the greatest area under this longevity/quality curve. Trying to have the greatest quality of life for the longest time period, just an acknowledgement that not only is length of life important, but quality as well, and that needs to be factored in.

In most cost effective analyses the terminology which comes out is dollars spent per quality adjusted life year. Well, what is a quality? It is not an unusual kind of duck found only in New Hampshire. It is a mathematical expression that takes into account the length of life and the quality of life.

So for example, if two years of life were added to a patient's life, and those two years were spent in absolutely perfect health, the quality factor would be 100 percent, and that would then add 2.0 quality adjusted life years to the model.

If, however, the patient had some other type of morbidity, such as fatigue, the quality of life would be reduced. Now the assignment of quality factors is a whole other science in itself or an art form in itself, but I have just picked 80 percent here to show you. You would take the 80 percent quality adjustment factor, multiple it by two years of additional longevity, and the intervention would then be said to provide 1.6 quality adjusted life years, and so forth and so on.

Death is assumed to have a quality adjustment factor of 0.

The end result then is the cost effectiveness ratio, or marginal cost effectiveness, where the differences in resources consumed between two different interventions is compared to the difference in the outcome they provide. So one needs to tally the cost of intervention A, and subtract the cost of intervention B, and divide that by the differences in outcome that they supply. That is usually expressed as the life expectancy times the quality of life between the two interventions.

Now this has been done numerous times. Many people have made their academic career doing these kinds of analyses, using fairly standardized approaches. What I'll show here is just some numbers that have been generated. There is no congressional mandate that any particular threshold be met, but very commonly accepted medical- surgical interventions have a cost effectiveness which numerically comes out to less than $50,000 per quality adjusted life year.

That is not to say that is what a year of life is worth, but that is just how it usually works out. So for example, giving Rh immune globulin prophylaxis to Rh negative mothers to prevent future children from having hemolytic disease of the newborn costs about $2,000 for every year of life, called adjusted year of life, that is saved.

You can see here is a range of examples. Even some things that the public often thinks of as being relatively exotic or expensive come in below this $50,000 per quality adjusted life year benchmark. Again, there is no absolute requirement that this be met, but it is very commonly found.

So decision analysis then provides a means of comparing alternative treatments, particularly those that have the same outcome, in an objective manner in order to determine the resources expended, and the outcome benefit that is achieved.

There are some limitations to this approach. For example, it is very difficult to value intangibles. There is no way to capture a non-economic variable very well unless it affects the quality of life or the longevity of the patient.

We have to take into account that not all expenditures happen in current time, and not all benefits occur in current time, and for both benefits and costs, a discount rate needs to be applied to future expenditures and future benefits.

It is very difficult to apply these models to a particular patient. When N equal 1, the statistical theories don't work real well obviously. That won't be a concern of ours today.

I will point out that every one of the models that you see in the literature has assumptions embedded in it. It is not possible to entirely recreate reality in a mathematical model such as this. So there are also assumptions that go along with it. One has to read the fine print, and the assumptions can in some cases, shade the results one way or another.

To actually apply decision analysis, it's a relatively straightforward thing to do. One first needs to define all the possibilities that could occur after choosing a particular course of action. That usually is seen as drawing up the decision tree. Then one has to define the probabilities of each one of these possibilities, and that is usually done by reviewing the literature.

Finally, one needs to determine the cost and determine the outcomes for a patient following each one of the paths through this decision tree. This is usually done either by reviewing the literature, or gathering information from your own experience.

Now to apply this to HCV, I would first like to share with you some data that we presented at last year's meeting of the American Association of Blood Banks, looking at HCV lookback in sort of a universal framework. That is, looking back to all donations occurring after 1985, once the donor was found to be anti-HCV repeat reactive.

Using number which were provided by the American Red Cross and estimates of how frequently individuals donate, we were able to predict that from 1990 to 1995, there were approximately 91,000 repeat HCV positive donors, that is, anti-HCV repeat reactive donors.

We then surmised that there would be five transfusable components on average from each one of these individuals, leading to about 450,000 components that were eligible for HCV lookback, extending back as far as 1985. This lookback system only works if the donor has donated a second time, or has a prior donation.

The actual lookback tree, to give you an example of what one might look like, is shown here. If there is a component on which lookback could be processed, one has to make the decision, is lookback done or is lookback not performed? If lookback is performed, is the patient alive or deceased? Obviously if they are deceased, there is no benefit that can be achieved by tracing them; that is one of the assumptions of the model.

If they are alive, they need to be traced to their current location, and they need to respond in order for them to get some benefit out of this system. If the patient is alive but is not located, or does not respond, you have the same situation as if no lookback has been performed at all. In that case, there is the possibility that the patient really had been exposed, that is the units that they had received really had been infectious, and that they may had gone on to develop chronic HCV. There is a possibility that might not have developed chronic HCV.

If the lookback recipient is alive and does respond, you have to consider also whether or not they were truly exposed, or whether the situation was one of false positivity. Have they developed chronic HCV? If they have, are they suitable for some type of therapy, which we have modeled here as interferon therapy. If they receive interferon therapy, do they get any benefit from that?

Now here are some of the assumptions, and I don't mean to bore you with all the details, but there are important from a logistic and practical point of view. We assumed that all blood centers had computerized records that could be used for tracing units. If a blood center does not, it has a big impact obviously on their cost.

We assumed that manual transfusion records were available for donations prior to 1990 in hospitals, but that computerized records were available after 1990. That's an oversimplification obviously. We assumed that all components distributed were transfused, and recipients only got one exposure.

We assumed that the average age of transfusion and notification was 60. This will become important in future iterations of this model.

We assumed that all the HCV cases eventually are diagnosed and followed, even if they aren't found through this system. We assumed that treating when a patient had asymptomatic HCV was effective, and that treatment later would not be effective. That's a large assumption.

We assumed that there were no quality of life adjustments for those patients receiving therapy. In other words, we assumed that therapy has no side effects; that's not exactly the case. We essentially considered all of the recipients just as one group.

Here are some of the probabilities that were used, assuming that 60 percent of the recipients were still alive, about a 33 percent chance of tracing a recipient and getting them to respond from the data based on the HIV experience; a 40 percent false probability rate.

About 70 percent of individuals with chronic HCV would be suitable for interferon therapy, and 15 percent of them would get some lasting benefit out of that therapy. In addition, we assumed that if the treatment were unsuccessful, or if there were no treatment, the HCV excess mortality would be about 0.35 percent per year.

Now for some costs. The committee was very interested in cost data at our previous meeting. Some of these are estimates. Some of these are based on published statements of cost. I show you here what it can amount to, and it can be significant.

For example, identification of past units, about $137 per donor, and an average of five donations being involved for that $137. It's about $27 to trace and contact a patient with a standard letter. If the individual needs additional care, additional follow-up, you can see the costs there.

We also had to come up with some estimates of longevity based on published estimates of longevity after transfusion, that is, if there were no HCV effect, about 12.4 years is the baseline longevity. With chronic HCV for which there is no treatment or no effective treatment the average decrement is a half a year, again, related to the fact that there is relatively little excess mortality of chronic HCV infection on average. The successful interferon therapy would take the longevity of the patient back to the baseline.

To show you overall cost then, if there were no true exposure, the cost per recipient, that is cost for completed lookback case would be $364. If interferon therapy were applied and successful, $4,000. If you were unsuccessful, you would have not only this expense, but the expense of taking care of the patient's HCV anyway, and so forth. So the expenses per case can be quite high.

Overall, the cost effectiveness calculated to about $88,000 per year of life extended, or quality adjusted life year. This then is slightly higher than the $50,000 that is seen for most commonly used medical therapies, but they are not entirely out of the ball park.

I would note what happened; how is it we ended up with something that wasn't quite as efficacious as we might have wanted to? We started out with 460,000 recipients to notify, of whom 60 percent were alive at the time of lookback; 33 percent of those were traced and responded and so forth and so on.

You can follow the numbers down to in the end the yield, in terms of patients actually benefitting were less than 1 percent of the recipients for whom the notification was originally directed. So overall less than 1 percent of recipients whose name came up as part of this lookback model would achieve any benefit out of having received the notification and gone through the therapy if that were appropriate; a much smaller number than most of us would like to see probably.

Showing the same thing here graphically; 91,700 anti-HCV positive donations, yielding 460,000 recipients to notify. We lost some to intercurrent death; some couldn't be traced; some wouldn't be able to be traced; some were not actually exposed; not all had chronic HCV; not all of these were suitable for interferon therapy, and we're left with less than 1 percent that actually benefit from the whole exercise.

Now in any cost effectiveness model certain values have to be assigned to each variable, but those are somewhat arbitrary. We don't know precisely the value of many of these variables. So what's called a sensitivity analysis is performed, where you vary the value assigned to particular variables and see what happens to the results, to see which variables have the greatest effect on the outcomes; to see what the outcome is sensitive to.

So for example, look at the line here at interferon benefit. Here is the number we used in our baseline analysis, but it's a very steep curve, so any change in this variable is going to have a large impact on the outcomes for example.

We went through and looked at that for each one of the variables, and you can see here that the number is not important. The concept that if you change the proportion of patients who benefit from interferon therapy, this would be a large change in the cost effectiveness analysis outcome. Also, the proportion of patients who are alive at the time of notification can have a big effect on the outcome.

The patient age is very important here as well. Notice that 60 years of age was our base case, but we could also look at younger groups of patients. Not too surprisingly, the cost effectiveness number gets better when you go to younger and younger patients, because more of them are likely to be alive, less likely to have died from intercurrent illness. Also, more of them are likely to be suitable for interferon therapy, so they can actually get some benefit.

Therefore, from a preliminary observation on HCV lookback we found that fewer than 1 percent of the recipients to be notified get any benefit, and the effectiveness was very dependent obviously on early treatment being better than late; but younger recipients would do better, and if we had more effective therapy, the outcome result would be much better.

This is, as I said, what was presented last year. Today I would like to propose some new data for your evaluation, looking at some alternative approaches to target the lookback -- so a difference use of the word "target" than other speakers have used -- to target the lookback, looking at those transfusions that have occurred in specific time periods.

Concentrate on those situations, particularly recent transfusion, where we have greater probability of tracing. We can get a younger age of the patient notification, and we can possibly base the notification on confirmed positivity or supplemental test positive positivity, as opposed to just the screening test positivity as we have had in the past.

So previously we looked at all donations back to 1985. One option would be to look back to donations of 1990 and forward. Another would be to look at donations as far as back as 1990, but use as the incident donation, only those donations that were found positive in the second version of HCV testing. A third option, only looking back to donations in the last two years, similar to one of the proposals before the committee.

So the option here in two and three, would then allow us to look only at confirmed positivity, or again, those donors who had had supplemental tests performed, and were positive in the supplemental testing. What would that do?

Well, it would allow us to focus on the people who are most likely found to really be infectious based on Red Cross data, using what is commonly known as "REBA 3.0." About 60 percent of their repeat reactives come out to be positive, and actually three-quarters of them are truly infectious.

Using these data in each year, looking back for two years, one would have about 23,000 repeat donors who are HCV positive by this definition, and an estimate of 3 transfusable components from prior donations for each one of these donors, yielding each year about 69,000 components per lookback.

Shown here are the three different options and some of the variables used. I would just like to point out that by looking at shorter and shorter time periods closer to the current time, we are able to reduce the mean age at notification, and have an improved probability of the recipient of that notification information being suitable for interferon therapy.

Also, by restricting our interest to those donors who are actually confirmed by supplement test positive for anti-HCV, we have a greater likelihood of really targeting those recipients who had been exposed to HCV through their transfusions. Some additional assumptions need to be included in the model. The most important is that all of the version 2 and 3 EIA tests are confirmed or supplementally tested.

Who would be involved in this? You can see the numbers here of recipients actually benefitting. If we were to look at more recent times, this number increases to the point where we are looking at just the last two years. About 1.7 percent of recipients who are notified would actually achieve benefit. It's not huge. It's not 50 percent. It's not 90 percent, but it is better than less than 1 percent.

The costs would decrease as the length of time period of the lookback decreased such that for example if you are looking at the last two years, the cost of the blood service industry of providing this information would be about $17 million per year, as opposed to about $173 million with an open-ended lookback back to say 1985.

The cost effectiveness is also improved with targeting more recent donations. You can see here if you look at just the last two years, the cost effectiveness is $42,000 per year of life extended, as opposed to what I showed you earlier at $88,000.

So what are we trading off here? You can see that as the treatment becomes more beneficial, the cost effectiveness decreases; but the cost effectiveness is already lower for any one of these options than the previous analysis where we looked at all donations.

Additional benefit would be achieved in terms of improved cost effectiveness if we had more beneficial treatments, however, the downside is that fewer patients in terms of absolute numbers would benefit, because we are not going to be going after, not going to be attempting to identify those donations that occurred prior to 1990.

So if you looked just here at the base case analysis, you see if we looked at all donations back to 1985, we would probably end up notifying about twice as many people as if we only looked at the last two years. The benefit here is that the likelihood of the recipient actually getting any benefit from the notification would be improved.

So where does this leave HCV lookback in the cost effectiveness scale of things? Shown here are some other analyses that have been done on the cost effectiveness of various blood safety initiatives and issues related to blood safety. You can see that it actually looks like a good deal whether you look at the prior analysis, the $88,000 per year of life extended, or the current analysis shown in the dark green. It looks like HCV lookback is a more cost effective approach than many other things that blood bankers have done over the last decade to improve the safety of the blood supply.

However, on a different scale now, if you look at HCV lookback compared to other more public health-type issues, you can see that in general it still does not fare particularly well unless you use the approach of limiting the lookback to those individuals who are most likely to get some benefit for limiting it to those donations that are most likely to have transmitted the disease. In that case then, that brings the cost effectiveness estimate of HCV lookback down to within the same range of commonly accepted public health measures.

So in conclusion, just note that the effectiveness of HCV lookback is very critically linked to the age, the date of transfusion and the effectiveness of therapy. The cost effectiveness is similar to other preventive measures if it is restricted to more recent transfusion. Obviously, if we are ever success in medicine at developing more effective therapies, this would affect the yield of the program positively.

Just a cautionary note, that again the model did assume that all version 3.0 reactive tests and the version 2.0 reactive tests that haven't been subjected to confirmatory testing, were subjected to confirmatory testing before the lookback process began, and that is not case as to what has happened in most donations today.

Thank you very much.

DR. CAPLAN: Why don't we open the floor for some questions and comments on this very interesting model. Anybody on the panel?

I had a question for you, Jim. If I followed this correctly, it seems to me that the real benefit is in the interferon therapy for the correctly identified HCV infected person. How would the calculations change -- not that you are going to do them on the back of an envelope -- but if you had people who stopped drinking or prevented transmission by going to safe sex practices in some small percentage way, or even factoring in just what we sometimes hear about people wanting to know their status; those sort of marginal harder to measure things. Any comments about how those would alter the cost effectiveness that you gave us just for the interferon therapy alone?

DR. AU BUCHON: Good questions. Patients may feel better having a transfusion, knowing that they will be notified if there is any potential problem of their unit in the future. That type of psychological benefit would not be captured in an analysis such as this, because on almost all patients it would have no bearing on their future health. No one is likely to require psychiatric hospitalization because they weren't notified, or knew that they were not going to be notified.

The benefit in this model was directed at looking at interferon therapy. Everything was associated with interferon therapy. If for example, interferon therapy and cessation of alcohol consumption would improve the outlook for the HCV notified recipient, then the benefit might not be 15 percent, we might to go to 20 or 25 percent, or some higher number. Obviously, that would improve the cost effectiveness.

That is why I showed you some of the sensitivity analysis curves to show that if therapy were more effective, the cost effectiveness looks better. So it's all been stated as an interferon benefit, but obviously what we are talking about is just medical benefit, which interferon apparently today is the largest interventional part.

PARTICIPANT: Ninety-one thousand is a problem. About how many might be first time donors? Do you know?

DR. AU BUCHON: The 91,000 number takes that into account already. I was using American Red Cross data where 88 percent I believe, shown on the slide, of their donations are from repeat donors. That has already been taken into account in the calculation.

PARTICIPANT: What were the products then that they were going to elaborate?

DR. AU BUCHON: It was assumed that these where whole blood donations that were going to plasma, platelets and red cells. They would not all be transfused. They would not all have been actually transfused and so forth. So the number of components, the number of donations there was estimated based on the Red Cross data from units prepared from number of donations collected.

PARTICIPANT: Only some of those units are going to have platelets, and only some of them are going to have fresh units of plasma.

DR. AU BUCHON: That's correct.

PARTICIPANT: Other units of plasma will go to fractionation, which will have some detergent treatment after the fact, so probably the possibility of imparting infection is going to be very limited, if at all in those [inaudible]. So that might be subtracted out. So I guess I question really the figures and the significance without having the data at how they arrived at for the total number of something like 400,000.

DR. AU BUCHON: This model only looked at the transfusion of blood components. It did not look at blood derivatives, because as you say, of the viral inactivation techniques that are used for blood derivatives today. There would be presumably no lookback that would yield any benefit.

PARTICIPANT: To get the 400,000 figure you would have to have four, at least five different products from that each unit, and I don't understand it.

DR. AU BUCHON: It was assumed that with each anti-HCV positive donation from a repeat donor, that there would have been five transfused components at some time in the past with that donor, over the previous 10 years.

PARTICIPANT: Then you also figured in the cost for management of these cases, but that cost would be there anyway, wouldn't it? The chronic hepatitis C problems of hepatitis.

DR. AU BUCHON: It was assumed that the -- and this may not be an accurate presumption as we gather more information about hepatitis C -- it was presumed that the vast majority of hepatitis C cases, although they may be chronic in nature, are asymptomatic, and that they are not always detected. Once detected, obviously a physician is going to be following them closely.

PARTICIPANT: So you would not cost in the treatment program of any sort, since the treatment is going to go on anyway?

DR. AU BUCHON: For those individuals who were notified of their potential exposure, and who were found to be HCV positive, they would then be evaluated for potential treatment, yes.

PARTICIPANT: But then treatment with interferon for example, would go on and be handled anyway if the patient came to the attention of a physician?

DR. AU BUCHON: Yes.

PARTICIPANT: So therefore that would not an additional cost to the program?

DR. AU BUCHON: That's correct.

PARTICIPANT: So that's got to be excluded, right?

DR. AU BUCHON: And it is not included here.

PARTICIPANT: From the presentation, it looks to me like the cost effectiveness was based upon the interferon treatment, and the recipients benefitted; approximately like 1 percent of them did. Now it seems to me also that the cost effectiveness bottom line could change, since you were saying there was a response to notification of 33 percent; at least that's what I understood it.

If these persons, 33 percent, did respond to the notification, I think the bottom line could change. It would allow for identification of these persons, resulting in follow-up limited treatment, which would be the alpha interferon. Also health maintenance, and also the availability that when new, future therapies came on the scene, they could be utilized. Then that would increase the cost effectiveness. Is that correct?

DR. AU BUCHON: Absolutely. If the patient doesn't respond to the notification, the system can't track that patient, can't apply the future benefits. That is entirely correct, which just reinforces the concept that the more recent transfusions that are then included in the lookback system are the ones that are most likely to be beneficial, because we have a better chance of finding the patient. They are probably still in the health care system, probably still with the same physician or HMO.

DR. CAPLAN: We have time for one more. Eric, over here?

DR. GOOSBY: I just wanted to get a little bit better feel for the impact of your [inaudible]. In looking at your sensitivity analysis, any minor change really in response rate will have a fairly large impact on [inaudible]. What exactly were you defining as a response that determined that 15 percent? If you look at data for people who were treated with alpha interferon, you get responses way up into the 70 percent rate, but persistence greater than two years, persistence greater than one year, what exactly were you looking to determine the 15 percent response?

DR. AU BUCHON: Fifteen percent was our reading of the literature as the likelihood of a long term response. Two years would be reasonable of PCR negativity; someone who had truly cleared the virus, cleared the infection.

DR. GOOSBY: It is also reasonable to assume that drops in viral load with hepatitis C improves one's clinical course, although that has not been completely proven. I don't know if the committee -- we passed out the Hoofnagle(?) article at the last meeting. I just want to highlight that for the committee, that that does look at some of that data, specifically looking at the less than optimal two year drop [inaudible]. In light of its kind of disproportionate impact on the calculation, it is important to keep that in mind.

DR. BUSCH: In your introductory comments you pointed out that the cost effectiveness is preferred, because you can really contrast various potential strategies and just make choices, rather than having to put dollar values in. You attempted at juxtaposing three options with respect to lookback in that vein, and that's very useful, because that is what the committee is struggling with.

I'm a little bit concerned that the last strategy, which you concluded seemed to be most cost effective, didn't incorporate an issue that we have chatted about. Specifically, the units transfused prior to 1990, that are from donors that are found positive now, unscreened product from donors known positive. The transmission rate of those -- probably those donors were chronically infected, and because they weren't screened, those units were going out, and it was very high. The transmission rate was probably about 70 percent, which is the rate of viremia among sero-positive units.

From 1990 to 1992, we know that first generation test was missing probably at least 10 percent or so of chronic carriers; probably even more. So again, a donor is negative on the first generation test, but then picked up by the second generation test, there is a very high probability that that donor was viremic and transmitting during that prior period, again, probably in the range of 50 or 60 percent.

Whereas, second generation screened blood on, which is really the period that you seem to be concluded was the most effective, in those units the risk is very small, probably 1 in 10,000, perhaps less. In addition, if those units were screened as negative by the second generation test, and either picked up subsequent second or third generation, what you are looking at is sero converters, because the third generation test, the additional pick-ups that have been obtained are either the rare sero converters, or people with very old infections, who are not viremic.

So those second generation screen products essentially have a zero risk of extraordinarily low probability of transmission, even though the rare repeat donor is found positive either from sero conversion or because the third generation test is picking up these very old infections.

So I'm just concerned that your model for the recent period didn't factor down the probability that that unit in fact infected the recipient. That you carried forward your earlier 70 percent probability, because I don't think those units from donors who sero convert now, have a substantial probability of transmitting. So I'm surprised that your numbers came out looking so good, because I think the yield of truly finding infected recipients as a result of second generation screening lookback is going to be very small.

DR. AU BUCHON: I understand your question, Mike. It is difficult to predict exactly what window period should be applied here, and interdonation interval period to predict the yield. Clearly, most recipients would be found potentially exposed if we look back the greatest extent, made the lookback program as broad as possible. In the first model I showed, that was back to 1985. I figured that was as far back as any hospital would have records that they could utilize.

If we compare the number of recipients who would benefit from the lookback program in that large umbrella approach, say looking only at second generation and later HCV positive individuals, the number of recipients benefit varies by about two. That is, we would notify and get benefit for about twice as many people if we looked back to 1985, as if we restricted it to just second generation and forward.

That may be an overestimate of the effectiveness of the lookback, restricting it to the more current period, but it gives you some idea of what we are dealing with. It is more cost effective approach to look at more recent donations, but it's possible we may have overestimated slightly, the yield from that.

DR. CAPLAN: Okay, we can come back to this. Jim is going to be around, and we can ask more questions about the model.

Thank you.

Is Lisa Ungerer out there? Ah, good. One of the things that the committee was very interested in was some discussion of some of the legal issues raised by lookback. Lisa Ungerer is going to tell us a bit about that.

Agenda Item: Legal Risks of Lookback - Ms. Lisa Ungerer, JD

DR. UNGERER: Thank you, doctor.

I think it was Dorothy Saylers who once that a man approached a rabid dog with the caution of a lawyer approaching an opinion. I hope that I'm not the evidence of that today, but I guarantee you I do not have a gold plated answer to all of the things that people have been talking about today, which are fascinating.

Let me start with some basic principles of law involving medicine. I'll start with my firm represents doctors in the San Francisco area. We have represented doctors in rural areas, and those doctors are going to be judged by the same principle of law, which is that a physician must act reasonably according to his specialty under the same or similar circumstances. That is, a doctor who is practicing at a huge medical institution and is doing cardiac surgery is held to the same standard as a doctor doing cardiac surgery in a 100 bed hospital.

The hospitals are going to be held to the same standards. Blood banks, being medical providers, would be held to the same standards.

When you come to the question of what is it that patients are going to be told, the primary cases, at least in my state, and I looked for others, but I didn't find them, the primary cases involve what a patient is told before treatment, not after treatment.

In my state it is Cobbs v. Grant, and what the patient must be told is what the patient would need to consider in order to make a reasonable decision about whether or not they should have the treatment. They need to know the risks, but they must know the benefits of the treatment as well.

In my state, there are some recent cases which talk about going back to the medicine itself to determine how far out you go to talk about risk. Do you talk about a risk that is known to exist in only a small minority of the people, of which this particular patient is not a category? For example, a risk only to diabetics, if this is not a diabetic. In my state there is law on that that says, no, you don't have to do that; other states, there is an open question.

With lookback, the question of whether lookback itself, whether it is targeted, whether it is all transfusion recipients, or whether lookback is not done and people come back by reasons of a community education program, can invoke different types of lawsuits.

Whether or not there is a duty to do lookback is completely unresolved, at least in my state. The one case I could find was of a patient who was continued to be seen by Doctor X. Doctor X received some information about a drug product the patient had used some years before. He was held to have a duty to talk to that patient when she came back to see him, but that was a continuing relationship.

When we are looking about lookback, particularly targeted lookback in the context of hepatitis C, we very frequently find that we have patients who have not seen these doctors for a number of years, and hospitals who have not continued to treat these patients.

Leaving the law aside, whether or not look back should be done has to go back to the first principle, which is whether it is medically reasonable, whether it is scientifically appropriate. In this context of this committee, we may need to include whether it is appropriate under the social sciences, because the question is not just one of medicine, but how and if and even more directly, the way to approach a patient to make it effective.

If you determine that the figures that Dr. AuBuchon had up on the board just before are a reasonable way to approach a patient, can contacting patients -- if you look at 460,000 donations, and you come up with I have been able to contact 54,000 people, is that good? Is that not good? That's something that you need to factor into this.

In doing so, you will help avoid what I'm going to call the avoidable lawsuit. The primary avoidable lawsuit that this committee can help to avoid is one where somebody says, I was incorrectly notified. The form of notification was not one that I considered to be optimal.

Bearing in mind that a community blood center, a freestanding blood center is not going to be contacting the patients directly. They will contact the hospitals. The hospitals will, if we use the current FDA model, contact physicians, and physicians will contact patients.

What is it that you are going to be telling those consignees to do, the hospitals to do, and ultimately individual physicians to do? If you tell them to do lookback, the way to avoid the avoidable lawsuit is to give the hospitals and doctors the information they need to communicate to the patient so that it is done in a way that the patient will understand it.

When you download the information from the Internet on the last meeting here, and you download the information from the Internet on the NIH consensus conference, you can't just hand that over to the patient; they are never going to understand it, and many, many physicians are not going to understand it.

So when you follow-up on what the CDC was doing, yes, that is a good start, but I guarantee rural physicians are not necessarily plugged into the Internet, and they are not necessarily plugged in to watching teleconferences via satellite. If that is done, would this committee recommend materials that physicians can give to patients? Or recommend at least that the three branches of the Public Health Service sign off on those materials before they are sent?

Another issue, and I think this may come up later with other speakers, and it has come up a little bit before is not only how far back in time do you go, and that also you are going to have to go with the medicine on, but how do you go about the process of contacting patients?

For example, I think it was assumed in the last meeting that a hospital or a physician would send out a letter to the patient. Well, if there is no response to that one letter, then what does the doctor or hospital do? Do you send out a second letter? A third? Do you do DMV tracing, if that is legal in your state, which some of it isn't? What do they do?

There is guidance needed there, because what is reasonable in that context may well depend upon how far back in time you are going, and whether or not you believe you have actually been able to reach some portion of the patient population who doesn't actually respond back to the same physician who contacted them. I don't know that there is any data on that.

The other thing that this committee may be willing to address, and if not, then they might be willing to describe a multidisciplinary approach so that someone in the Public Health Service can address it is what is to be said in these letters of outreach to get people to come back, to get people to respond? Because the other avoidable lawsuit is the fear of lawsuit.

You noticed a portion of people who were going to test hepatitis C negative. In my state we have a case that concerns AIDS, that someone who feared they were exposed to AIDS -- in that particular case, it was by an accidental needle stick -- but proved to be HIV negative on a number of occasions, was not allowed to sue, because they had not evidenced exposure to the virus. They were HIV negative.

In other cases, and in other states, there are lawsuits progressing for fear of AIDS where the person says, I know I was exposed, usually by sexual route, and I know that this test says I'm HIV negative, but I'm afraid, and I'm suing.

So the law itself on fear of cases is open. The way to approach the patient, the prophylactic approach to the patient is the bedside manner. The bedside manner doesn't come across too well through a letter, but if there are communications specialists who can reach out to the patients to bring them back in to the doctor, then give the doctor something they can do with these patients, so that the patient who turns to be HCV negative doesn't go crazy with worry for the next 20 years, that will help avoid one of the avoidable lawsuits as well.

The last kind of lawsuit I want to address I'm not going to call avoidable or unavoidable. I'm just going to call it transfusion associated hepatitis C. I know of three states where there have been hepatitis C lawsuits. There is one in Florida in progress. There are several in Louisiana in progress. I'm aware of one in Oregon where the defendants got out on motion papers, a summary judgment, because the plaintiff's didn't have a qualified expert to talk about the medicine.

The cases in Louisiana are a little bit different. They are very old cases from prior to 1982. It is a time frame no one here has talked about. The suits are largely based on a law that Louisiana had at the time where hospitals were subject to strict liability in transfusing blood, and the lawsuits there I understand have been largely defended by settling them.

The lawsuit in Florida is progressing. I'm not entirely sure how that will go. It involved transfusions of a patient who believed they got hepatitis C. I have heard varying reports of whether that was initiated by lookback or not.

There are additional questions that may be involved with respect to lawsuits that can come out of lookback, and I don't think that anybody who is connected with the voluntary associations, whether hospitals or doctors, has considered the recordkeeping capabilities. There is recordkeeping in terms of being able to trace recipients.

There is also recordkeeping in terms of being able to defend yourself if a lawsuit happened. If they have tossed out their donor records prior to 1985, you can also be sure that someone has tossed out what their protocols and procedures were prior to 1985 as well, and those would be very hard lawsuits to defend. If you can't say what you were doing, it's hard to say that it was medically appropriate.

Why do I say that this is going to be the forum to come to patients and physicians if lookback is decided, why this forum to try a patient-oriented, a communications approach? It is my understanding that this forum as a public forum incorporates a multidisciplinary approach to these issues of medicine, and is a forum where contact can be linked between the three agencies of the Public Health Service.

In other words, rather than falling back on certain voluntary associations to try to figure out the practical aspects of what should be done, can this forum go forward and pull the Public Health Service together? Work together within the forum in terms of either the best or most efficient method of communication, whether that is community education or some form of lookback, and then go forward from there?

There was discussion by Dr. Margolis that the CDC has been relying a bit on voluntary associations. I tell you that I don't represent the voluntary associations that he was referring to, the liver association and what I believe to be an association of doctors involved in hepatology. There is one court who spoke in other contexts about voluntary associations who involve themselves in health care.

That court felt very strongly that any voluntary association who sets forth health care recommendations, does so at its peril. So that if it is found later not to have been omniscient, it may still held liable. I would not bet here that you could make a broad-based pronouncement and have someone else, some non-governmental entity pick up the ball.

Even though it's going to be a nasty amount of work, if lookback is required, this forum, with the Public Health Service may be the one that is going to have to say what should be done and how.

DR. CAPLAN: Comments, questions?

PARTICIPANT: I had a question. Lisa, were you suggesting that if we do lookback, that the shorter period of time it is done, the less likely you're going to get a lawsuit?

DR. UNGERER: No, I'm not suggesting that at all. I would think though that if you go back -- I'm told that a lot of hospitals retain records for eight years. In my area, some of them retain them for much longer than that. If you go back to a point beyond where hospitals are retaining records, it would be very hard for them to do what you have asked. The person who ultimately comes back and says, I wasn't contacted, I'm going to have a form of notification lawsuit; that is going to be a hard one.

That's going to be a hard one for the hospital, because should they have anticipated that this would come up some years from now may well be part of their defense, and it may be hard for them to defend against that in the absence of records.

PARTICIPANT: Is it important to establish a date like we are discussing right now in that when you had a first generation test and a second generation test? The reason I ask that is would that be a vehicle which would provide protection for the government? The reason I ask that is because with lookback by notification of the recipient provide for a legal recourse known as causation?

Can a person go back, in other words, and say if you didn't establish that date, can they go back beyond that date and say, if you didn't notify me, say we don't decide to notify the 300,000 people prior to our first generation/second generation test, would they be able to come back with some kind of a legal recourse?

DR. UNGERER: They may, but you mentioned the government, and it is my understanding that the government would assert its immunity in that case. For others, I think that if the person is ultimately diagnosed as having some effect from hepatitis C, they may go back and say for example, you did a cut off as of 1990.

Say I was infected with hepatitis C in 1987. Would I say, gee, you should have notified me. I could have done X. I could have done Y. I would think people might file those lawsuits, but I can't give you any quantification of how many people that would be; whether they would have any success with those lawsuits; or whether indeed sending out a mass notification of any kind, whether the community education or attempted targeted lookback would avoid those lawsuits, because with your targeted lookback, remember, you are not going to pick up everybody.

Even with an effort to send out letters to all transfusion recipients, we know from certain studies, at least in the San Francisco area, it is thought that since people didn't respond back to the hospital, that not a huge portion of those people got the letters. So the notification is not going to be a shield. I don't see it as being used as a shield, although you could try. I don't see it as an aggressive step for example, that a hospital could take to insulate against future lawsuits.

PARTICIPANT: But it would be advised to establish a date, just like HIV did?

DR. UNGERER: I think it would be advised to establish dates when things happen medically, so that when medicine changed, if it is reasonable to do lookback now, why is reasonable now, versus 1990 versus 1991, versus 1995? I think those things would be very important to establish, because I was hearing that said today, and it was very close to being articulated, but it really wasn't spelled out in a way that I thought most people could understand it. If medicine has changed, it would be useful to articulate how.

DR. CAPLAN: Dr. Hoots?

DR. HOOTS: Unlike HIV, the burden of proof it seems is different. Here you have got a disease where certainly we know it is transfusion transmitted, but as we have already today, and as we heard before, a number of individuals likely were infected, in fact assuredly were infected by other means which would be virtually impossible to pinpoint.

So in terms of the plaintiff's burden of proof, how protected are entities -- blood banks, hospitals, physicians -- just on that premise alone? Because if we acknowledge in whatever recommendations we make, that HCV is an endemic, if not epidemic disease, that it is in the population from a number of sources, some of which are virtually impossible to ascertain on an individual level, therefore any ascertainment that does occur makes no assumptions about the source of the infection, rather that it is important for the person to be identified for their own individual health needs, rather than [inaudible].

DR. UNGERER: I agree that would be an extremely important and helpful thing to put into the deliberations of the committee, and to include in any aspect of communication to physicians. If you were to create something for patients, I can't tell you how you could incorporate that word-for-word. It would be nice if you could, but the point would be to get them back in, and see if they could be followed. I would agree, that would be helpful.

DR. ALTER(?): I'm Miriam Alter from CDC. I have a question regarding how the court might view different types of notification. We are talking about lookback, and about notifying only certain people who received transfusions before or after a certain date. In fact the intent -- Dr. Margolis, correct me if I'm wrong -- the intent is to notify the public in general about risks of HCV, including that associated with transfusion, regardless of what decision is made about lookbacks.

So would the court look differently a Public Health Service effort to notify the public about their risks of acquiring HCV from a transfusion in terms of in general, making public service announcements, advertisements in the public magazines and newspapers and those kinds of things, versus a letter sent by the hospital or physician?

DR. UNGERER: I think some of that depends on the level of effectiveness that you could ultimately show. It may well be that the public approach could be more effective in bringing people in for follow-up, but sitting here today you can't prove which method is going to ultimately be more effective.

DR. ALTER: You will never know.

DR. UNGERER: That's part of the problem with my profession. I always get people five years after the fact saying how good does it happen. To sit and prognosticate in the future what lawsuits may occur is a little bit on the academic side, and that is why I say you have to go with the medicine. You have to go with what is medically and scientifically appropriate, and what everyone forms a consensus on they think is going to be effective and useful dealing with patients who may have a serious disease, or who may be asymptomatic, or may be HCV negative.

DR. ALTER: Could a patient say -- and I may have asked this question, and you may have answered it -- could a plaintiff say, my neighbor was transfused in 1993, and he got a letter. I was transfused in 1989, and I didn't. I found out because I heard it on television. Would that person have more of a suit or a stronger case than the person notified by letter just because they found out by mass media instead of a specific letter?

DR. UNGERER: I'll commit myself. I'll say that no, I don't think they would have a stronger case. The problem is getting them to recall that after the lawsuit is filed. The only way you would know that really is if that prompted them to do something.

If I heard it on the television and didn't click the television off and go back to making dinner, if I actually did something, then that person wouldn't necessarily have a notification suit, because the notification had been effective. Failure suits come in when the notification doesn't work.

DR. ALTER: I see; and they find out at some point later on that they have this disease, not because they heard, but because of another reason, they find out. Then they realize that they were in this risk group.

DR. UNGERER: Yes, that is correct.

DR. ALTER: I see.

DR. CAPLAN: Let's do one more. Carolyn and Peter, and then I've got to give Ron Gilcher five minutes to come up here. Let's do Carolyn for the last question.

MS. JONES: I understand that lawyers have to give the worst case scenario, and that is your job, but you started off your presentation talking about informed consent. Even back in 1985, when patients received transfusions, it was part of the liturgy for them to know that blood could transmit diseases. At that time we couldn't specify all the disease that are transmissible by blood, and we still cannot.

So part of this has to factor in that patients at that time were made aware of the possibility of transmission of unknown, yet undiscernible diseases. How does that play in the scenarios that you have sort of worked out here?

DR. UNGERER: With respect, I would disagree that many patients were made known of unknown and undiscernible disease as of 1985. As of 1985, some patients may have been informed of a risk of HIV, which was no longer a serious risk, because of testing. They may have been informed of risks of hepatitis in general, which might cover hepatitis C here. I don't think anybody appreciated there might be something such as CJAD, which we don't quite know today whether or not it is in the blood supply.

A number of patients also who received transfusions would have done so on an emergent basis, not multiply transfused patients who come in on a chronic basis. They may not have been informed of risk at all if it was truly emergent -- your trauma victim, your accident victim. So those patients in particular may say, I don't know this was coming at all, and many of them may not even know they were transfused.

DR. CAPLAN: Okay, thank you.

DR. GILCHER: Thank you, Art, for giving me the chance to talk to the committee.

There are some points that I would like to bring forth for consideration to this committee, because our charge and our challenge is very great in terms of the HCV issue on lookback. The points that I'm going to mention this morning are on the slides that I am going to show, as well as the handout that I have given to the committee. I didn't have sufficient handouts for those of you out in the audience.

Many of these points have already been made by speakers this morning, especially Dr. Margolis and Dr. AuBuchon. I'm going to use as the model, the Oklahoma Blood Institute of which I am the president and CEO, and also a blood banker by profession.

A quantum leap occurred when anti-HCV version 1.0 came into being in May of 1990, as you have heard this morning, going from no test to a test, which unfortunately had somewhere between a 35 and 45 percent false positivity rate.

There was also another significant leap forward, not as significant, and that was to version 2.0, which at our blood center went into place on March 14, 1992. Many of the false positives were eliminated. Version 3.0 was a slight improvement, but not a lot compared to the other two.

Now the points that I have on this slide have already been mentioned this morning, but I think they are worth reiterating; mainly that there are probably at least 4 million Americans and perhaps more as we have heard, if you include those who are imprisoned, affected with HCV. Depending on who you read, there are somewhere between 10,000 to 24,000 deaths per year attributed to HCV.

Again, as we have heard, only a small percentage of the infected people are from transfusions. In fact, the numbers that you have heard this morning would place that at between 7 and 8 percent. Meaning that at least 90 percent of the HCV in this country is not from transfusions.

Another important point is that the infecting transmitting donors who do not donate after May 1990, that is not tested, will not be available to the lookback process. We must consider that.

Then blood unit tracking. It is very clear that we can track 100 percent of the units to the hospital, but once you are at the hospital level, it is very difficult to track back to the patient. Hospitals within our system, many have already gotten rid of their records.

Now there are three time periods that I felt were important, and this was really the gist of why I wanted to present today, and in fact this has been in a sense presented already by Dr. Margolis, and really by Dr. AuBuchon.

The three time periods that I see of important is before anti-HCV testing, and I am going to use May 1990, although one could argue that that could be May 1992, since the version 1.0 test had a high false positivity rate.

The second time period is the current HCV testing and call that from May 1990 to present. I'll define present as the point in time where we decide as a committee, what we are going to do, when we don't really know when that will be; whether that will be in 1997, or that may be in 1998.

Then a future HCV testing, which is beyond present.

Let me remark then about my points that I feel this committee should consider for each of these periods. In period one, which is before testing, because hospital records may not necessarily be available, I believe it is going to be very difficult to reach the patients. It is what I call educational publicity that may be most important here, and we have heard that this morning; there are other names to it.

That is, voluntary testing that would occur because of educational publicity that would be put forth to the general population. As we have heard, who does it? Who pays? Who notifies? Who follows?

The legal concerns from these alleged transfusion associated HCVs are important, and again, Lisa Ungerer has addressed that to some degree as well. There are going to be the true transfusion associated HCVs in this group, but there are clearly going to be individuals who will come forth, who got their HCV from another source, and the onus will be unfortunately on the blood center to prove that it did not come from that donor.

If we go to period two, which is after testing began until "the present," and do we recipient lookback, we have previously untested donors who know come forth as a donor, and we detect them. So they are now a positive donor. There really are two kinds of donors that I'm talking about. Then we have the previously negative donor, who comes forth now and is tested positive, and that is the sero converting donor. Those are really the two kinds of donors with which we deal at the present time.

Again, we have the same kind of legal concerns when there is the possibility of an alleged transfusion associated HCV infection. There is going to be the true transfusion associated HCV, and of course there will also be those who got that infection from another source.

But it is period three that I think is an important one for us to consider. This is what I call future considerations. There are so many changes that are occurring at the present time. Let me address it first from the standpoint of recipients, and second from the standpoint of donors.

There are a number of ways that we could approach it from the recipient standpoint. We could pre-test all recipients, and if they are negative, we could then do a post-test at some definable point in the future. We could away repository presamples on the donors, and on the patients, but in this case the patients are recipients, and only do a pre-test if the post-test was positive, and that is to determine if that recipient actually had the infection before they were transfused. We could only do a post-test, or in fact we could do no test, that is, a pre- or post-test. These would be considerations with respect to cipients.

From the donor standpoint, we could collect repository samples for HCV PCR, and I'm talking then about a new generation of testing. If the recipient were to test positive on their post-sample, we could go back, and obviously this donor tested negative by conventional testing, and then do a PCR test to see if they are positive.

Or the direction that we are moving now, and within a year or two years we will see PCR testing technology in place in this country for HIV and for HCV. Could we use this as a pre-transfusion mode of testing, and then not even be concerned about recipient testing, because the number of recipients who would be infected once PCR was put in place would be extremely small?

Then there could be variations of a theme on that, and that is where PCR-type of testing could be used ahead of time to test first time donors, and a variety of different kinds of algorithms could be put in place. This is currently being looked at by many blood centers. These are the points that I wanted this committee to consider.

Now the last slide is really an anecdotal story connected with it. My blood center has a repository, which on September 1, 1997, will be 10 years old. There are approximately 1.6-1.8 million samples in that repository; every donor, on every donation. We have approximately a 2 milliliter segment that we have stored at less than minus 20 degree centigrade.

The story relates that I'm going to tell you now very quickly to an attorney who regularly sues the Blood Institute in cases of HIV and otherwise. That attorney called me about a year and a half ago said -- she calls me Ron now, because she knows me so well -- and said, Ron, I would like to see you.

I saw her and she said when she came to my office, "I have two transfusion recipients who allege that they got post-transfusion hepatitis C from units of blood from the Blood Institute." She said, "I know you have the repository and I can subpoena that, but what I would like to do is to have you test the units if I give them to you, and I will trust you to do the testing. If the units test negative, that is the samples in the repository test negative, I will encourage my clients not to sue you, but if any of the donors test positive, then we will in fact file suit."

Well, God looked over one shoulder, but not over the other. There were two units on each case. These four donors had not come forth as donors since testing had been put in place. They were donors in 1988 in one case, and 1989 in another case.

In one case the two donors on the one recipient tested negative, but unfortunately on the other, one of the two donors tested positive. We cannot be sued for not doing the testing, but the basis of the suit is that we improperly selected that donor.

So this is another kind of issue that has to be considered, and I believe this committee must consider those kinds of issues as well.

Thank you for the chance to speak.

DR. CAPLAN: I feel we are moving down a path where we are getting more and more cries of the heart here from the law into the Blood Institute, but probably no one is going to come with more zeal to this than our last speaker before we take the lunch break, and that is Howard Kaplan, about the impact of lookback on hospitals.

Agenda Item: Effect of Lookback on Hospitals - Dr. Harold Kaplan

DR. KAPLAN: Thank you.

I'm going to take a look at the hospital impact or implement issues. I thought we could look at it, since we are considering a kind of Public Health medication, that we might look at it in the context of this medication being administered through hospitals -- compliance, efficacy, potency, dating period, purity and safety -- to get back to familiar ground.

For compliance, what is the capability of the hospital to comply? What is the infrastructure? That has been mentioned briefly, and there is some data to this, but let's imagine for a moment you are in Paris, and you decide to use your American Express Card. Getting credit involves a 46,000 mile journal over phones and computers. The job can be completed in five seconds.

This is from the "Micro Revolution Revisited," by Large. The micro revolution hasn't really visited the hospitals yet. Imagine if you were looking up for a notification, a record. Two out of three hospital transfusion services have no computerized information systems. They rely on hard copy, not necessarily just paper; some microfilm.

This is based on a study that was done around 1965 [1995?], reported in 1966 [1996?] at meetings. It hasn't been published yet, but I believe it is going to be published in Transfusion by Marfie(?) and her co-workers. Thirty-six state, 910 responders with a heavy preponderance of transfusion services -- and most of those being small hospitals. That's the point about everyone being held to the same standard.

Now let's take a look at what we know about efficacy. This is unpublished study by Ron Strauss in Cornell at the University of Iowa. I think it was completed in 1966 [1986?]. The methods here were the donors who were repeat reactive by HCV 2.0, and they retested prior donors who tested positive by the 1.0, and then confirmed with REBA 2.0 or 3.0.

The notification was through the physician and it was a mail notification of the physicians in which the letter explained HCV epidemiology, the rationale for lookback, and provided counseling advice. To make this easy, the response form and envelope were included, and I believe it was even a stamped envelope.

They identified 21 donors who fit those criteria; 11 of them had no pre-sero conversion donation. Of the 10 with the pre-sero conversion donation, there were 110 component produced; 68 of those 110 recipients had died. The remaining 42, their physicians were notified by mail, as we have just discussed. There are 11 no responses, and 31 responses to this notification of the physicians.

I have handouts for the committee; I'm sorry.

Twenty-one were actually tested. In five the physician chose not to test. The patient had hepatitis, or had prior existing disease; was unable to test. Those are the 10 out of the 31. So the 21 that were tested, all 21 were negative. So out of 110 if you will, components that required the lookback effort, no test positive, no case found. It was a very small number.

I don't want to make more out of this, because when I go into the realm of mathematics, I am clearly out of my competence, but I do know there is the old idea of if nothing is wrong, is everything all right? Here the heuristic rule of three, if we put three over N, we get the percent of missed events with a 95 percent confidence limit interval, and therefore 3 out of 110 or 0.027 percent; 3 per 10,000 trials perhaps.

I know we are basing this on a very small piece of information, but it is the only one I could find, and that incidently, hadn't been published.

Now we have considered efficacy. Let's consider potency in the context of how strong a hit will there be in the hospital to this? This model, as all models, are models for understanding, not as a belief system. The numbers I'm using here are from New York, primarily from the New York Blood Center, and the initial repeat reactive rate here is 0.88.

The Red Cross has figures somewhat lower than this, 0.7 I believe. I have seen some figures that are up to twice this value and more, but I think this probably does represent looking at this, looking at New York state and others, a reasonable model. Everybody seems to get to the same place, somewhere a little bit above 0.2 percent repeat reactivity.

These then, in this model of 100,000 donated per year blood center, this idealized 100,000 unit per year. The red would be the number of units, and the assumption is that we are using a conversion of about 2.5 for each unit made, 2.5 components produced.

We can follow that through and you see the drop off. These first two years are not surprising. Here is where we have the scrubbing effect of getting the reactive donors out of the system, and also coupled with of course the false positivity.

Now taking the 100,000 unit blood center, going to 1,000 at the hospital, my own experience, and based on what we use from that 100,000 donor center, we probably use somewhere around 20 percent of the components produced. So we would probably get somewhere around 1,500 notifications for this 7 year period of time.

Now I have not corrected for the number of first time donors.

The workload impact, and this is an estimate, but a conservative one, and it is in a blood transfusion facility in which there is computerized support, looking from receiving, logging in, looking up the disposition in the computer, getting the chart, notifying the physician, reviewing the chart, written notification, follow-up in recordkeeping, a total of about 90 minutes total time is a conservative estimate with computer support.

If the older records are off-line, we would have to add time to that. Obviously if it is a paper system, and particular paper systems as you go back this far, tend to be filed off site. Off site ranges from warehouses to in some of my prior experience, very wet basements.

Now if we take a look then at our 1,000 bed hospital, one and a half hours per notification, 45 notifications, 2,318 hours. The average FTE of an employee in a hospital is about 1,800 work hours per year. So we come up with about 1.3 FTEs for the 7 year back log at our hospital. I would like you to hold that thought, because it is for per year.

The next thing we want to consider in this medication model are the dating periods. There are two dating periods, the one we talked about at some length, which is how far back we go, and the other dating period is how soon does the hospital react to this new information? Can we take a year to notify once the decision is made, and the information is available?

There is a great deal of pressure for rapid notification. As Yogi Berra said, "It gets late awful early out there."

Now there are a number of forces, the culture at the hospital. We want to intervene. We have the information. Take action. In the regulatory environment what we are required to do, we are going to take that action and document that we have done it as soon as we could. Finally, the litigious environment that we live in propels that forward.

Now if we take the time frame of 12 months, then we have the 1.3 FTEs. If we do it in three months, we need 5.2 FTEs. Now there is not a cost increase, but it would be density of effort; if you will, mission overload, or mission load. We can talk about whether it is overload later, but that would be in three months.

I didn't put one month down here, and I have to admit that my mind boggled when I worked with quarters, and the middle points are in error. Six months should be doubling with this model, and four times what it would require if we did it in a quarter. So it makes a big difference as to how much time we are going to take to carry out this task.

Finally, we were talking about the impact, the infrastructure, how effective it is. Let's take a look now at the purity of the data; the noise to signal ratio. Here we know we have certainly in the early time, a 50 percent -- I think we have had some conservative quotes on false positivity today -- but we will consider 50 percent a close enough approximation.

Now that was at a time, the first two years, at a time when there was no licensed supplementary test available. So if this early time were to be corrected, the 1990, 1991 into 1992 testing were to be corrected, it would require retrospective testing of repository samples, which are far from standard, although I do believe, as Dr. Gilcher points out, his institution and at others, including I believe, the Red Cross, that has been carried out. It is not carried out throughout the country.

In fact, it should be very clear to everybody that today this supplementary testing is not carried out as a routine in many centers.

Now fortunately, with current generations of the HCV EIA, that is a lot closer to true positivity, but still clearly there is a false positive noise.

Now if we look at again the 12 month-6 month-3 month model in which we work to carry out whatever the plan is if there is targeted lookback, what I have done here is just with the red part of the stacked column, indicate the burden related to the 1990-91 time. It just shows again, it is half the burden.

Now we have considered the usual and the very important issue of safety. There is an underlying assumption here that when we talk about cost, we talk about dollars. I would like to just restate the obvious, that there are risk trade offs that go beyond the dollars. Risk trade offs are always a fundamental problem in decision-making. The efforts to combat a target risk can unintentionally foster increases in countervailing risks.

We have talked about false positive notification. I think that is a very heavy burden for someone to go on that odyssey once they get this notification. That is not trivial.

Undesired notification -- some work done by Paul Holland in looking back at known positive units that have been transfused based on subsequent testing of repository samples, some 19 percent of the people did not want to participate; did not want to know this information. So we talk about countervailing risks, and one of the absent voice. There are many absent voices, but these are some.

Then finally an interest of mine, which I will keep uncharacteristically brief is error induction. I will use homeostasis of error, which we can all, I think, appreciate. It is a state where we come back to the same error level, the same risk. Anti-lock brakes have been a technological improvement with the potential for increased safety. Once we have anti-lock brakes, we like to consume that margin of safety. We drive faster and closer to the guy in front of us.

Here the workload in the hospital with hospital staffing -- I have had the pleasure of right-sizing in my laboratory -- we have across the country, supervisors doing bench work. We have a very heavy workload. We probably have some margin, but we start consuming this margin as we add tasks.

These are very closely coupled systems, and very linked. As we increase workload, the potential for error, particularly we are not going to hire clerical people to do this record review. It is not going to be additional staff probably. Much of that will be absorbed by senior staff, who carry that out now, because this review of records is not considered a trivial activity, and it is carefully done with some pretty senior people, because of the implications of notification incorrectly, or not notifying.

So we should look at risk superior alternatives, as phrased by Grayman Wiener(?), and this risk superior alternative may be some combination of the education and things other than targeted lookback by itself.

Now this is a simplified matrix by Grayman Wiener of looking at risk trade off. Here, if we look, does this countervailing risk versus the target risk affect the same population or a different population? Is it the same type of risk or a different type? I think here we move down to risk transformation, where we are affecting different populations potentially in terms of the safety of this activity, and the risk will be of a different type.

Finally, the infrastructure to allow compliance is less than optimum. The farther back we go, the harder it is to assure any reasonable completeness in the hunt. Efficacy with only the small information doesn't seem to be different than what was predicted.

This will have a significant impact on hospital workload, even with the conservative estimates. The dating periods, the issue of how far back we go is not the only issue. I think it is important to talk about in what time period should the activity be carried out at the hospital.

The importance of the false positives, and certainly how this plays out as risk trade off.

Thank you.

DR. CAPLAN: Thank you, Dr. Kaplan.

Maybe we have time for one question, if there is a panel member or audience member who wants to put just one.

I'm just curious, when you projected those FTE hours for record review and so on, that comes from other similar sorts of studies where people --

DR. KAPLAN: That was our experience with for example HIV lookback.

DR. CAPLAN: And you were completely computerized on that one?

DR. KAPLAN: Yes. Well, those things that at least we can look up, yes.

DR. CAPLAN: All right, thank you.

What I would like to do is give the committee a little bit of a lunch homework assignment. When we come back, we are going to hear a bit about some of the details of course hepatitis C, a little bit more information there. We are going to hear about lookback, the experience in some other places on the globe. Then we are going to move to our first discussion of where we are at with respect to our recommendations.

My lunch assignment is this; I would like you to lookback at the now writ in stone Penner-Hoots recommendations. We also have them written on an overhead, which we can show you later, but please review those, because clearly we are going to go back there as a starting point I would imagine, this afternoon.

You might make some notes about things that you have heard today or thought about since the last meeting that you think might lead us to want to modify, change or add to what those statements say.

There were some agency responses that I think you got in the mail too that had some comments, many of which may interest you. I think it would be useful to review those as well.

So if you could at some point during our lunch break, take a look at those documents from the last meeting. Fresh your memory on those. Take a look at the agency comments. Then maybe make some notes, a few summary points of things that you think we should return to in modifying, editing, adjusting this document.

The work plan we will hear a bit more information. I think we have gotten a lot of valuable information from this morning's presentations, and I think it will be illuminating for us hear in particular how some of these other nations have dealt with it.

Then we are going to have it on our plate to sort of wrestle with this for the rest of the day, and come back to it again tomorrow morning, where I hope we can finally at least finalize a pretty clear set of recommendations in answer to the questions we got asked on this issue.

So that's the game plan, and if you do that during the lunch break, I think we will get that much ahead.

[Whereupon the meeting was recessed for lunch at 12:45 p.m., to reconvene at 1:45 p.m.]



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

DR. CAPLAN: Let's get underway. If I could have Dr. Gordon come down.

We're going to have two presentations on the course and outcome of hepatitis C, first by Dr. Stuart Gordon, and then by Dr. Leonard Seeff. After that, we will take a few questions. Then we'll be moving over to the lookback experience in Canada and the U.K.

Dr. Gordon.

Agenda Item: Course and Outcome of Hepatitis C - Dr. Stuart Gordon

DR. GORDON: Thank you, Dr. Caplan, Dr. McCurdy.

As Dr. Seeff will undoubtedly point out in the next presentation, the outcome of hepatitis C is a function of several variables, host related and virus-dependent. We sought to determine whether mode of transmission affects outcome. This presentation summarizes our abstract that was reported at the American Gastroenterological Association in May of this year.

This concept first arose in the late 1980s with liver biopsy studies. It was observed at that time that the then non-A, non-B chronic hepatitis was morphologically more severe in patients with transfusion-related infection than sporadically transmitted disease. The size of the initial inoculum was implicated as the cause of this phenomenon.

After the discovery of the hepatitis C virus, Dr. Miriam Alter described the natural history of community acquired hepatitis C in the United States, and also observed that severe disease histologically was more likely in blood transfusion recipients than injection drug users.

We originally described the histopathology of hepatitis C as it relates to mode of transmission in 1993. In this report we observed that the total histologic activity index score among blood transfusion recipients was significantly higher, signifying more aggressive liver disease histologically than among intravenous drug users.

Therefore, we attempted to determine in the present study whether our previous histologic observations translate into clinical outcomes, and answer the question, does post-transfusion hepatitis C pursue a more aggressive clinical course than disease that is parenterally acquired.

Our study took place at William Beaumont Hospital in Royal Oak, Michigan. This is a 1,000 bed medical center located at the Oakland County region just north of Detroit Michigan. The demographic composition of the hospital system market is summarized here. The region is 90 percent Caucasian, and the socio-economic nature of the region is considered upper middle-class, with an average annual household income of approximately $58,000.

We studied the clinical course of 627 consecutively evaluated non-alcoholic patients who were referred to our institution over a seven year period for further evaluation of hepatitis C. Ours is not a transplant center, and therefore no patient was referred specifically for transplant evaluation.

Forty-five percent were transfusion recipients. Forty-two percent or 262 patients acquired the disease via parenteral exposure, primarily former injection drug use, and 83 patients or 13 percent were without reported risks. Incidently, in an anonymous questionnaire that was subsequently mailed to the no risk individuals, a significant portion of these patients subsequently reported and admitted to prior injection drug use.

Liver histology was available in 463 patients, and showed non-cirrhosis, varying degrees of chronic hepatitis in 59 percent; cirrhosis in 37 percent or 173 patients; and a hepatocellular carcinoma, all of whom had underlying cirrhosis in 4 percent.

Follow-up was defined as the time from the initial presentation with abnormal liver function studies, until June 1, 1997. The duration of follow-up was from 1 to 24 years, with a mean of 45 months.

This slide shows the distribution of age of hepatitis C acquisition, and age at presentation according to gender. The overall mean age at presentation was 51.3 years for women, and 44.8 years for men. Not shown on this slide is that prior blood transfusion was more common among women than among men; 63 percent of the transfused group were women. Parenteral transmission was more common in men; 71 percent of the injection drug users were men.

The age at acquisition as a function of mode of transmission is outlined in this slide. The mean age of hepatitis C acquisition among transfusion recipients was 36 years, whereas the mean age of hepatitis C acquisition among the parenteral group was 19 years. Although the transfused group acquired disease at a later age, the presumed duration of disease was similar in both groups.

By univariate analysis, liver pathology appears to be a function of age at presentation. The age at presentation was 42.5 years for non-cirrhotics; 55 years for cirrhotics; and 66 years for patients with hepatocellular carcinoma. The presumed duration of disease was similar between groups.

This slide points out histologic severity as a function of mode of transmission. I point your attention to the middle group of bars, where the blue bar shows that 118 patients, or 68 percent of the cirrhotic patients reported prior blood transfusion. Only 40 of these cirrhotic patients, or 23 percent acquired infection via parenteral transmission.

Looking at it the other way around, 54 percent of those in the transfused group were cirrhotic at presentation. Among those who acquired the disease parenterally, only 21 percent were cirrhotic at presentation. By multivariate logistic regression analysis, we found that the mode of transmission was more important than age as a predictor of cirrhosis.

During the follow-up period, 59 patients representing 31 percent of the cirrhotic group developed hepatic decompensation. This table then shows the patients with liver failure, and compares them to the 568 patients who did not as yet decompensate.

By univariate analysis, both the age at presentation and the presumed age at acquisition were significantly older in the group of patients who decompensated. The transfused group was significantly more likely to develop liver failure than those who acquired the disease parenterally.

Again here, multivariate analysis showed that the risk of developing hepatic decompensation was more strongly related to the mode of transmission than to the age at viral acquisition.

Kaplan-Meyer(?) probability figures for lifetime risk of development of cirrhosis based on mode of transmission are shown here. In our population, 25 years after acquisition of HCV, the cumulative risk of cirrhosis in the transfused group was 52.5 percent, and was 35 percent in the parenteral group.

Probability figures for cumulative risk of liver decompensation among all patients based on mode of hepatitis C transmission are shown here. Twenty-five years after disease acquisition, the risk of liver failure was 17 percent in the transfused group, versus 6 percent in the parenteral group. This risk was not related to gender, prior anti-viral therapy, or disease duration.

In summary, in our population the probability for development of cirrhosis after 25 years disease duration was 52.5 percent, and 35 percent in the blood transfusion and injection use groups respectively. This probability of liver-related morbidity after 25 years disease duration was 17 percent and 6 percent in the transfused group and drug use groups respectively.

We conclude that post-transfusion hepatitis C is more severe histologically and clinically than parenterally acquired disease. Both the histological severity and the risk of hepatic decompensation are more closely related to mode of transmission than to age at viral acquisition or to duration of hepatitis C infection.

DR. CAPLAN: Dr. Hoots?

DR. HOOTS: Is there anything that suggests that there is a different genotype in the two populations?

DR. GORDON: We did not look at specific genotypes in this population?

DR. HOOTS: How about previous studies with comparing just transfusion and parenteral? Anything in the literature to suggest that there is --

DR. GORDON: Genotype I is more common worldwide, and anecdotally we have seen more genotype I among both groups. There has been some to suggest that genotype III is more common among injection drug users, but we didn't look specifically at genotypes.

Dr. Alter?

DR. ALTER: This is what clinicians see, and I think it is very helpful to have a really nice picture of the disease that is most commonly seen in this setting.

I have to, however, argue one particular point, and that is your estimate of the probability of cirrhosis and decompensation . You are looking at the probability given that someone has chronic liver disease, that they will go on to decompensation, but you don't know that that is the probability of someone who gets infected via transfusion or any other route of actually going on to decompensation, because we don't have the group of individuals in there who didn't get sick, or weren't sick enough to be seen in this setting, or weren't referred to this setting.

So that to look at the probability, you need the entire group of infected individuals, not just those who came to medical attention because of their symptomatic disease, or because they were screened.

DR. GORDON: I don't know how you get at the issue of the denominator, and that is true.

DR. ALTER: In this type of study I just think we have to limit how much we can conclude from this study on the natural history when you don't have the cohort of infected people to follow.

DR. GORDON: Analogous though to other cohort studies representing those patients referred for further evaluation of hepatitis C. It does represent all patients referred over a several year period, so presumably it reflects a representative cross-section of what is seen in the community in terms of initial diagnosis of hepatitis C.

DR. ALTER: So I absolutely agree with you, because this is what clinicians are seeing, but I don't think it is representative necessarily of all of the individuals who get infected. So I just think we have to be a little cautious.

PARTICIPANT: Any serologic data or viral load data that you attempted to look at?

DR. GORDON: Not in this study. We previously looked at viral load as it relates to mode of transmission, and found no convincing correlation between mode of transmission.

PARTICIPANT: How about to development of cirrhosis?

DR. GORDON: Nor that; we found no close correlation in a previous study, although we did find higher levels of virus among those with more aggressive liver disease. This is an issue that has been debated. We have found that there was an association. Others have not found as close an association.

PARTICIPANT: [Inaudible].

DR. GORDON: We did; 150 patients were treated in this group, and we looked at the role of anti-viral therapy and found that it really did not influence outcome.

PARTICIPANT: [Inaudible].

DR. GORDON: We had a larger group that I didn't get into. We excluded alcoholic patients, surface antigen positive patients, HIV positive patients as well. These were all excluded from analysis. There were also 10 patients who died from non-liver related disease during the course of the study.

PARTICIPANT: Two questions; one is a follow-up to the question on anti-viral therapy. Did you have a small percentage of responders in that group, and that's why you didn't see any difference in outcome among the responders?

DR. GORDON: Correct. We did have a cohort of responders both among the cirrhotic and the non-cirrhotic group, and we entered that in and found that that was not a significant variable in terms of influencing outcome.

PARTICIPANT: [Inaudible]. What type of follow-up post-therapeutic intervention?

DR. GORDON: We just looked at those who were treated versus those who were non-treated in terms of overall outcome.

PARTICIPANT: So it wasn't on response?

DR. GORDON: It was based on response -- no, it was based on those who were treated.

PARTICIPANT: Not based on response?

DR. GORDON: Correct.

PARTICIPANT: Okay. My second question had to do with do you have a feel for whether persons with a history of transfusion might be overrepresented in a referral group, because the medical community in general is more aware that transfusions is a risk factor for hepatitis C, and therefore might be more likely to screen individuals who have received a transfusion in the past?

DR. GORDON: That is always a possibility. It is hard to say. These were all patients who were detected in general with abnormal liver enzymes. They were detected by a variety of means. Screening wasn't the predominant method by which these patients came to referral, but it is possible that they may have been detected by that method.

PARTICIPANT: Do you know why they were tested, why their liver enzymes were tested?

DR. GORDON: All of these patients were referred by primary care physicians in general, because they were found to have abnormal liver biochemistries. A small cohort were picked up by blood banks, et cetera, but as a rule most patients were just picked up by primary internists because of abnormal liver enzymes, not because of any screening.

PARTICIPANT: When I said screening, I was referring to enzyme screening as well. I just was wondering why they had been screened. Is it because they had a history, or because they were going in for a routine physical?

DR. GORDON: Mostly routine physicals. The younger injection drug users are often picked up when they go for insurance physicals, disability policies, et cetera. They are picked up incidentally. So they are picked up for a variety of different reasons. It's hard to say whether we are selecting out the transfusion cohort.

PARTICIPANT: Just a couple of comments. I think the point about your extrapolations to overall probability of progression are well taken. If you look at early AIDS calculations, the time from exposure to presentation, they were much, much shorter due to the fact that people were presenting, and we were basically catching the sickest people in the early studies based on clinical diagnosis, and tracking backwards. As we know, those numbers have quadrupled in terms of estimated time from exposure to disease.

Early on in HIV for the nineties there was also data that indicated that transfusion was associated with more rapid progression to disease, and that has completely fallen apart once age has been factored in, as well as other variables. Clearly, your data has major implications to the discussion here about a broad public health campaign, because if transfusion recipients are, as you seem to show, four or five times as likely to progress clinically, then they may warrant a more aggressive tracking, and more aggressive notification method than would the general population.

I'm concern that it just doesn't make sense to me that they would be more prone to this liver disease absent other confounders. They are not any different than anybody else. It's either their underlying health status at the time of transfusion, or their age and their immune control. What is your hypothesis for the biological basis for this accelerated progression?

DR. GORDON: For a nice discussion, the current issue of Hepatology which just came out this month has a paper from France which had similar results in terms of mode of transmission, and there was a nice discussion in this regard. The theory that has been proposed is size of viral inoculant, and this is the most plausible explanation for this phenomenon.

Age is obviously the major confounding variable. We have two disparate populations, young men who are injection drug users, and older women who are the recipients of blood transfusion. By multivariate analysis we found that mode was more important than age, but clearly this is the most confounding variable here.

The plausible explanation other than possibly genotype is the size of the initial inoculant.

DR. CAPLAN: All right, thank you very much.

DR. GORDON: Thanks.

DR. CAPLAN: We have Dr. Seeff next. Then I think we will actually take one more break right after that; a short one, allowing for some slide an AV manipulation.

Agenda Item: Course and Outcome of Hepatitis C - Dr. Leonard Seeff

DR. SEEFF: Thank you, Dr. Caplan.

I was asked by Dr. McCurdy if I would review the issue of the natural history of hepatitis C, which I would plan to do with your indulgence, by spending a few minutes talking about how you go about doing such a study, and the difficulties intrinsic to doing such a study, and my view that the perfect study has not been done, and will never be done for a variety of reasons. I will present that here.

I would also like to discuss my review of what we know about this disease, and then present some data from a study with which I have been involved, a very long-term follow-up study which in fact is a cohort study.

Let me start with what we do know about this disease. I think there is no argument and everyone agrees to this that about 70-85 percent of people who are acutely infected with the hepatitis C virus will not lose the virus, and will progress to chronic hepatitis. I think there is no argument about this phase.

Where we are struggling and everyone seems to have varying views on this is what happens after you get to this point of chronic hepatitis, because there are potentially four outcomes. One is that one could recover completely; or one can go on to develop asymptomatic chronic hepatitis, and live out one's life with asymptomatic chronic hepatitis; or develop decompensated chronic hepatitis, usually with cirrhosis; or end up with what is most feared, that is hepatocellular carcinoma.

So I think that there are two general views that have been expressed, and this the controversy. That is there are those who believe that if one is infected with the virus and lives long enough, that inevitably one will go on to develop severe chronic liver disease, unless you die from some other disease first.

The other possibility is that in fact not everybody, regardless of how long you live with this disease, will in fact progress to end stage liver stage. It will be a subpopulation. The question we have if that is the case, what are the factors that may promote progression, or on the other hand, hinder progression? I guess that's where the controversy lies at the moment.

So the trouble we have with this disease is that the vast majority, and I believe more than 85 percent of those who develop acute hepatitis have no clinical symptoms. So we don't know when the disease begins. If you are going to try to do a cohort study, you have to know where the disease begins, and we don't know in the majority of people.

The other interesting thing is that most people who get the disease and go on to chronic liver disease, don't have any symptoms either. So it's a pretty silent disease, both in its acute and chronic manifestations until you reach towards the end.

The third issue of course is the fact that this is a slowly progressive disease that takes two to three decades at least before you reach end stage, and therefore it is very tough to do a study in which you are going to follow a disease for this period of time.

Here is an example when I talk about the lack of symptoms. This happened to be in -- somebody, I guess, asked this question a little earlier. We did a screen at my hospital in January 1994, of what we hoped were going to be 1,000 admissions to the hospital, and we ended up with almost 900, and to our utter amazement found that 21 percent of people who walked into the VA hospital here in Washington, D.C. were anti-HCV positive.

If we asked about a history of acute hepatitis, only 8 percent knew that they had been infected. So the majority had no idea that they were infected -- excuse me, this was 10 percent, and a history of having had acute hepatitis was present in only 8 percent. So the majority of people simply didn't have any clinical evidence that they were acutely infected.

With respect to duration, as I'm sure you are aware, the studies that most people focus on are the one from Japan, Dr. Kiosawa(?), and the one from this country, Dr. Tong, and most recently a study from France, Dr. Poinard(?). These were studies, at least the first two, in which they studied people already in end stage liver disease, and then tracked them back to when they thought they might have begun with the disease. On the basis of that, they came to the quite interestingly similar figures.

The first two studies suggested it takes about 10 to 13 years before you find chronic hepatitis; about 20 years before you see cirrhosis, and almost 30 years to the development of hepatocellular carcinoma. Recently Dr. Poinard, in an interesting study which has other implications, suggested that the average duration from infection to development of cirrhosis is about 30 years.

So it is a disease that goes on for a long time, and therefore is very tough to do studies on. So in order to do a study, the obstacles we face are: (1) that the disease acutely is not identified in most instances; (2) the duration of development of serious sequelae is markedly protracted, and therefore takes your entire life to come up with a conclusion on this outcome; (3) we are in a new phase now where there is treatment, and that most people are being treated one way or another, and therefore it's possible that may modify outcome.

As a consequence, I believe we will never be able to do a true natural history study. We will be able to do a natural history study in people who are treated, and that may modify in fact the natural history.

So how do we go about studying this disease? I believe that there are three approaches. Either you do a prospective study that begins with onset of the disease, but as I mentioned, the onset is rarely identified, so you don't know when it begins. It's tough to match people if you want to have a case control study if you don't know when the disease begins. The duration of follow-up is enormous, so it's very tough to do such a study.

Or you can do a retrospective study, and that is looking at people who come into a hospital with chronic liver disease. Try to track back to the time where they develop their disease, and then to assess the duration of time. This is the issue I guess that Dr. Miramota(?) has been referring to. This omits those people with unidentified subclinical illness, which I will show you in a moment, who never come to that hospital, because they have no clinical evidence of liver disease and never identify.

The third possibility is to do a so-called retrospective prospective or nonconcurrent prospective study. That is to go back to early studies in which a prospective study was done which permitted a case control study to be done, pick it up in the middle, and then to go forward and see what happens thereafter. I'll show you two of those in a moment.

There is a problem with that also, because you have historical evidence of the disease, and it takes a very special setting to identify people who would fit into this category. Obviously, if you have an intervening period of time where you don't see the patient, you may miss some valuable data.

So what has been done? I'm going to skip this by just to show you that there have been a number of studies that can be divided up into those prospective studies beginning with acute disease. My next slide will show you prospective studies beginning with already established chronic liver disease. I'm simply showing you these to show you what studies have been done. A summary of this is shown here.

In five studies in which prospective follow-up has taken place in the people with acute disease, and this is follow-up studies with a mean of about 8-14 years, clinical symptoms have been noted in this studies in about 10 percent of incidences. Cirrhosis among those biopsied has raised between 8 and 24 percent. Hepatocellular carcinoma in these studies have been rare. Liver-related mortality has ranged between 1.6 and 6 percent.

There is undoubtedly morbidity and mortality associated with this disease even when you begin from the beginning, but it is of modest frequency with respect to morbidity or mortality. On the other hand, if you look at people who come in with already established liver disease, and there are four here, although I'm going to show you a summary of three, of Tuckahashi(?) and Yano(?) from Japan, and Dr. Myron Tong(?) from the United States, these were instances in which people came into the hospital with already established liver disease, and then they followed them to see what the outcome was.

In summary, in three of those four studies -- the other one was a very small one in Australia, and I have dropped that out -- over a mean period of follow-up of 4-11 years, cirrhosis was found in 8-46 percent; liver cancer was found in 11-19 percent.

In fact, in some of these studies patients already came in with established liver cancer, and the mortality was extremely high, so that there is in these studies where you start with already established chronic liver disease and follow them up, there is a high frequency of mortality and morbidity over a relatively short time period. That is quite different then if you start from the onset of the disease.

Well, what about the retrospective prospective studies? I'm aware of two. One is one I will be discussing, which is the one that I have been involved with at NHLBI, National Heart, Lung and Blood Institute Transfusion Study; then there have been a number of studies that I think many people know, the immunoglobulin studies that were reported from Ireland and Germany, and a summary in these studies, there is a modest frequency of mortality and morbidity, and we are going to have to discuss what modest means.

Just one reference slide to the immunoglobulin study, and I'm sure you are all aware of this. This is a study in which people received the Rh immunoglobulin that happened to be contaminated with hepatitis C virus. A large number of people were infected. This is a follow-up of 232 of these women over a 17 year period. Average age at assessment when these were evaluated on recall was 45 years.

About a quarter of them did complain of mild fatigue. If you look at the enzymes, almost 40 percent had normal enzymes; 52 percent had enzyme values between 40 and 100 ALT values. Ten percent had ALT values greater than 100.

Liver biopsies showed mild chronic hepatitis in 55 percent, moderate chronic hepatitis in 38 percent, severe chronic hepatitis in 7 percent. When they looked at fibrosis, 1.8 percent of the patients with these liver biopsies had fibrosis, and early cirrhosis was found in 2.4 percent.

So this was a rather staggering piece of information, because this suggested that if you looked at this particular cohort of individuals and followed them for 17 years, progression was very, very slow and minimal. Only 3 percent at most have ended up with cirrhosis. Well, that's one study.

We have been doing a study, and I'm just going to flash this up for a moment just to show you a lot of people are involved in this. This is a study in which we went back to post-transfusion hepatitis studies that were done in the early seventies. There were five studies. They were two VA cooperative studies. There is a long-term NIH study that has been ongoing here, Harvey Alter(?). There is the so-called TTV study, another NHLBI study, and a study done at Walter Reed.

What we did was we accumulated the data from all those five studies. The importance of those studies was that all of these people were screened prospectively when they came in for development of enzyme abnormalities. So we were able to identify all hepatitis not on the basis of symptoms, but on the basis of serum enzymes. It was pretty much the same among the five studies.

So we combined these studies, and you can see a lot of people, including Dr. Schiff, who I don't see here, but I gather is on your committee, was one of the people involved. We have been doing a study now which is about 21 years at this point.

The first part of it was a mortality study. We wanted to find out what happened with respect to mortality, and we began following these people about 18 years after they had been transfused. We did a mortality study looking at the NDI tapes, Social Security death tapes. We have death certificates on every person. We also have medical records by the way, on every single one of these people from every admission between the time they were transfused and the time we started the study.

I have spent hours going through 6,000-7,000 records, and I think people here will be well aware of the fact that death certificates are fine, but they are not terribly accurate. So we have some data that we can compare death certificate data with what we see from the medical record review, and also interview of proxies, of family members. We have autopsy records on virtually everybody. We have these hospital charts, and as I said here, family proxy interviews were done.

Well, what did we find? As a single summary slide, we published and it caused a lot of consternation in people's minds, including ours, because we didn't expect this, was that at the end of 18 years at our first cut -- excuse me, I forgot to mention we had a control. This was a case control study. We selected two transfused individuals who were carefully matched by about eight or nine different criteria with the patients who were called non-A, non-B hepatitis, and so therefore we had a case control study.

So 51 percent of the cases had died, versus 51 percent of the controls, which was clearly not significant.

Another cut of this, at 20 years of age showed that again, there was no significance. When we restricted this to hepatitis C, and it turns out that 70 percent of the cases were hepatitis C, 30 percent were no. When we looked specifically at the hepatitis C group, all cause mortality, there was no difference.

There was, however, when we looked at non-A, non-B hepatitis cases and liver related mortality, based on the examination of the medical records at 18 years, there was a slight increase in mortality, and it was significantly higher among the cases than the controls. So we did see a slight increase in liver-related mortality.

Now this complicated slide I'm afraid is the survival curves. We have in red are the cases which you probably cannot see. In yellow are the controls. Here we have the data shown with respect to the age at the time of transfusion. It turns out that in this particular study if you were under the age of 55 when you were transfused, you were more likely to die if you were a control than if you were a case. On the other hand, if you were over te age of 55, there was a higher mortality rate among the cases as compared to the controls.

Now this is not the only study. There is another recent study from France, Dr. Poinard's study, that would suggest that age is an important determinant with respect to outcome. We used to think that well, if we are going to talk about treatment, should we be excluding people have their CABG say when they are 65 develop the disease, because after all it takes 30 years for problems to occur, therefore why treat them?

Well, it turns out I think that there is a reason to consider treatment if treatment works, and that is because there seems to be a more rapid progression among those who are infected over the age of 55, at least in this study, and in some other studies, than among those under that age.

Well, the next phase of the study, which is turning out to be extremely interesting is our morbidity evaluation. That is, we are trying to follow all the people who are alive in the study, among those who have not died, and we have tried to recall all the cases and their controls. In every case we did initial historical, clinical, biochemical and serologic evaluation.

We then followed each subject with biochemical testing at least at three month intervals for at least three values, over a period of six months in order to assess and determine whether the patient had chronic hepatitis by our definition, which was that there had to be at least two out of three of those enzymes that were abnormal.

Later on we added a probable chronic hepatitis, even if there was a single abnormality, if there was no other basis that we could find for abnormal enzymes. In many regards those two could look very similar.

We tried to get biopsies wherever we could, or if they wouldn't permit us, we tried to get a surrogate test, which mean upper endoscopy or a CAT scan or something to see if we could find evidence of chronic liver disease.

Well, we ended up out of the 568 cases, which again I'm sorry I forgot to tell you was the number of cases that we are following, and the 984 controls, and we could find 205 patients and 335 controls who were living and in whom we could do follow-up, and most of the data come from this particular group.

This is a single summary slide of where we are at this point, and I will show it to you in a moment in graphic form. Biochemical evidence of chronic hepatitis, that means the persistence of serum enzymes, has been detected in between 15 and 18 years later in 53 percent of cases. Ninety percent of the people who are anti-HCV positive originally are still positive. That means that 10 percent by the way, appear to have lost antibody.

Seventy-two percent are HCV RNA positive, which means that 28 percent appear to have lost virus. The percent in whom we were able to detect chronic hepatitis was somewhere between 24 and 36 percent. The percent with cirrhosis was somewhere between 12 and 17 percent, so I think that there is the figure then that has been used of about 20 percent, which I didn't want to accept before, because I thought that was 20 percent of those biopsied.

These are data that extrapolated back to the entire group, and we think that something like 12 to 15 percent of people have got cirrhosis at this point, and you can see there is about a 2 to 1 ratio. Twice as many people have got hepatitis without cirrhosis, as compared to those with cirrhosis.

Percent of people with overt liver disease, clinical evidence of liver disease -- we are now talking about 18 years later -- about 13 percent have got one or other manifestation. By that I mean it could be splenomegaly, it could by thrombocytopenia. It could be viruses, which we call severe disease, or it could be a psyches which we call severe disease, but there is one or other clinical manifestation in 13 percent of people.

What I don't have over here is the cancer rate. At this point we have 400 people who are hepatitis C positive in a mean of 1974. We have 4 people who have primary liver cancer, giving a cancer rate over this period of time of 1 percent.

Now what was very important is when we biopsied these patients, and we compared the histologic finding of chronic hepatitis and cirrhosis with clinical manifestations of liver disease, what was fascinating was that 94 percent of these patients who had chronic hepatitis without cirrhosis, had no clinical evidence of liver disease whatsoever. Six percent had mild chronic hepatitis, and none of them had severe chronic hepatitis.

On the other hand, once they had developed cirrhosis, 14 percent had no clinical evidence of liver disease, but 43 percent had mild, chronic hepatitis, and 43 percent had severe chronic hepatitis. So the development of cirrhosis at this point in time, as far as I can determine, dictates an outcome that is likely to be as we move into the third and fourth decades, severe. This is where we are going to run into problems.

Do we have any studies that might help us? Well, here is a study from Italy recently published in Gastroenterology; Dr. Alberti's group, Dr. Fatovich(?). What they have is almost 400 people with compensated chronic hepatitis that they were able to, by some retrospective analysis, follow-up to see what's happened to them.

The probability of survival in this particular group at three years was 96 percent. At five years it was 91 percent. At ten years it was 79 percent. On the other hand, if they had an episode of decompensation, the mortality increased to 50 percent. So clearly cirrhosis in the patient who is decompensated has a poor outcome.

Here were their figures with respect to hepatocellular carcinoma, decompensation and liver disease mortality. They figured out that the annual incidence of the development of hepatocellular carcinoma was 1.4 percent; decompensation about 4 percent; and liver disease mortality about 2 percent.

Now this is the outcome of our clinical study at this moment, our follow-up study. If we took everybody who developed transfusion associated hepatitis in 1974 who was anti-HCV positive, where are they now in their follow-up? This says 1994. It should be 1997, excuse me.

Fifty-one percent of the people who originally seemed positive are still anti-HCV positive, are still viremic and have biochemical evidence of chronic hepatitis. Interestingly enough, 22 percent are sero positive and are viremic, but have no clinical evidence of chronic hepatitis, and this is a new group that people are now beginning to worry about. We are seeing more and more of this in individuals who have biochemical normality, but are anti-HCV and PCR positive.

Seventeen percent were anti-HCV positive, but were HCV RNA negative, and they have no chronic hepatitis. Eleven percent had lost absolutely everything, and this has been very surprising to us. They were neither anti-HCV positive, nor RNA positive. They had no hepatitis.

So on the basis of this, it looks as though 27 percent of patients who were acutely infected in the early seventies have now resolved their infection, and 49 percent have got infection, but it is minimal liver disease that we are seeing at this point in time.

So to summarize this study, this is a quick summary that 51 percent of the patients who had acute post-transfusion related, transfusion-related hepatitis C in the early seventies remained viremic, but biochemical evidence of chronic hepatitis. Twenty-two percent remained viremic without biochemical evidence of chronic hepatitis. Seventeen percent retained antibody without RNA or biochemical chronic hepatitis. Eleven percent have neither antibody, HCV RNA, nor biochemical evidence of chronic hepatitis.

Liver biopsies revealed chronic hepatitis alone in about two-thirds, and cirrhosis in one-third. When we looked at clinical evidence of liver disease compared to the liver biopsies, we found that clinically evident chronic liver disease was detected in just 6 percent of those with biopsies showing only chronic hepatitis, none of whom had advanced liver disease, but in 86 percent of those with cirrhosis, about of whom had advanced chronic liver disease.

So our conclusions is that the perception of the severity of chronic HCV infection depends on the stage at the time that you begin your initial evaluation. If you begin studies, such as the prospective studies or the retrospective prospective studies that I have shown you, from the beginning of the disease, you see that is a slowly progressive process of relatively modest severity, at least in the first two decades.

I suspect that those who have cirrhosis are going to run into trouble as we move into the third decade. The question we have is how many people with chronic hepatitis without cirrhosis will advance to cirrhosis and run into trouble?

Studies beginning with already established overt chronic liver disease define a more rapidly progressive course, with evidence of severe sequelae, but this is a biased group. You are looking at people who have come in with already established liver disease, and all that group of people who don't have disease that never get steered to attention are missed out in an analysis of that nature.

Now if indeed there is a limited number of people who progress, the question is, what is it that is responsible? These are the things that I believe we need to focus our attention. We have to look at host factors: age, race, gender, immune status, genetic predeterminance. We need to look at viral factors such as: viral strain; quasi-species; viral quantity, as has been suggested several times; mutant strains; extraneous factors such as alcoholism; co-infection.

I might mention we have tested our group for hepatitis G. We did find hepatitis G in the people who are C positive, but do not find that it has any effect on outcome. So G plus C, at least in this study, doesn't seem to be any worse than C alone.

I see this disease, and I have shown this before, and this is a slide that the AGA has, as an iceberg. I think that more of the disease is undetected than is detected. I think it is the detected disease that gives us the greatest consternation and appropriately so. This is the group that is defined already with I think potentially severely progressive disease, and this is the group that we need to look at.

Our problem is that when we see a patient in a clinic who is anti-HCV positive, or has an abnormal ALT, we cannot determine in an individual patient, as far as I can tell at this point, who is going to do poorly, and who is not. That is, of course, what we need to find out, because I think that we have to do is to look for the risk factors, and deal with those.

Also, we must be able to tell our patients what to expect. Should they expect that they are going to die from this disease, as so many people believe, which is not true in most instances, or are they going to do fine? That's what I think we have to struggle with.

Now this is a slide that Harvey Alter lent me. This is his perception of the outcome, and I'm done. Sorry if I'm taking too long. He feels that over a period of 20 years, this is the rate of progression of the disease, perhaps reaching 10-15 percent who are going to end up with clinically apparent liver disease.

Then there are three potential outcomes. Either this line can go straight, or it can go up like that, or it can go up like this. His data and our data would suggest that it is going to follow this course, and that the majority of people with this disease may yet outlive it. We cannot tell, however, which it is going to be. So we struggle with this issue, and we have to continue to look at this disease in as much detail as we can.

I think I'm about done. I have one slide, and this is from Miriam Alter. She gave me her slide, because I don't have anything in our study that permits us to examine the risk, comparing disease acquired by transfusion versus risk acquired by other sources.

This was the study that was referred to early of Dr. Alter's, in which she looked at the patients with community acquired hepatitis C. This is the blood transfusion group. This is the injection drug abuse group. The interesting thing is that chronic hepatitis occurred at about the same frequency.

If you looked at histology, there does appear to be a worse outcome among those who were transfused, versus those who were drug abusers as a basis for their acquisition. The numbers, however, are very small. We have got only one out of five here, so this 20 percent really is somewhat inflated probably. Also as I think was pointed out, the age at acquisition of the disease was not taken into account, and so therefore, that may play an important role as well.

I think I'm done. Well, thank you very much.

DR. CAPLAN: Questions?

PARTICIPANT: Question one, corollary question two. Why do 11 percent clear their hepatitis C versus why don't the rest do it? What are the factors that you know or that we know enhance clearing versus the factors that do not enhance clearing?

DR. SEEFF: Well, first of all, we have to confirm this unequivocally. I'm giving you somewhat preliminary data. We are in the process of double checking all of this, because this is a figure that is twice as high as is currently accepted. Data from Harvey Alter would suggest that about 15 percent of people recovered from this disease. In our case it was almost double that, 30 percent.

Once we have got all the data in, we will know whether this is accurate or not. Why it is happening, I don't know. It's a hugely important thing. I have an enormous database. I'm busy looking at all of these various subgroups now in great detail to see if we can account for many of these things that you see here. That is a question that deserves further examination, but I can't give you the answer at this point.

DR. CAPLAN: We're going to take a brief break of let's say 15 minutes. Try and be back in 15 minutes.

Is Dr. Robinson here?

DR. ROBINSON: Yes.

DR. CAPLAN: I think we're going to go with you first, because I think you had a plane to catch after this meeting? Is that true? So we'll have Dr. Robinson up next when we come back from the break.

[Brief recess.]

DR. CAPLAN: I would like to get us underway again. I think we are going to find the next two presentations of special interest, because they involve lookbacks in other settings. First, we have got Dr. Robinson, who is going to tell us something about a lookback in England. She is the National Medical Director of the Blood Authority for England. She has left me a number of materials which is brought about their lookback, which I will have copied and made available to the committee by mail after the meeting is over.

Agenda Item: Lookback Experience in the United Kingdom - Dr. Angela Robinson

DR. ROBINSON: Well, thank you very much for inviting me to your committee. It's been very interesting to hear your debate, because as you will discover we were where you are perhaps two years ago, and have gone on from there to actually do an HCV lookback, and I'm sure you will be interesting in where we have got to. A lot of the issued you have discussed, I'll touch, or you can ask me questions afterwards.

Just to give you a little bit of the background of how we work, the first thing I'm sure you are all aware of is the whole of our transfusion services is part of our National Health Service. There is no Red Cross. It is government run, and it is a non-paid, voluntary system.

The Department of Health has an advisory committee, which is the Expert Ministerial Advisory Committee on Microbiological Safety of Blood and Tissues of Transplantation. It is a policy-making body, and it is the only body that can come up with what will be mandatory screening tests. I, as an individual, cannot start in any blood center, a screening test unless it has had ministerial approval.

It is chaired by one of our chief medical officers, deputy chief medical officer. The members are chosen by the minister and the chief medical officer. The observers -- you will be entertained to see the territorials are Wales, Northern Ireland and Scotland. In fact what it means is it is a U.K.-wide advisory committee. What it attempts to do is come up with advice that we all do at the same time, same date, throughout the whole of the U.K.

We have our own professional advisory groups within the service. Again, we try and make those U.K.-wide. It is a little complicated, but these produce guidelines. They are not regulations. It is the group that I would tap into as a professional to say, well, what do we think about this issue; what is the latest medical and scientific thinking; should I take this into the ministerial advisory committee as something that we should do?

So we have a red book executive committee that is like to send us some guidelines for the service. Under that, the most one in this context is the Signing Advisory Committee on Transfusion and Transplanted Infections, which I am a member of, as I am also a member of the MSBT.

The other thing it is important I think to be aware of which enabled us to do this study in the way we have is that we became a truly national organization in 1994, and have got ministerially approved reorganization changes since November 1995. So it is have been a very busy time for people like myself and my colleagues.

What it means is that all of our blood centers -- and again, our blood centers are large. We did have 14, and we are now down to 10 that do processing and testing; but 14 centers still exist that give advice and follow-up, et cetera, where consultants actively interface with the hospitals.

It does mean that as the national medical director, I can exert executive authority over all those centers. If I say something should be done, it should be done. Ha, ha. Well, you have to remember I was actually one of what we call the barrens. I used to run one of the transfusion centers in Yorkshire, so I'm very well aware of the culture that exists out there.

Just to give you a little bit of background, I have to correct that data, I'm sorry. We actually started in September 1991. Now that is significant to this audience, because it meant we started with the second generation test. So we haven't got the problem of the first generation test, the high rate of false positives. We actually started with the second generation test, and with the REBA, so we had a means of confirmatory testing.

That was again, a debated decision by that advisory committee. Until the second generation tests came in, we didn't start universal screening. The other important thing is that every center started on the same date throughout the U.K.

But the decision was made in 1991, no lookback. I wasn't in the committee then, so I can't really tell you why the decision was made not to do a lookback. My understanding is that it was believed that the hepatitis was relatively mild, and there was no treatment, so why inform anybody, and it took such a long time, so let's not do it.

It was reviewed at various meetings, but it was that standing advisory committee in 1994, where according I think to your legal advisory that you had, we considered that it was medically and scientifically appropriate at this stage to undertake a lookback.

Part of that was because more of the natural history and epidemiology of the disease was being known. Interferon alpha was about to be licensed in our country, and trials were suggesting there was a treatment available, and we had the capability within the Service to actually undertake it on a national basis.

This committee, the Ministerial Advisory Committee, considered in September 1994, felt that there was a duty of care, particularly to the younger recipients who were likely to live that long time span, and set up a little core group to report back, of which I was a member.

This discussion there went, well, yes, we recognize the duty of care to the pediatric patients who have been transfused, you can't give two levels of care, therefore the recommendation was that if you were going to do it, you do it for everybody.

Duty of care is actually a legalistic word, as you know, and therefore we had to check all this out with the legal people, who, although it is not written down, believed that a case could be argued that if you had the information that the intervention treatment was available, and you failed to actually do your utmost to let that patient know, then you were failing in a duty of care.

In January it was announced and a working party was set up. On the January 11, 1995, there was a public ministerial announcement with press release, question and answer, brief. A help line was set up. There was a cascade throughout the Service to let all the doctors, together with the question and answer brief, we're back to what's the message; consistency of the message; what are you going to do; how are you going to do it; and when. So through that cascade, every registered practitioner in the U.K. actually got a copy of the press release of what was happening.

Just to compound and confuse, the media got involved and did a wonderful group. "Panorama" is a BBC documentary, and it did a wonderful -- that's sarcasm by the way -- documentary on bad blood. Why hadn't we told all these people that they had got hepatitis C? There were of course 500,000 in the U.K. that have got it. The implication was that they got it from blood transfusion.

It was an appalling program, and very, very difficult for me. I was interviewed on that, but I was not at that stage -- the interview of course was done in December. I was well aware of what we were going to do, but I was not going to let the "Panorama" documentary people know that that was our decision.

We certainly didn't want them to imply -- you have seen the process you went through. From 1991 to 1994, it has been heavily, however, the result of that bad blood program, as we anticipated, was that there were a lot of worried multi-transfused individuals out there, or anybody that had ever been transfused.

So we actually had to set up a help line, and also made arrangements with our public health laboratory service and with our GPs, that is our primary care physicians, that if someone was really worried, and had been multi-transfused, then they would be allowed to be tested and the government would pay for it.

We had to give the preliminary actions for all our blood centers. We then had to get our working party going pretty hard to work out how were we going to do the rest. How to get the counseling guidelines; how to get a uniform, consistent message; how to make sure everybody was informed as someone said earlier, in the right language and in the right way.

I have given some of the documents to the committee members, what was released at the time. I have brought some highlights out. Tom Santville(?) was the minister at the time, and this was part of his public announcement that was released throughout the media in January.

You will see that although recognizing that we have one of the safest blood supplies in the world, we might inadvertently suddenly have infected some people, and therefore we are going to follow them up. Wording is so important.

Another bit about wording. We shall do all we can to care for patients who have become infected in this way through counseling, and where appropriate, treatment.

In the chief medical officer's release, he said a very significant group of words, which again I think might be helpful for you. It said, we will follow-up all recipients where records allow. The classic yes minister cop out, if the records don't exist anymore.

I should add that in our country officially you are supposed to keep medical records for 30 years. That is the medical record document of patients. But of course as we discovered, that is not always the case.

Now I realize this is small, but this is the algorithm of how we actually set about it. In looking at this algorithm, it was trying to make sure we covered everything, to make sure we all did it in the same way. The first thing was to make sure that only those who were confirmed positive in that second generation test actually were the components that we listed, and also make sure that only the components that went to hospitals were listed.

In our country, every hospital blood bank has a consultant hematologist in charge. So our link between the blood center and the hospital was through that consultant hematologist, and through that blood bank. So that was the first step, was to get all those components listed.

The other decision we made was that any donor who proved to the HCV positive when we started screening, we worked on the assumption that all previous donations were likely to be positive. That was based on some Scottish data where in fact they had archived material back to ten years, and although they were small numbers, it did suggest that most of the donors who actually found to be HCV positive at the time in September, nearly all their previous donations were positive.

The algorithm goes on, and I won't go into great detail, but what I am trying to illustrate is that we tried to cover every option. For every option, we had a standard letter to notify the hospitals through the consultant. A standard letter that had gone through the legal people to make sure that what we said and how we said it was appropriate.

We had a form, and we tried to produce forms that were simple, easy to fill out for every component such that we could build up a database. There was the LBF-1. That was for each component made from any of those donations that were positive, there was a form which had the necessary follow-up material that the technicians in the lab could try and fill out for us, and also on which we would find out who was the consultant in charge or the GP of that patient.

Then there was the suitability of the recipient for counseling, and did they want us to do it, or did they want to do it themselves. This is a very important point, and I have heard that discussed today is how can you insure that the counseling that the patient receives is of a high standard, is consistent, and you are all saying the same thing?

We actually produced a package. If the consultant or the primary care GP wanted to do the testing and the counseling themselves, then they were sent a package of this is the information, together with a question-answer brief for all the simple questions they might be asked.

Finally, whoever did the testing and counseling, this is the form that comes back to me at the moment, so all the living recipients, I've got the test results, and a little bit more information back.

All the letters to the GPs, consultants, patients, they were all standardized, legally approved in terms of what we said, how we said it, and how we communicated and contacted the recipients. I think in that case, yes, everybody did use those letters.

The other difficult part was who was going to do the testing, and who was going to pay for it? This is where I did have fun and games. There were ten laboratories that were recognized as being able and capable of doing a reliable HCV PRC test, and this was the algorithm they were supposed to be doing, which they had to contact the recipient; screen them by an EIA; if that was positive, to go ahead and do a REBA; if the REBA was positive, that came back positive, and there was no need to do a PCR. They would be onward referred to a hepatologist.

If they were negative, then a PCR was done. If they were either indeterminant or negative by REBA, a PCR was done. That is where I had fun and games, because when I got the results back, everybody had done their own serological algorithms, and it made interpretation very difficult.

It was highlighted in this advice from our chief medical officer. It was felt terribly important that the patients hepatitis C status was done by an approved diagnostic microbiology laboratory that really had the capability of doing a reliable PCR test for hepatitis C. Even with that proviso, some of our results had to be repeated and sent to the reference laboratories.

I won't go into the, because there is limited, in some of the counseling guidelines. They are in the documents I have given you.

What I will do is what are attempting to do is get a lookback central database. So we know the number of donations, the number of components that were made, and the number that were issued to hospitals. If they have gone to another hospital, we had to go back and go around again. The LBF-1 was that form that every hospital got where we issued the components to.

There was no further action if it was found the patient was dead, not traced, or not transfused.

The number that were followed up, of those that were alive, and the LBF-3 is that final form that comes back to me. At the moment, I have only got the data on all the living recipients. I haven't got all the data on the recipients that subsequently died, but I'm in the process of getting that from each of my blood centers.

I suppose the bit you have all been waiting for is how far have we got. Remembering this exercise began properly in April 1995, it has proved a huge exercise. Again, to put into context for you, our population is about 50 million compared to the states of say about 250 million. So we are sort of that much smaller, but it gives you an idea of how you can relate your figures or your potential figures.

So you can see that the number of donors identified pre-1991, was nearly 3,000. The number of components identified and notified to hospitals was just short of 12,000. I have to say this was quite close to our estimates that we had made beforehand, like you have heard from James AuBuchon. We were pretty close to what we thought we would find. We thought we would find 3,000 living recipients from the number of transfusions that we do.

Again, to put it into context, we collect about 2.5 million donations, and we think approximately 800,000 patients receive transfusions per annum. I think we worked on the assumption that a donor would have given at least three times before, and we would have made two components from them.

You can see that the hospitals actually managed to identify 6,000 recipients, and we have followed-up nearly all of those that have been notified to us. You can see that nearly 4,000 actually have subsequently died, and that there are just over 2,000 living, of which we have actually followed-up and counseled 1,500.

Thinking of the time span involved, this has taken us a long time to get to that stage, and most of that has been the hospital records. That has required a lot of stick to actually try and get the information, because a lot of the records, like they are here, are paper records. It is time consuming, and does have to be done by someone who knows what they are doing.

Now I have here, and I apologize, I have only got this analyzed up to March, and we have more than this now, but this is the data that I have got back on that LBF-3 form. So this data is where I have actually got the results and entered them on my central database.

I looked at all the results, and tried to be -- regardless of which serological algorithm they used -- tried to be very objective about what which were positive, which were negative, which were indeterminant, and which I can't really determine which are which. So in my central database, and these are all the living recipients by the way, none of the dead ones, I have got just over 1,000.

Just to put that in a slightly better form, or more graphic form, you can see that of the living recipients who we believe were put at risk by receiving an HCV positive blood component, 581 were positive, 369 were negative, 75 indeterminant, and 27 I got no results at all.

The thing I should also say about when we started testing is that although we did the second generation HCV EIA and we did the REBA, we did not always confirm by PCR. So in fact it could be that some of those that are negative didn't actually receive a HCV positive component.

Again, a preliminary analysis of the data I feel fairly secure about, just to give you an idea of male/female split; how many actually are parents; who sort of ethnic origin I have got so far. You see the vast majority are Caucasian.

I think you will find this one interesting. Again, I have to remind you that these are the living recipients, not the recipients who have died. You can see in the age group, this is the age of the recipient at the time of transfusion. The blue are the recipients who have been found to be HCV positive. The green are the recipients who are HCV negative.

You will see interestingly in the younger age groups, there appears to be a greater likelihood of possibly the infection being cleared if they were transfused with positive material. Whereas, as you get older you see there is a much greater proportion of the recipients who are actually HCV positive.

We are analyzing that in more detail, but it does appear to be that there is some age factor involved in whether the donation transmits or not, or whether in fact that recipient is able to clear the infection once it has been transmitted.

I also had a chance to look at the component types to see if that had any influence on the ability of the infected component to transmit. You will see that we have got FFP, platelets, red cells, whole blood, other unknown. You see really that each component would appear to be equally able to transmit. There is no difference, although I know some of the numbers are small.

Now you are wondering how far should you go back. We actually tried to go back as far as records would allow, I had been instructed. In fact, some hospitals and centers have been able, as you see, to be able to go right back to 1978. If you take 1985 -- 1985 was when we started HIV screening. In 1983, we started having the HIV information leaflets. So our screening of donors from sort of 1983, onwards should have actually taken out donors who were previous IV drug users, et cetera.

What we found in our HCV positive donors who we see, and we counsel all of them, many of them maybe have been -- 70 percent of them actually were IV drug abusers in the seventies; maybe did it once or twice, completely forgot about it, and therefore didn't regard themselves as a group that should be excluded from donating.

So I thought you might be interesting in actually the span there of our ability to actually lookback. The other rather interesting feature, earlier on it would seem that our selection of donors maybe has got worse, rather than better in terms.

Now I have only got the information on the forms, and I'm following-up information now on all the recipients to get more information and to get donor linkage information. Of those recipients that have turned out to be positive, 3.4 percent of them, 20 out of 581, have evidence of symptomatic liver disease.

Just as a by-way, we did this at the same time, because we do not test for anti-HCV core, and therefore we just wanted to check in this multi-transfused group, whether in fact they had also been transmitted with hepatitis B.

The information I get on the form is I get their ALT results. You see that 20 percent have ALT results higher than twice the high reference. There are some AST results, and there is bilirubin. I haven't got the liver biopsy information yet. All those that proved to be positive would have been referred to hepatologists, and if they had raised ALTs, almost certainly they are going to have a liver biopsy.

It is an enormous amount of work to do a lookback like that. It involves our own transfusion staff. It involves the hospital blood bank staff. It involves a lot of consultant time. Therefore, we felt if we are going to do this, we want to try and get the maximum benefit out of this major piece of work. So what we have actually got funding to do now is we have got a national registry of transfusion acquired, known date hepatitis C infection. Obviously the core of this will be transfusion acquired.

I can already hear some arguments, it won't be 100 percent accurate, because some of those components that we transmitted, we can't be absolutely sure were positive, although the vast majority were.

Then any other HCV infections from other sources with its known date will go in this central database. We are now actually getting all the additional information from the donor link up material, and getting a link post between ourselves and our PHLS and CDSC. We are going to institute an annual follow-up. It will provide a resource for hepatitis for researchers if they want to do a particular study of HCV, then maybe Dr. Seeff, we can provide them with an appropriate cohort of a known date infected patients.

The aim of that registry is such that we can monitor natural history and the long-term outcome of HCV. That we can institute an annual follow-up. We can maybe update the estimates of time from HCV infection to biochemical, histological to actually being related illness and death; monitor the numbers of new infections as they get reported; and share this information both nationally and internationally.

Again, thank you for inviting me to come, and good luck.

DR. CAPLAN: Why don't we see if there are any questions for Dr. Robinson about the lookback experience there?

PARTICIPANT: Dr. Robinson, of the 581 living identified positives, do you have any data that would talk about the number that have identified an entered into care?

DR. ROBINSON: Yes. I can't give you that figure now, but yes, it actually was relatively few. We do have that information on the database; those that have already been notified. That is on that LBF-3 green form. So I would be able to go to the database and say, how many had already been notified. It's actually quite a small proportion of the total number.

PARTICIPANT: Two questions. One, you indicated that you coordinated having PCR performed on the recipients who were tracked in the lookback, and who were either screened negative or REBA negative. Do you actually have evidence on the yield that? What hiccups of infections were found by PCR that had missed by standard serologic methods.

DR. ROBINSON: I was just trying to check that, because I knew you would ask me that question. When I last looked at it the night before, we had four cases so far that were PCR positive, but EIA and REBA negative. It might be slightly more than that. In each of those cases, they were immunosuppressed individuals.

PARTICIPANT: It might be interesting to do linkage studies on those cases with the donors.

The other question is, prospectively you have focused on your program prior to implementation of HCV screening. What is the British policy and experience with continuing doing looking back associated with sero converting donors subsequent to screening? Do you do that, and do you have comparable yield data to parallel against the retrospective?

DR. ROBINSON: Well, when we instituted the lookback from September 1991, at the same time we then made sure that any sero converters were also followed-up. It is a separate database, and there are far fewer. The donor base now is so clear, our instance of sero conversion is very low. It's not going to yield any huge amount of information.

PARTICIPANT: I wonder if you would comment on how the differences in the medical care system of the United States versus that in Britain might affect your ability or one's ability to do the kind of lookback that you have described?

DR. ROBINSON: I think probably the biggest difference is the one that all the transfusion centers, all the blood centers are actually within the National Blood Authority's control. So you can actually say, this is what you are going to do. This is when you are going to do it. This is how you are going to do it. You still get variations, but you stand a bigger chance of doing it.

Now the same applies to the hospital. We do have some private hospitals, but most of them are within the National Health Service. So again, you can use some stick to say you must do it; that you are obliged to do it. They will shout and they will squeal, and central monies were provided. I think it cost us about just in testing alone, I think they set aside something like 250,000 -- that's pounds sterling.

The biggest difficulty at the moment is the hepatologists are really struggling now, because these patients are being referred to specialist liver centers. They say they haven't been given sufficient monies to supply the interferon. There are some combined interferon/riboviron(?) trials in progress. If they get referred to a specialist center, then they will get the money centrally funded.

The other thing was at the same time, there is a huge amount of governmental monies on HIV research, but very little on hepatitis research. There has been a shift now with actual tenders for studies to be undertaken in hepatitis, and the hepatitis C field in particular, of which the registry was fortunate in getting that funding, because that is a resource that they will all need to have access to.

DR. CAPLAN: Just one question about the content of the counseling. It basically says what? We have some of these materials here.

DR. ROBINSON: You have that. It was interesting that I think your legal person was saying you've got to say it in the right language. So the way we did that was first of all, you had the ministerial press release, and then you had the sort of department of health press type of language.

Then we had with the help line, it was very much a lay person's question/answer, which again you have that very basic, even to people who don't know what a transfusion is, or if they have ever had one. The difference between a product and a component is quite complicated.

The information that was cascaded to our primary care physicians, which we call our general practitioners, was a full -- basically, we assumed they knew nothing, but gave them the scientific information that they would require; then for those that were interested, more in depth follow-up. The main message was if you have got one, and you do the counseling yourself, you use this brief to counsel them, and you refer that patient to a specialist at a unit if they are positive.

We felt that was very important that a uniform message was given. We did have one charity -- I think we call it the Liver Trust -- that wanted to get involved, but actually we maintained control of the information ourselves. I think that being said, if you do it here, then that is very good advice. Do it from your committee.

PARTICIPANT: One of the difficulties we have in this country is people are so mobile, they move an average of once every seven years. Also there are so many fragmented media choices, it is hard to use them The London Times and get as many people, but are people as mobile in the U.K.?

DR. ROBINSON: They are as mobile, but the key really was -- you probably don't have a similar system, but with our general practitioner system and our national health system, every patient is registered with a general practitioner. They have a sort of national health number. So if you can find the consultant on the medical record, in the medical record at the hospital there will be the GP. Even if the GP has not got that patient, that patient will have registered with another GP, and therefore the records get transferred.

So we are probably better able to trace patients within our health service than you are, through that sort of primary care registration system. Everybody does it, because it costs you nothing, and you usually have to see your GP at some stage during your life.

DR. CAPLAN: Let's do maybe two more.

PARTICIPANT: [Inaudible].

DR. ROBINSON: Yes, it's the process we have actually got with the investigation. I think that there are several stages along the way where you have got hiccups. First of all, the list of components goes to the hospital. At that point, you may lose some, because the hospital blood bank records have been lost or burned or are untraceable. So you lose some there.

The next one is can you find the medical record? And we lose a few there.

We got to the point where in fact we have just over 6,000 recipients that we were able to trace from 11,000 components; about 7,000. If you look at those, two-thirds were dead, but it meant we had actually been able to trace and find them and they were dead. One-third were living. Of that third that were living, on the main date that I have got, which is asking around -- I haven't got the full information back, but on inquiry and updating -- we are about 600 short in terms of have they actually counseled them yet. I believe that probably we will be about 18 and 20 percent where we will never be able to trace.

The form was done in such a way that you will counsel that recipient unless you can state valid reasons as to why you're not going to. They had to be either the patient was dying or it would upset them, or a certain age. So there is only a small proportion of those were where the consultant said, we do not think it is advisable to inform that recipient. There are a few.

PARTICIPANT: [Inaudible].

DR. ROBINSON: Once informed, the vast majority agree to be tested.

PARTICIPANT: A politically difficult question, but in England you, one, didn't introduce anti-core. I don't know if you did ALT. Two, there was a delay in HCV testing relative to the rest of the world, first generation to second generation. How much did that influence the resolve behind the country's efforts into this lookback program?

DR. ROBINSON: The main reason was it was becoming quite clear that HCV was not a benign disease. It was also very clear -- actually, I have to tell you at the time we as transfusion centers, wanted to do it. If you are going to start a test, and in fact I've got into the NSBT and said, look, if we ever do this again, can we say that lookback and mandatory, unless you say why not? I liked your legal person who said, is it medically and scientifically appropriate to do so?

The question there was, well, we don't know what sort of disease it really causes. There is no treatment, so why tell anybody?

PARTICIPANT: The value of your delaying implementation, by the time you started, you had a good test and a confirmatory --

DR. ROBINSON: The only thing I would say about that is you can imagine the potential litigation now. We've got four groups. We've got the pre-1985, the hemophiliacs. You've got this very critical period between 1989 and 1991. You've got the pre-1989, where our government would say, we're not liable; no test. So I think the test cases will be coming up soon. That period between 1989 and 1991, why didn't you?

A lot of our argument was there were false positives. Well, there were false negatives as well. There are a lot of arguments. I give us a 50-50 chance. We started testing with a test that was reliable, that we could confirm, and we deliberately delayed until the second generation test came in. I think that is going to be interesting.

DR. CAPLAN: All right, I have to let Dr. Robinson go, because she has got a plane to catch. I want to thank you for your presentation.

I'm going to ask Dr. Gill to come down, and he is going to tell us a bit about the Canadian experience with lookback.

Agenda Item: Lookback Experience in Canada - Dr. Peter Gill

DR. GILL: I was asked to talk briefly, for ten minutes or so, about the Canadian experience, and to concentrate a little bit on the cost of what it is costing us to do this lookback.

I would bring you into the picture, and as very good introduction to it all from Angela Robinson, about the difference in the systems between the U.S. and Canada, and the difference between Canada and the U.K.

In Canada we have been subject to a royal commission of inquiry into the tainted blood scandal going back to the mid-eighties, and the reason why we delayed HIV testing between March and November. The launch of that inquiry led into concerns about another potential blood scandal, that is the hepatitis C virus infection of the Canadian blood recipients, and why we didn't introduce ALT and anti-core from 1986 on, when the United States introduced it.

So the HCV side of the story arises mainly out of that, and that royal commission of inquiry conducted inquiries across the country from 1993 on, in which the main stories coming out at that time were the stories of the recipients. So on television every second night or so, you would get a group of people who were victims of the Red Cross tainted blood scandal. So that the general public was alerted to hepatitis C through the commission of inquiry.

What we did in the Red Cross in 1994, was look at what had been going on in terms of the pressures that we were feeling about doing something about hepatitis C, the pressures that we were feeling from litigation on HIV in which we were advised that we should do all in power to identify recipients of tainted blood.

So that in 1994, we formed an ad hoc committee of a few of us within the Red Cross, and two or three epidemiologists from the Public Health Service, and debated the issue of hepatitis C lookback. We came to the conclusion that we would do hepatitis C lookback based on litigation. If we had the knowledge, we should go and track the recipients.

If there was an opportunity to do something with respect to lifestyle change, the consumption of alcohol, that was significant. It being an emerging virus, we knew discoveries were going to be made every X month. In terms of the evolution of treatments, we expected to occur over the next few years, therefore we decided to go ahead and do a targeted lookback.

Now at the same time, there was a movement based on the pressures from the Creaver(?) Commission's early report in February 1995, that hospitals should do all in their power to notify recipients of blood going back to 1978, as to whether or not they may have acquired HIV and hepatitis C.

So the hospitals were put under a lot of pressure. When we launched our hepatitis C lookback program, we did get funding from the Canadian blood agency, and that is an agency that represents the provinces. The provinces are the funders of the blood supply, not the federal government. The 10 provinces more or less went along with what they were going to do. It required of them to be prepared to provide the testing facilities, which again is an eventual responsibility for the recipients of blood.

So we decided in mid-1994 to work with the liver foundation on an education program, recognizing that both the public and the physicians needed an education program. The liver foundation convened a meeting of 25 hepatologists in February 1995, came up with a program of education for the general practitioner.

We put on a program jointly with the liver foundation to advise general practitioners of the kinds of patients that might come to them from either the donor base or the recipient base in the Red Cross, as well as the general education program for physicians. That was put on in very city where there is a blood center, just to get the whole program going for the public in general.

In addition, the liver foundation made a brochure for physicians out of the program, and distributed it to 30,000 general practitioners. They put a brochure out for the general public. They had a 1-800 hotline for physicians, and a 1-800 hotline for the general public. So this is how we introduced our educational program in March 1995, at the same time as we introduced our lookback program. So our lookback program spans from March 1995 formally.

So to give you some idea of the background across the country, this is the national data on hepatitis C, in that we started testing in 1990, as soon as the test became available. The same week you started in the United States, we started testing for hepatitis C with the first generation.

Because we were not doing ALT, we were not doing anti-core, and as soon as the test became available, we introduced it in our centers. This worked out around 1,550 positive cases on an annual basis out of about 1 million donations a year.

We introduced EIA-2 in 1991, and at time we could purchase it. Even though it was not licensed in the United States, we purchased it, because we were allowed to with agreement to buy a test ahead of licensure if our own regulatory agency approved it. So we brought EIA-2 in 1991.

We introduced REBA at the same time, and test thereafter. The basis of these figures are using EIA-2 and REBA, and that became our standard test.

All the donors that were found repeat active on EIA-1 and were submitted to a REBA-1 testing, we happened to have a stored sample on them in the National Reference Lab, and so we retested all those with REBA-2 to clear up the false positivity of the EIA-1. That data was used to start to lookback at the recipients of the previous donations.

Just to give you some idea of what we are looking at in terms of the donors, the ratio of first time donors to returning donors is about 1.5 to 1, but the prevalence in first time donors is about 14 times higher than that in returning donors.

In terms of what we were looking at for potential incidence of sero converters, these are the numbers we actually did find in terms of donors who previously tested negative on an EIA-2, and then were subsequently found to be REBA-2 positive on an annual basis. This is out of about 1 million donations a year, or about 750-800 donors a year.

So we picked up 10 in 1993; 6, et cetera. So that the actual sero conversion rate to hepatitis C is very low in the donor population. Most of our concerns in lookback were donors who had not been screened previously, and had returned after one, two, or three years after testing started. Therefore, we had a need to identify the recipients of those donations.

So we looked at it in two parts, the ones that we identified which were positive before March 1995, when we introduced the test, and after March 1995. In terms of the completed lookback, I don't know the denominator here in terms of the number of returning donors. That figure is still being found.

You can see that components transfused, and in terms of the recipient status, 32 percent were alive; 62 percent dead, and 6 percent not located.

In terms of the positivity, 64 percent of those that we found alive were positive for HCV. After 1995, I am not quite sure of the bias in these figures here, but insofar as we are testing ongoing, it would look to be for some reason or other, that the bias is towards not finding positives at the same rate as prior to 1995.

In terms of looking at the year of transfusion, again it rather very much follows the British story that some of the hospital records may go back to as early as 1982, 1983, 1984, but the bulk of them are concerned with more recent transfusions.

If I remember correctly, when I was a royal [inaudible] in Montreal, we had to keep our records for seven years. I don't know whether that is still so, but probably it is on average about seven years. You keep them for eight years to cover yourself for the seven years. Somebody mentioned that I think here in the United States, eight years is a common time to keep your records.

In terms of age, of those that are alive, with implicated units you can see that given the gender of the age when transfused, there is a cluster of female gender outweighing the male gender in childbearing years. Something like 66 percent of the recipients are under the age of 60 at the time of transfusion.

That is the age when we started the lookback. There is a shift this way, as you might expect.

Now in terms of the cost, I managed to get some figures out of one major center about where they are at currently with the lookback program. They have identified 5,361 blood components from hepatitis C positive returning donors, and 3,207 of those components were assigned to the local hospitals.

In terms of the cost of setting up the lookback program and chasing at the hospitals, and in terms of following-up at the hospitals, because of our being found at fault for not doing all in our power to notify hospitals and follow-up with the hospitals, whether they had themselves instituted proceedings to find the recipients, we started sending registered letters automatically to the CEO of the hospital, as well as the blood bank director. We repeat that twice, to make sure that we got some sort of response back to get the status of the actual recipient tracing in the hospital.

So in terms of those costs, the labor costs versus other costs, this is what it has been costing this particular center; something like $55 a component for the total, but $93 for the ones where the hospital destination has been assigned. So that is the current cost to a center, and probably that is fairly similar to most of our centers.

We have 17 centers conducting these programs. They are not uniformly computerized, and of course neither are the hospitals in terms of record tracing. Some of those records are sometimes found accidently. This particular center moved house a few years ago, and suddenly came across a box of records going back to 1986 or something like that.

Now in terms of the hospital side, I managed to get some figures about ten days ago from one of our major teaching hospitals as an example of what it costs them to follow-up the notification by the Red Cross, and how they went about it over the last four years.

This hospital received information from the Red Cross on 182 lookback cases of hepatitis C. Of those, 159 were deemed to have been transfused, and of those, 121 of the patients were deceased. Of the patients who were still alive which they could follow, there were 29. The patients who previously knew they were hepatitis C positive were 6. There were 9 new cases located after these 159 transfused patients; 4 with chronic infection; 5 with no chronic infection.

The age range here is considerably higher than what I alluded to in terms of the general national picture, but this is a tertiary care hospital, and there is probably some bias in this age package, with an overall mean here of 67 years, compared to something like 50 years with the other. So that is what they found in the hospital.

What it costs them to find those patients is listed here, and the breakdown of the costs in the hospital for finding those 9 patients. The tech time, again, it was alluded to before, we don't use causal people, causal employees to do this work. They have got to be hospital employees who are thoroughly knowledgeable, as the data arising out of their investigation is of significance with some potential litigation if they fail.

So these are the costs incurred at the hospital level for following-up those 159 transfusions. If you look at those costs in terms of per transfusion, of the 159 it is costing something $780 per transfusion, however, when the recipients are alive, there is further testing and counseling, which raises it to $855 per transfusion for each recipient who is alive.

If you divide the $124,000 by those 29, of course it is running around $4,000 to identify each live transfusion recipient, if you want to look at it that way. For the ones which are newly diagnosed, to identify a newly diagnosed hepatitis C patient, it is costing something like $13,784 per patient.

Superimposed on what we have been doing as a lookback, there have been new initiatives at the provincial ministry of health level, because Creaver also recommended hospital notification programs in quite a big way. Some provinces have already got class action lawsuits for hepatitis C taking place; the province of British Columbia, and the province of Ontario in particular.

So that the current atmosphere in the hospital sector is to do all they can do to notify transfused recipients. So there are hospital notification programs taking place. BC have almost completed their notification program. In their notification program, because health care is a provincial responsibility, the medical records' addresses of recipients, et cetera, can be accessed through the central provincial database on Medicare.

So that the process is ongoing now. The province of British Columbia, the province of Nova Scotia, the province of the New Brunswick are all into full fledged hospital-based notification program.

We alluded to the potential, when we had our committee meetings, of the pediatric hospitals in particular should consider a notification program. Several hospitals have done that across the country. There is a report of the costing and the kinds of efforts that are involved in such a program, which has just come out by Head L'Etale(?) in the Canadian Medical Association journal of July 15th, with an editorial commenting on the approach to hospital-based notification programs.

So in Canada we targeted lookback programs conducted by the Red Cross blood donor service. We have the hospital-based notification program superimposed on that, and somebody has got to sort out where these recipients are going to be assigned to in terms of who found them. That is what is ongoing right now. The material you got, I presented to you, is an interim sort of assessment of things as they stand at present.

Thank you.

DR. CAPLAN: Maybe just a couple of questions. That was the first talk I have heard in some time where someone couldn't say, but at least we have escaped the legal concerns of the United States.

PARTICIPANT: [Inaudible].

DR. GILL: The numbers shared from both the U.K. and Canada fit very closely with the model in terms of likelihood of the patient still being alive, and also the costs involved. They are not precisely the same, but appear to be relatively close.

PARTICIPANT: In addition, the number of positives, 67 percent is what I believe you are receiving. Is that what is assumed?

DR. GILL: It depends on what time period the donation falls in, the incident donation falls in, and whether we are dealing with version 1, 2 or 3 positivity. Generally close, however, the model that I used overall was anticipating a higher infectivity rate than what has been seen, but that was based on version 2 or 3.0 REBA positivity.

I believe that is what you are using in Canada. So possibly the model that I have been using has been overestimating the yield slightly.

PARTICIPANT: But the cost factor is much lower in Canada?

DR. GILL: No, the costs are relatively close; within 20-30 percent.

PARTICIPANT: I thought I saw much higher figures.

DR. GILL: No, I don't believe so. I mean I will have to look at this more closely, but I believe the cost per each one of the individual operations and the overall costs are fairly similar.

PARTICIPANT: [Inaudible].

Agenda Item: Open Committee Discussion

DR. CAPLAN: All right, thank you very much.

I picked up an article that Dr. Gill was kind enough to leave me from the Canadian medical journal from July, which I will also try and get copied and sent out to you.

One of the things that occurred to me while I was listening to the talks during the day is at the last meeting of the committee there was a request made that we get more information about a number of specific topics. My sentiment from this day is that will be the last time you will ever ask us for that again.

We have got a lot of information, and I know the hour is late and we are trying to process it and mull it over, but I would ask your forbearance that we do try to at least open the discussion today, because that will get our minds churning a little bit in the evening hours about things that we have to get back to tomorrow.

My proposal would be this, that we at this point take about an hour to talk some about some of the issues that we have heard, maybe using let's call it the Hoots-Penner statement as a template. I know that it was written with great care, and extreme precision at the last meeting. It is probably beyond revision, but in the off chance that we might want to add or subtract something from those first statements, maybe we could even put them up there, Paul, in the spirit of editing.

If that is all right with the group, why don't we maybe put those up there, but open the floor for some comments about things that you think ought to be included in the recommendations that this body wants to make. I would say you could base that both on what we heard today, and some of the reactions that we got from agencies that were written and distributed out.

So let me do that, open the floor up; see if there are any particular comments or themes that you would want to see us try to reflect in answering the questions we have been asked, and in making our recommendations.

One of the things that I don't think we said much about in the two statements that I recall is how would we like to see outreach coordinated with the medical community? This has struck me as something throughout the day that is crucial if we are going to go out and recommend some form of lookback, some form of discussion with people who might have been exposed, whether it was last week or ten years ago.

We have got to have a prepared medical profession, health care profession to deal with questions and have someone there to counsel. I would be curious to see what others think about that. Certainly this committee is in a position to ask the agencies and PHS to prepare some sort of plan or move forward in that regard.

We may want to say more than that, but it does strike me in listening to the experience in the U.K., in listening to some of the things that have come up as matters of legal concern, that we have got to have, if we do go in the direction of recommending some sort of lookback, even if at the level of just public service announcements, we are going to be stuck with the obligation of making sure we have got counseling, and an informed set of answers available. People expect that, and deservedly so.

PARTICIPANT: A question on policy. How specific and in how much detail would this committee be providing? I would propose that we would be giving the broad focus, and then from there it would be a matter of details being worked out by the FDA and others as to specifically how it would be carried out.

DR. CAPLAN: Well, it is fair to say -- and if Eric or Paul want to jump in -- as I understand this, as I looked at the committee's mandate, we have been asked to answer questions in the level of detail and specificity is really in our hands. We could get quite specific and say, A, B, C, D, E, F and do this, this and this.

It is also fair to say that some of the agencies are certain there, ready with expertise and knowledge about communicating risks or launching a public education campaign, and we could say look, the mandate falls to you to follow these general, broad guidelines about what we are trying to get to.

Personally, I think we are safer if we try to paint with a little bit of a broad stroke, and not get into too much of the detail about how to carry forward, but it is not -- my understanding is it is within our charge to get specific if that's what the committee wants to do.

PARTICIPANT: There is one very important issue, which is does the committee in premise accept the first statements of either one of these, which is that there is the responsibility to identify, using whatever means necessary, individuals who are HCV infected?

Because if we say yes to that, question, then immediately step way beyond lookback, and start I guess putting a moral onus on ourselves to make recommendations about how that is to be done, because we all agree I think, after seeing all this data, that lookback is only going to accomplish a smidgeon, even at the very best, of the total infected population in this country.

So therefore I think in terms of specificity versus generality, I think we have to solve that question first before we can go on. I think both drafts kind of take that in the spirit of things, but I think it is important, because we are going to get really -- the specific question was with regard to lookback, and we may have to get very specific about lookback.

There is then the question we have to decide it seems to me is do we also want to get rather specific about how else to identify individuals who are at risk for what we have seen to be a pretty bad disease?

PARTICIPANT: While I generally think that more information is better than less information, as you may have gleaned from my presentation, I would be much more comfortable endorsing a specific recommendation about a patient's right to know about potential hepatitis transmission in certain circumstances as we have been discussing today, than a broad statement about a patient's right to know about any infectious disease potential transmission that could have occurred through a past transfusion.

I would prefer that we focus on, even as a statement or principle, the hepatitis C, because different infectious disease risks carry with them different scenarios.

DR. CAPLAN: One of the things that is on the table as certainly part of the Hoots set of recommendations that we started out looking at last time is the notion that different periods of time in terms of when exposure occurred possibly through blood transfusion, might lead to different answers about obligations to inform, and obligations to try and look back.

One of the things I would also like to have some feedback about is whether people still believe that if you will, what we have to say might be time sensitive in terms of how we deal with individuals. Does it make sense to say, well, people have a right to know, or that people have a right to know relative to the time at which a particular test was available? Is that morally relevant or not in terms of what we have to tell someone if we couldn't detect a particular virus prior to a certain date?

I guess adding on to your comments, I'm kind of curious about what we think about the evolution of knowledge in this area. Certainly for me, one of the things that does become relevant is it is one thing to tell somebody, and feel that people have a right to know when they have been exposed to risks. If you know it, they should know it too.

Not all risks are equally nerve racking, and we might dispute about how much harm or quality of life impairment has to go along with that, but a second set of fact has changed that definitely makes a difference to me personally, and that is, what can be done?

There weren't interventions before; there are some interventions now. Some of them are lifestyle, and some of them might be interferon. Some of them might be information about even being a donor again and this sort of thing. So we have new knowledge that in some ways I think shapes my attitude about obligation to disclose that is in part based on patient benefit, recipient benefit if you will.

PARTICIPANT: I agree with you and others that the patient has the right to know. My concern, as I said earlier, is that what will happen is that we won't reach the patients if we pick one particular method. I think the Hoots proposal is time sensitive.

I think that reaching the patients prior to testing is going to require one approach, whereas after testing is going to require either one or two additional approaches in order to effectively reach. It doesn't matter whether we have intervention strategies or not, ultimately we will.

I think that we must reach those people at this point, but we really have to find a cost effective and encompassing way of doing that. That is really what I was talking about earlier.

PARTICIPANT: I just want to say I would really endorse what Art and Ron just said in terms that this is very -- and still is, but the period that we have been really focusing on has been a very dynamic period, where things change and are continuing to change. I like Ron's approach in the slides he put up in terms of trying to define some general time periods, or broad periods of time.

I guess I would say that perhaps we could really even stratify even in more finer detail, that period of 1990 to 1992, 1993, when we had first generation and second generation, with the availability of supplemental testing to really try and nail down with much more accuracy, who might truly be infected.

So I would really be supportive of at least the committee working through important points in that time period, and having a good understanding of what happened, and hence what information could be made available to people who might have been recipients in that period of time.

PARTICIPANT: If we put a date on something, does that then imply you really have to try to speak to everybody individually? Does it almost eliminate a general method? Because a general method of communication, public service announcements, et cetera, wouldn't really identify that date. It would just say, if you have had this happen, then you should go for testing, or if you had this sort of lifestyle.

I understand what you are saying, Mary, and it makes sense from the fact that the test is more definitive later, but I am just concerned that it boxes us in if we put a date on it.

PARTICIPANT: So we would be saying people who received product, who became positive in 1988, would not be notified, and if they found out inadvertently, and they asked you why weren't we notified, you would say well, we have a test, and tried to just draw the line there. I don't think that is going to float very well if someone wants to litigate.

Or even if you look at it rationally, it's drawing a line that is very arbitrary to begin with, and making a difference based on that arbitrary system of it would be less costly, number one, which doesn't work very well these days. Then secondly, that it was just because the testing was done. The individual is still involved, and has received the infectivity, and is proceeding to deal with it.

PARTICIPANT: I think maybe that's getting back to what I was saying before. I agree with you; I don't think you can do just that. I think if we can come up with a strategy that says that we think we can protect a person who was exposed in 1988, in a better fashion by another strategy besides straight lookback, then I think we are on both scientifically good ground, and perhaps even reasonable moral ground as well.

I don't think we can afford to just say, well, you were infected in 1988, and that was a difficult time, and therefore we have to just kind of forget about you. I think part of the reasons for the lines to be drawn was saying okay, if we are going to apply precious resources to trying to identify every foreseeable person, is this the strategy to do it with?

I think the lines were drawn to say that for people who were transfused in this period of time with these tests, there is a reasonable expectation that you could identify them if you used classic lookback methodology. If, however, they were infected in 1982, or maybe even 1988, perhaps the yield would be so low that the resources you applied to find that person may be better used by another methodology, be it PSA, be it direct physician education to seek his or her population of patients in for testing that could be at theoretical risk.

I think that was kind of the way the lines were drawn, not to say that we're kind of shunting away from the 1988 people or the 1982 people or whatever.

PARTICIPANT: The minute you separate -- in other words, you are discriminating. You have no really solid basis for the discrimination. If you use cost, that is not going to work, because the total cost situation from the HIV status was so tremendous, that I think we are still reeling from the fact that there were some things that we didn't do when we should have done it.

So I think at this point we are in need of being able to proceed to show the public that we are worried about the infectivity is.

PARTICIPANT: [Inaudible]. If you take on the idea that we have already discussed [inaudible], a lot of those records for 1982, 1983, possibly even 1988, may not be available. I think if you put together a strategy that says from this committee if we are going to go in that direction, that it does not seem prudent to include that time period in a targeted lookback, but rather public service announcements or whatever to reach that group, because it's not the cost.

I think if it was just an issue of cost, yes, it's [inaudible].

PARTICIPANT: But what if the records are available?

PARTICIPANT: If the records are available, I think the language that the -- the idea of putting something in place either from this committee or anything that comes from [inaudible], such a litigious situation, that they want to [inaudible]. I think they would be remiss or very stupid considering the times that we live in [inaudible].

PARTICIPANT: Although we may not like to admit it, I think costs are indeed an issue today. There are not unlimited resources available for health care. As Dr. Hoots said, it is not so much an issue of how much something costs, but where we can get the best yield for the resources that are available.

I would love it if Congress were to allocate $1 billion for this effort, and we could do everything for every patient, but that is not the case. We are going to have to be able to work probably within our current resources, that is, no new funds available for this endeavor, and we have to be sure that we reach as many patients as possible to get the message across.

If there were one patient transfused ten years ago who was going to come down with HCV, and we were out to find that one patient, and it was going to cost us $200 million to do it, we would say, we can't do it. That's too expensive to save one patient's life.

If there were a million patients in the last decade that we might help, and it would only cost us a dollar a person, we would say sure, let's go do it. We are somewhere in between those two extremes, and we don't have unlimited funds to do it. So I think we do need to direct our attention using different means for different time periods to get the message across in the best way.

DR. CAPLAN: One thing the committee may want to give some thought to before I turn the mike is the question -- I was trying to link this before, making a recommendation about lookback, trying to inform, how one gets information to people who may have been exposed. Do we want to say anything about the obligation, not only to counsel, but to provide the means to treatment?

If you are really looking at your budget, interferon for the uninsured, or interferon for the working poor with not much in the way of insurance, there are some real resource dilemmas there, and there will be some real problems if you tell somebody, well, we think you have this and this, but there is no money in the kitty to give you anything in the way of treatment.

So we may want to make a recommendation that if we are going to move in this area, we've got to move with a commitment to provide treatment access or not, but at least note it or say something about it.

PARTICIPANT: I agree that we need to focus on time periods. It is critical. The data we saw from Canada and Britain was all lookback predominantly to people who received unscreened products. There, there was substantial yield. In contrast, the data from Iowa was recipient lookback of sero converters, screened blood, zero yield.

So the Catch-22 is that the people where we will get yield, are the people who received unscreened blood prior to 1990, and yet the truth if there will be essentially zero yield to what's been proposed as ongoing prospective lookback fortunately. The numbers may be great, but the proportion of exposed, infected people from that early period is very small, not only due to absence of records, but due to the fact that we introduced surrogate tests in the late eighties; that we waited until now to even begin to move formally towards this process.

The other is that we did have a year or two of screening with first generation tests where there no confirmatory tests, and places do not have those samples to go back to. So the first year or two of screening, which culled the blood supply of all these donors who were transmitting, we can't go back and sort it out and effectively do it.

So when I look at it, we are left with a general notification campaign for the period of time when recipients were at substantial risk; 1 percent of all recipients per unit were getting infected prior to 1980, and even higher in earlier years, whereas now it is essentially trivial. The real issue is what do we do with those interval years, and where do we really draw the line on where to focus.

I think the message of the broad public health message should be to the period of time when there was substantial risk. If that message includes people getting transfused today, it is diluting out the message, because there is essentially zero risk today.

PARTICIPANT: I think everybody is talking about numbers here, and maybe we ought to focus back on people. If one of your family members had received blood in 1988, would you like to know whether there is a 50 percent chance that that individual might have come down with hepatitis C, and you would like to be directly notified?

If you look at numbers, I think you can play that game, but we already done the numbers with ALT many years ago, and we lost. [Inaudible] at the blood situation now and the blood business whole is very, very poor. I think the minute you begin to back off and start playing numbers games, then I think we are going to be still more liable than we have been in the past, saying that we are not very interested in the public. We're interested only in getting something that is going to be cost effective, and not going to create a lot of problems for us.

No, I think what we are really saying is how we can be effective in reaching segments of the population at different points in time. That is what I see as the real issue. To use certainly a family member, I can understand that, but if you don't reach them, they are never going to know. So we have to define strategies.

One strategy that is used in a particular time period doesn't preclude its use along with another strategy in another time period. All I'm saying is I think we should look at each time period. We should attempt to define it, and talk about the most effective way, number one, Art, of reaching the people.

I still say that treatment is secondary. You can't treat until you have defined who is or is not infected, whether they were infected by a transfusion or not. So treatment is really secondary. It's reaching the people, and in effect testing those individuals, or at least allowing them the right to be tested, but they have to know.

PARTICIPANT: I think what we are talking about is the cost risk here. Take your scenario. If you say to everyone who has a family member who was transfused in 1988, you need to be tested [inaudible]. It's a very expensive proposition, but it will be very effective if we get the message out to those individuals. Even though it may cost more dollars to do it, it may be a far better use of those dollars than it is to try to do lookback, and ignore the 90 percent that we'll never identify because the records are so screwy.

So I guess that's what I'm erring on. In fact, I'm not for a minute saying this is going to be a cheaper way to go. I think this is going to be a more expensive way to go, but I think it's going to be a more efficient way and a better use of dollars, even though it is going to cost more money to do it.

PARTICIPANT: One thing I think that I want to talk about is the public's trust here. I think the main thing we have to do is try and re-establish that trust, especially in the blood supply; not so much in the government as a whole, but in the blood supply.

I think if we can do this in a general way in which we do not leave anyone out, whether it is in a general statement or in a way of doing this through a lookback or however, I think that's one of the things we all have to remember, is that the public's trust in a sense has been violated. A lot of the citizens do not have the trust that they should have in committees such as this; in the blood supply as a whole based on things that have happened in the past.

Whatever we do, whether we are talking about the human factor or the cost involved, that has to be a part of it. We have to do this in such a way where people realize we acknowledge the mistakes that were done in the past, and we're trying our best now to include everyone, and not exclude anyone. I just think that that has to be a part of whatever we do.

PARTICIPANT: I think as I have been listening today, and listening to all the different perspectives, and the statistics and data that we have had, it is evident that this is a serious either epidemic, endemic, whatever we want to call it. I think we need to address it in a very respectful way.

I think as we look at lookback, to me I'm just seeing that we need to have probably a targeted lookback, as we have discussed. It sounds to me as though we need to fall down on some kind of a date, whether it was the second generation specificity test in which we want to use as that criteria of where we look. Do we go from that date forward? Then what do we do from that date backwards?

Do we address it with a public health service program to inform those prior to that date? What kind of program should that be? I'm seeing that that is probably the way that we need to go, but I don't think we can go in either direction until we kind of set a date on what we are going to use as our criteria for doing lookback. Then we can do notification and so on and so forth.

DR. CAPLAN: One thing that I hear coming across is if we divided our task in the right to know into identifying people, and then having them counseled properly, it kind of leads us to maybe different strategies about how we lookback.

One way to identify the pool of people is to do public service announcements, outreach, educational programs for primary care providers, health care providers and so on. Another question then comes up, well, we don't want to necessarily panic everyone who had a blood transfusion before 1988, and have them all in a tizzy, so can we do this in a way that assures maybe notification, but appropriate referral to someone who can then get a test done, and counsel them, and have education that is meaningful to them without causing undue concern?

Having said that, I guess we have a couple of issues that would lead us maybe to draw some lines. One is tests and when they became available, but I think what is really fueling our concern about looking back into the distant past is records, and the ability to find names and samples, to follow what Mike was talking about too. They are not there either after a certain time, or weren't collected in that way.

So it seems to me we might want to be forthright and say, look, it would be great if we could lookback and identify exactly who this was, but we may not be able to, given the state of our records. The next best strategies we have got are kind of general announcements, physician and health care provider education, and perhaps some form of special populations. I keep thinking that it can't be quite the same if we are dealing with hemophilia or sickle cell or Cooley's anemia to find those people. They are going to be easier to find than some other people in the general population.

So we may be able to acknowledge that we are talking about here is an inability to get the kind of records and information that would lead us to it. It's possible in other words to not just say, as records permit. We could say, well, we're looking back in this way, because it is our belief that we can't do anything else but to lookback in this way for this particular population prior to date.

The cut off date then become not when the test is introduced, it becomes the seven years or the eight years when people start hefting the records out to the basement.

PARTICIPANT: I agree very much. I think the only disagreement I have with Keith on this is if one has general statements that are floating around in the public domain that if you have been exposed, you have a blood transfusion in 1988 or 1987, you should go in and get this, how much different is that sort of a message than if you get a letter saying you have received blood in 1988. An individual who has subsequently been tested was found to be positive for hepatitis C. It may be that there is a chance that you will have this infectivity, and therefore we would recommend.

Now that is sort of push -- drives people in, whereas I think the general statements coming through the television and in the newspapers and so on have a lot less impact, because people never think it is them. That is just my opinion.

PARTICIPANT: I guess I'm moving somewhere in that same vein in that on any problem that is as complex as this, I guess I want to avoid the tendency to say let's answer one question. We have set that aside, and now let's answer a second question. The first question, date of lookback; second question, how do we then do this other type of notifying?

If we were voting right now, I would be very reluctant to raise my hand and say we only look back to date X period, and now let's talk about these PSAs. I'm going to say date X is fine with me. If I understand these types of announcements, just a PSA is not going to do it, and I think we all understand that.

So we're going to have to spend some time trying to be creative. Use people that understand how to attract people and the special audiences that we need to draw in, like you mentioned, Art, before that is going to make some sense to me. So these two questions in my mind, are very closely interacted.

PARTICIPANT: At the same time, I think that we have to certainly recognize the fact that this will take us only so far, because I think as we heard in several presentations today that suggests that there are other risk factors; the group of IV drug users, the group of individuals that find themselves in penal institutions.

We don't have enough information also about women in their childbearing years, and the possible vertical transmission to their infants. I think that as we begin to take information out to the public, we have to recognize that we can't just back and relax because we've done that, without at the same time trying to gather the other data that helps us to really understand the full impact of hepatitis C in a population.

PARTICIPANT: To me, if we are going to need to begin to talk dates, I think the introduction second generation screening test is the critical date, because subsequent to that time the risk of transfusions has declined to essentially zero. Prior to that time, there was a substantial risk, 1 in 1,000, and before that up to 1 percent or even higher as you go back in time.

So the message in any kind of general risk message has to be the blood prior to that date carried a substantial risk of HCV, whereas blood subsequent to that time did not. That then is sort of the public general notification focus, which I think does have to have an enhanced transfusion risk. It can't just be that little piece that the CDC puts this 5 percent of risks.

I think it's our obligation to bring the risk up to attention, and some of the data we saw today about increased progression potentially may even make it more important. But then if we focus on the pre-1992, pre-second generation screened blood for that, then what do we need to do subsequently? I guess that's where the ongoing sero conversion lookback becomes important, and part of the FDA overall armamentarium with respect to recipient risk for safe blood.

Today's blood is very safe, but one of the catches we have to catch the rare transmission is lookback from sero converting donors. So I guess I support doing that lookback, albeit it will be extremely low yield, back to the point of second generation screened blood, because we've got sort of two overlapping or contiguous methods; one for the pre-risky blood, and two, for the ongoing safe blood.

PARTICIPANT: Clearly, one of the things that has to be a part of what we do is an auditing and monitoring process to determine how effective that what we are doing really is. We know the number of transfusions that occurred in each time period. We can make an estimate or guesstimate of the number of people who are still alive. Then we can determine by whatever we decide to do, or what we feel should be done, how many people are we reaching?

If we are not reaching them with the technique or the method that we use, then that isn't effective. So we really have to have as part of this -- theoretically it would be nice to have some pilot projects out there, where we could go and take certain areas and do a small test to see how effective each of these concepts is, but we are going to have to monitor and audit to see if we really are effective. If we are not, we are going to have to come back to the table.

PARTICIPANT: One avenue in terms of identifying some people that might be at high risk other than IV drug users and people in prison would be people who have the need for transfusions on a regular basis or blood products on a regular basis. There are obviously national genetic support groups and things of that nature that we can use to help get that message out, because they are more in tune with these groups than we are as a whole.

They have daily contact or whatever with these groups, and they can help get that message out. So if we can tie this all together so that like I said, we don't exclude anyone per se, I don't have a problem saying let's start with the 1992 base in regards to the testing. I think that because of the amount of people prior to that who have had transfusions for whatever the reason, if we can include them, if we can at least try our best to identify as many of those groups as we can, and then enhance that by certain public general statements or something.

By general, I don't mean anything very simple to the point where people just gloss over it, but I mean there is more than one way to do this. I think we can identify certain groups in one way, and go after other groups in another way. So we'll just have to figure out a way to tie all that together.

PARTICIPANT: I agree. I think the other thing a number of us were talking about over lunch, I think we can kind of harken back to the early days of HIV before we were trying to exclude donors. In terms of identifying recipients there clearly, I think, need to be certain subsets of individuals identified in the population who may have received transfusions and not be aware of it, including patients who underwent procedures for major trauma [inaudible].

Their parents may be vaguely aware that they might have gotten blood. As a pediatric hematologist, I can tell you that comes up a lot in HIV. Did your child get transfused? Well, I can't remember. I think he did, but I'm not sure.

One of the other caveats that we can put into this is to say that not only if you received a transfusion you are at risk, but if you are in these groups, you may be at risk, and therefore you may want to consult with your physician, or with the physicians in the past to help you ascertain whether you may be at risk. If there is any question, be tested.

PARTICIPANT: Well, with family members I can just sit here and tell you about family members that I know of personally who have had transfusions that just aren't aware of the circumstances. So if we make a general statement and we get this out the right way, then there are going to be people getting on the phone and calling their aunt or their uncle or whoever, and say, I think I just heard something. I think you need to call your doctor and be tested.

There are different ways to do this, and we just have to explore all of them. I think that is one of them. If you make the message important, people will listen to it. They will call the people that they think might be a part of this. That's another way of getting the word out.

So there are a lot of different ways of doing this besides just a public announcement. We can do this. We are not going to reach everyone, let's just face that. We're not going to reach everyone, but I would like to say we are, but we are probably not.

The more people we reach the sooner, the better, because as we are sitting here talking, someone else is being infected. I think that at some point we are just going to have to get the best plan we can together, and at some point if we find ways to improve upon that plan, let's do so, but let's not stop the ball completely, because we all want the perfect plan. That is probably not going to happen.

I don't want to say let's waste the money either. I just think that there are enough ideas here that if we get this ball rolling, we will start getting in touch with people, and making people aware, and word of mouth will do a lot that we just can't do.

DR. CAPLAN: Tell me from the world of primary care, is there a question, did you receive a transfusion prior to 1990 or 1988 as a standard part of a good history? Is that done? So that could be done.

PARTICIPANT: Just like you do vaccinations.

PARTICIPANT: [Inaudible].

PARTICIPANT: It seems to me that part of the process for trying to get information like this out to these groups that will inform the public in ways that we have not always been so successful, that the academic training programs early on in this process, even before a formal statement is prepared, need to incorporate into the curriculum, into the material that is taught to students that all of these conditions have some very important population impacts.

So that it is more than just talking about the kind of [inaudible] in the context of a clinical syndrome, but rather, it does affect a lot of people. Part of the history gathering, and the interaction with an individual patient or the family or with the community is to better understand how some of these conditions really potentially affect a lot of people. [Inaudible], that perhaps allows us to do the kind of testing that informs the process.

PARTICIPANT: What I was going to say is that again, attempting to reach people, there is one clever idea that possibly could work. It would be two or three years away, and that is using the Census. If they would include in the Census in the year 2000, a question asking if people had been transfused or a member of the family.

Then literature would be mailed back to those individuals who indicated a yes, informing them of the possibility of having been exposed, depending on the time. I'm just trying to think of ways that this could be done, not necessarily now obviously.

PARTICIPANT: [Inaudible].

PARTICIPANT: We're back to separating notification versus specific versus general. I think you summarized it pretty well. Specific, we have limitations in trying to find them, as Carolyn was saying, because there aren't records. But there are records in some places, and up to maybe seven or eight years, but maybe more.

So one has I think, an obligation to pursue that as far as one can, because you can't discriminate and say there is a first class and a second class citizen. If the line is drawn here, and we're not going to bother notifying you, even though we have records.

Because you are liable to go back to that question I asked earlier, Freedom of Information and find out, well, these people knew something about me, and they decided that they could make decisions about my health, and I can't, because I don't have that information. The minute we begin to superimpose ourselves upon that sort of decision-making, I think we are in really big trouble.

The other is the general, which I think Keith is pushing, and others. There are many ways of getting at this problem of getting the information over to the public, so the public can respond, and see that it is significant, and see that they maybe should go in and get some testing done. That has to go to the medical profession and the general public.

We have lots of ways of doing that. I don't think that is going to be an issue. I think it is getting people to respond who have been getting commercials thrown at them on television about five seconds, and expect them to jump up and do something. People just don't do it. That's why I like the idea of having notification for as far as we can on a specific level, enough though it's costly.

Now I have run a Red Cross center for 15 years. I know what it is like to do the HIV lookback. I have done it with a lot of griping from the staff. I also sit on a blood banking board at two of our local hospitals, so I know the griping that will come when they have to start digging through, but they can do it, and it has been done for the HIV. It is a little expensive, and it will probably end up costing us a little higher for the blood.

I remember about ten years ago someone saying, boy, before you know it -- and I think Jim, you were at that meeting -- blood is going to cost $100 a pint, and we all laughed. Then it was about $42. Now it is about $90. So it probably going to be like the cigarette situation, they'll just pass it on.

We'll end up all paying for it, and it will $120 or $150 a unit, but to know that that unit is safe will be worth the extra $50 in my mind, as opposed to trying to get by with a lesser amount.

PARTICIPANT: I just wanted to seek a clarification from Dr. Penner, and sort of going back to Mike's comments and others have made about trying to draw a line in the sand. Is the starting point for your doing targeted lookback as long as records permit, the donor who has generation two EIA with a supplemental, to somebody as best we can tell is truly infected? Is that your starting point?

DR. PENNER: No, I think we will be even be using that 1990, where 50 percent are false positives. We would be going back from 1990.

PARTICIPANT: So then the message is a little less clear, because you are going back to people and saying, I don't know whether you got blood. There is a 50-50 chance your donor was infected, as opposed to people in 1992, where you could have an increased level of certainty.

PARTICIPANT: Because it comes back to my mind of saying, but I would rather hear that I have been exposed to something with a 50 percent chance of giving me a problem, or not hear about it at all. So I think that is the kind of issue that we are making decisions on for our public.

PARTICIPANT: In addition to cost and yield, I think we should remember Dr. Kaplan's comments today about competing risks. The notification process is not without some inherent risks for the recipient of that information. The recipient who may not want to know that they have been exposed, but beyond that, the burden of false positivity. The fact that even with a negative result, the recipient cannot be assured that he or she does not have hepatitis C.

So I think we should approach the targeted lookback effort with the best information that we have. I don't feel that we are being irresponsible in our medical duties to say to the public, to say to individual patients, we decided to take this process from point A onward, and not further back, because that was the best information that we had in order to base our system.

PARTICIPANT: But you are going to be saying the same thing to the individual now, that perhaps you will not have been infected by that product, even though it is the latest testing, and we're sure it is positive. You don't know, so you are still getting the same message that individual receives the blood, you may be positive or you may be negative, but we don't know until we test you. So I don't see any distinction there, any difference.

PARTICIPANT: On a probability.

PARTICIPANT: Well, I think that is true. The probability that you are infected with the new testing is going to be much higher. I would like to reiterate again, I spent the last ten years developing a product to treat this disease. You create a patient when you tell them that they are HCV RNA positive, when you tell them they are C positive.

I think it is all fine to go back and identify these people, but I think we need to think about what we are going to do with them once they get that letter, because there is a lot of misinformation in the medical community about how you deal with this patient.

They may go to their local physician, and they will be dealt with inappropriately. It will cost them a lot of money. They will become very angry. They will become very confused. They may not get diagnosed. So I think as another part of this, I would like to see us think about once we create these patients by telling them they are positive, what we are going to do with them.

I'm having anxieties about creating patients based on some of that false positivity that we were seeing with those early assays. I have seen a lot of those patients come in to clinical trials thinking they are positive, and then when we really get a close at them, they are not.

So I think we need to think about those things, because I really have a lot of concern, because once a patient thinks he is positive, he begins to think like a patient. If you look at some of the quality of life data that I have looked at in baseline HCV patients, they change their demeanor. They change their psychological make-up if you will, once they find out they are chronically infected.

So I think we have to be a little careful about who we identify; how sure we are when we tell them they are positive. Then we at least give them some methodology on who to get to and how, so they don't end up in the wrong place, and end up with another misdiagnosis, and a lot of money spent based on a letter we sent them.

I think it's important to be able to identify them, but I think we have to be able to handle them, and be pretty sure when we identify them that they are real patients.

PARTICIPANT: [Inaudible].

PARTICIPANT: I'm trying to think back to a Jim-type model in terms of cost. Before we started screening, after we introduced ALT and core testing, we were talking about 3 per 100 risk, is that right for HCV?

PARTICIPANT: That is probably a little high. I'd say the general population prevalence of 1.8 percent is probably on the high side, probably pre- anti-core. Certainly post-HCV exclusions of 1985, I would say you're probably around 1 percent. Once we added anti-core and ALT, once we started prospective screening in 1990, we were dealing with about 0.4 percent, 4 per 1,000 infected donors found is the risk.

PARTICIPANT: When you are talking about trying to educate the population to what is high risk, the concept in the article on cancer risk in here is really a lot with how you frame it. If you are talking about a disease like HCV, where the potential outcome is severe a significant percentage of the time. Even with a 1-2 percent risk, that is a very high risk.

Then you get a question, well, if you cut your losses in terms of not having to test people who were really not at risk by not doing lookback or by not testing everybody, do you have cost savings over the long haul? That's what I'm having trouble dealing with.

It seems to me that with this high a risk, that if you take the ones who are in that treatment range period, that testing everyone will cost a little bit more, but the yield will be extraordinary. The miss rate will be extraordinarily low. That may counterbalance the increased costs, which is what I was saying before.

So rather than going through the extra cost of tracking those in that period from say from 1988 to 1992, why not just advocate there being tested? I guess that's what I'm saying. I obviously haven't worked out the numbers. Do you have a comment about that? I don't know.

It just seems to me that when we see how much it costs to do a lookback anyway, why not just test all those individuals, or recommend that they all be tested, and try to get them in to be tested?

PARTICIPANT: But you are still going to then be separating groups out and saying some are going to be notified this way, and others are not simply because we have decided that we can use the money more effectively this way than that way. When you begin to discriminate, then I think you have set yourself up for some legal action.

PARTICIPANT: Except that if your yield is extraordinarily high -- there is a remote chance that you will miss that person because they didn't come in to be tested, but it seems to me that for the population at large, that is a risk that is well worth taking, because you are going to identify so many more.

Are we going to make decisions based purely on legal grounds, or are we going to do what is right for the United States' population?

PARTICIPANT: I certainly don't want to miss your point, John, because that is very critical, but I am also thinking of something Larry said a little bit earlier. There is a trust factor here. The type of points that Keith is raising really to me says what is this message or messages that are going to be sent out there, and how are they going to be delivered?

I think that I'll just call it a PSA for now that we are talking about in the abstraction is an incredibly complex idea. It scares me on one hand to say, we're just going to put out that general message, because I can see that general message causing more problems than it is going to create. Yet if it is done well, it could be as Keith said earlier, much more expensive, but the yield could be better. Under those type of conditions, if we could work that out, then I am comfortable with it. If we can't work that out, then I'm coming back to where you are at.

PARTICIPANT: Decisions whether or not to implement a lookback program don't always depend on the sensitivity and specificity of the test among others; the probability of false positives. No one is considering, no one has ever used ALT or anti-core results to identify a lookback situation, because we know the yield would be incredibly low, and we would be chasing our tail.

First generation anti-HCV testing, well, it was better than the two surrogate tests. It wasn't all that much better. Essentially what we are trying to do I think, is decide how good a test has to be before we think that the yield will be good enough to engage in the effort. Clearly, the surrogate test did not have a good enough predictive value to be used for lookback.

Anti-HCV testing probably does, but the question of what version, when in that spectrum do you say enough is enough? I think there are some differences of opinion.

PARTICIPANT: Fifty percent isn't bad. I would take those kind of odds.

DR. CAPLAN: One of the things I can do for a brief respite from turning this ground over is Paul reminded me -- you will love this non sequitur -- that we have a legal obligation or a duty or some clear cut requirement upon us to approve the minutes of the last meeting. So why don't I entertain a motion to approve them, and then we will do that.

[The motion is duly made and seconded.]

All in favor? Opposed.

[The minutes were unanimously approved as submitted.]

There you are.

It seems to me one of the things we could do is perhaps try to summarize a bit some of the things that have come up. We are actually making some progress not toward necessarily consensus, but toward picking out some things that I think we will have to hash at yet again tomorrow, one of them being how do we insure coordination between looking back or notifying, and making sure that there is someone there with counseling and an appropriate set of answers when people come with questions.

Another is whether we are going to be time specific and why, which is to say is it due to the specificity and accuracy of the test that drives time sensitivity about looking backwards, or is it due to recordkeeping, or is it due to the ability to perhaps pick out subpopulations and get further, in terms of feasibility in what is doable with them?

Does it make any sense in terms of yield to simply go back and recommend testing? I can't help but think that one of the things that we're aware of is that there is supposed to be a bigger push for hepatitis C, however it is caused anyway. That is going on, and to some extent some of the PSAs or whatever they wind up looking like should be tied into this general effort to alert the American people about this 4 million plus problem, epidemic of hepatitis C.

So that is there, and that is in the background. It is not just inventing the blood transfusion fueled communication problem, but that is out there. How does that piggy-back in I guess is what I'm trying to say in English? How do the two link up?

We certainly heard some comments too just in this discussion about the importance of monitoring and auditing, and being ready to step back and say, well, even though we recommended X, we may want to change direction here in terms of where things are going, and that seems sage advice.

We heard people also mention that we've got to be sensitive in the recommendation to making sure that people's attention is commanded, but we don't scare folks out of their wits. That is always a fun enterprise, although I have to say that if Americans are good at anything, it is probably the ability to do exactly that in an advertising/communication mode. It just costs something.

The last general point that I picked up is that we need to make sure too that we are sending out consistent, uniform messages, whatever they are, so that people come to have trust and faith that the blood supply not only is safe, but the people who operate it are being square with them.

It makes me think that in our recommendations we need to have a few words about why it is that when there was no lookback, there suddenly is lookback. If I was being wary of experts and pointy heads in Washington or Bethesda, I guess I would want to know why are they all of the sudden deciding to change something? I think we can answer that, but we better, otherwise we are not going to have trust cemented that way. We may stir up distrust inadvertently.

So we do need to say whether it's because we think treatment modalities have shifted, or we understand more about the danger of the disease than we did perhaps in 1988, or whatever we want to say, but we better say something, otherwise we are going to look like we are just back filling, and that is not a good position to be in.

I'm going to try in the privacy of the wee hours to capture some of those points. I was taking some notes, and maybe I'll trot them forward tomorrow. I was trying to summarize a little bit also to get you thinking along those lines.

If I missed something, or there is some other important issue that you think -- or recommendations -- I ought to capture, then I wish you would take a little time maybe to jot down some notes, or to write even a sentence or two about what you think that ought to look like. My summary was just meant as a jog. It wasn't supposed to be comprehensive or definitive.

So if there are other things there that you think ought to be there, please spend a little time between now and eight o'clock in the morning tomorrow, all your waking hours, trying to put some of those on paper so we can make sure to capture that.

What we are going to do is we will reconvene at eight. We have got brief presentations by Mike Busch and Ed on communication of risk, which will undoubtedly undermine everything we've done now. We'll be cast back into the sea of despair.

Then we'll get right back to this, but our mission tomorrow, instead of kicking around just general themes, will be to really true and focus in on that language. I'm going to have Paul or somebody up there with that pen probably writing on the overheads pretty frequently to see if we can get some consensus.

So I hope this gets us to a place where we are ready to do some hard thinking and do some hard work, and maybe push something out by after lunch tomorrow that we can all sign off on.

PARTICIPANT: I have been considering for a while today pointing out to the group that the Heart, Lung and Blood Institute has had several national education programs, some of which some of you may be familiar with. There is a cholesterol education program, a high blood pressure education program, and fairly recently an asthma education program.

It is conceivable that some of the expertise that has been used in the Institute to set up these programs, could work with other parts of the Public Health Service in setting up a somewhat similar program. There was also a National Blood Resources Education Program. Each of these has a professional and a lay educational component to them. I suspect that Heart, Lung and Blood would not be the likely lead agency in hepatitis, since our part of the operation is relatively small.

It is conceivable either as a result of what is mentioned and discussed here, and what is posed for the future, that we could find out whether the Institute could lend expertise to setting up a somewhat similar program. I honestly don't know how effective they have been as far as measurement of their ultimate goal, but I sure hear a lot about some of them on the radio.

DR. CAPLAN: Before I break, I'm going to ask if anyone either on the committee or in the audience wants to offer any final comments or thoughts?

PARTICIPANT: [Inaudible]. We have to be very careful as we identify individuals as HCV positive in terms of their jobs, in terms of where they live, in terms of family issues and all of that. I think we need to [inaudible].

The other is just more for information. Has any research been going on that perhaps Paul could show the committee related to any prototype HCV vaccines that are under study?

PARTICIPANT: There was a conference on the main campus here. [Inaudible].

DR. CAPLAN: One reason sometimes to lookback with as much precision as you can, and we have said this a few times but I forgot it in my summary, is so that you are able to move forward should the situation change. So one reason for more precision in going back, even it if is hard to do it, is that if you get a vaccine or some situation changes on the therapy side, you are ready to move. That may be the justification for the ongoing perspective lookbacks even now at the small yield rates, or to pick up the PRC undetected that Ed was talking about.

That is still a good point, stigma. We've got to pay attention to that. It has certainly come up in many other areas where pre-existing conditions might disqualify somebody.

The ability to then say, well, we've got information that could benefit you; because we did it now, we can use it soon.

Any other comments? All right then, we will meet here at eight o'clock.

[Whereupon the meeting was recessed at 5:30 p.m., to reconvene the following morning, Tuesday, August 12, 1997, at 8:00 a.m.]