Regulations and Recall
The Food and Drug Administration (FDA) is the principal regulatory agency with respect to blood and blood products, but it exercises its authority largely through informal action. For example, recall—the removal of a product from the market—exemplifies the relationship between the FDA's potent formal powers and its informal modus operandi. Recall is a voluntary act undertaken by the manufacturer but overseen by the FDA, which has authority to seize or delicense a product. Because the FDA's resources are quite limited, and because the regulation of blood and blood products is based on scientific consensus (much in the same way as the establishment of scientific facts generally), the agency relies upon the blood industry and others for advice. The FDA's Blood Products Advisory Committee (BPAC) is a venue for consensus building about blood regulatory policy. In an industry in which company and product reputation is critical to market success, the FDA's collegial and informed approach is usually effective.
This chapter presents an analysis of the FDA regulation of blood and blood products during the study period 1982-1986, when the HIV virus began to threaten the integrity of the blood supply and before tests and procedures were established and disseminated sufficiently to control, if not virtually eliminate, that threat. The analysis extends to consideration of a policy concerning the identification and notification of persons who may have received contaminated blood or blood products that was not formally put in place until 1991. In carrying out the analysis, the Committee examined the FDA's exercise of its regulatory powers and the actions it took during four critical events. For each of these events, the Committee posed a series of five hypotheses to explain the
FDA's actions. These hypotheses focused consideration on (1) the reach of the agency's legal powers, (2) the information available at the time in relation to relevant public health considerations, (3) the FDA's resources, (4) the FDA's institutional culture, and (5) the economic costs of particular actions and the prevailing political climate.
A later section of the chapter includes a short discussion of other federal agencies and of private organizations involved in the regulatory process. The section immediately below introduces the framework of analysis: critical events, related regulatory policy questions, and explanatory hypotheses.
FRAMEWORK OF ANALYSIS
The history of the threat that HIV infection posed to the safety of blood and blood products, and the eventual suppression of that threat through the combination of private and public actions, can be told from numerous perspectives. Any one perspective will to some degree falsify and distort a multifaceted narrative. This chapter structures the story in terms of the Committee's interest in the appropriate development of FDA recall and lookback (recipient tracing) policies. The Committee highlights a series of "choice points" (either events or nonevents) that were presented to the FDA as increments of information about AIDS as a disease—and about possible means of preventing its spread through blood and blood products—became available. 1 In particular we consider the following critical events and questions discussed in the following sections.
Critical Event 1
On March 24, 1983, the FDA issued letters to all blood banks, plasma centers, and plasma fractionators requiring particular practices related to donor
screening and the segregation of high-risk plasma supplies (Petricciani 1983a,b, c). Further regulatory action was possible at this time, although not necessarily prudent. The Committee sought to understand why, for example, manufacturers were not asked to withdraw and/or recall all existing products derived from donors who were not screened for AIDS on an orderly basis (i.e., on a schedule that would not seriously have affected the availability of blood and blood products)?
Critical Event 2
On July 19, 1983, the Blood Products Advisory Committee (BPAC) recommended to the FDA that recall of AHF concentrate and other plasma products not be "automatic" when those products had been linked to an individual donor that had been identified as having or as suspected of having AIDS (FDA, BPAC 1983). The FDA accepted this recommendation and, shortly thereafter, adopted a policy that a recall decision regarding AHF produced from a plasma pool that included material from a donor who was later found to have AIDS should be evaluated on a case-by-case basis, which would take into account for each suspected donor the accuracy of the diagnosis, the occurrence of the symptoms in relation to date of donation, and the impact of the recall on the supply of AHF (Novitch 1983). Was the FDA's acceptance of the BPAC recommendation a well-considered judgment? Even if it were, was the case-by-case recall policy sufficiently well specified and supported that it could be implemented effectively by the industry participants involved? Why was the policy applied only to AHF concentrate and not to cryoprecipitate, which when used can involve multiple infusions of single donor units that are stored before use?
Critical Event 3
Heat-treated AHF became available in 1983 (Persky pers. com. 1995), and by 1985 heat treatment had been accepted as effective in inactivating HIV (McDougal, et al. 1985). Why did the FDA wait until 1989 to require the recall and destruction of all untreated units?
Critical Event 4
By 1985, there was wide consensus that transfusion of HIV-infected blood products led to HIV infection in the transfusion recipient (Goedert and Gallo 1985; Curran, et al. 1984). Recipients became infective immediately but
remained asymptomatic for approximately five years (Bove 1986) during which time their intimate contacts were at risk of acquiring HIV. In 1988, a presidential directive to the Department of Health and Human Services was issued on blood safety based on the report of the Presidential Commission on the Human Immunodeficiency Virus. The Department was instructed to formulate a policy for tracing recipients of possibly infected blood products ("lookback") and informing them of potential risks. However, the FDA did not issue recommendations until September 1991 (FDA, BPAC 1991a,b). Why were a presidential directive and six years of agency development (1985-1991) necessary to put such a policy into effect?
FDA Regulatory Authority and Practice
Since 1972, the FDA has been the principal regulatory agency with respect to blood and blood products. The FDA exercises its regulatory authority in an environment in which other agencies play important research, public information, and coordination roles. The agency clearly is dependent upon the information generated by these other federal agencies and by other public and private organizations. Nevertheless, the FDA's statutory regulatory authority is very extensive [federal Food, Drug and Cosmetics Act and the Public Health Services Act, codified at 42 U.S.C. § 262] (Hutt and Merrill 1991).
The FDA licenses all collectors and producers of blood and blood products; licenses those products themselves; and prescribes general regulations applicable to the production, testing, labeling, recordkeeping, and most other aspects of the blood collection, production, and distribution process. It is authorized to enforce its requirements in court by seizure, injunction, or criminal prosecution, and administratively by suspending or revoking the license of a producer of a blood product.
Interestingly enough, the formal powers of the FDA do not include mandatory recalls of products, although recall decisions loom large in the analysis in this chapter. Technically, all recalls of blood and blood products are voluntary [21 C.F.R. Subpart A and Subpart B, 1992]. The FDA has such broad authority under Section 262 of the Public Health Services Act that it could probably create a mandatory recall process by regulation (Falter 1994; FDA 1988). However, it has never adopted regulations of this type. Most recalls are manufacturer initiated, although some are undertaken at the urging of the FDA. Manufacturers recall their products in order to protect the public health, protect their own reputations, avoid civil litigation, and avoid the utilization of the FDA's formal powers of seizure, injunction, criminal prosecution, license suspension, or license revocation. Recalls are thus undertaken "in the shadow" of the FDA's formal powers. Moreover, the FDA has regulations and guidelines
concerning how recalls will be conducted. Those documents describe a process in which manufacturers must notify the agency of all manufacturer-initiated recalls. The FDA works with manufacturers in every recall to determine the breadth and depth of the recall activity that is required [21 C.F.R. Subpart A and Subpart B, 1992] (FDA 1988).
The legal informality of the FDA recall process is emblematic of the agency's general approach to regulation. This informality is not necessarily evidence of a weak regulatory posture. Indeed, informal action by the FDA will often be more effective than formal action. The agency owes its ability to act informally in part to its extremely powerful statutory authority. An FDA action to suspend the license of a producer of blood products, for example, has the immediate effect of shutting down the manufacturer's business. Suspending a product license or instituting a proceeding to seize a product already distributed will almost surely destroy the market for that product. Given these formal possibilities, the FDA can expect significant cooperation from industry when it suggests or recommends changes in existing practices.
In addition, the plasma fractionation industry, like the pharmaceutical and food industries, is highly dependent upon the retention of public confidence. Reputational effects loom large and manufacturers generally have strong incentives to self-police and to be seen as "good citizens" who comply with all regulatory criteria. Manufacturers must also consider the costs of civil litigation should a violation of FDA requirements cause harm to persons using their products. It is perhaps worth noting that so-called "blood shield laws" protect those who supply certain blood products, such as whole blood, from civil responsibility for non-negligent injuries (see Chapter 2).
Perhaps most importantly, the regulation of blood and blood products is a scientific enterprise that operates within the broader norms of the scientific and medical communities. Those norms include broad-based cooperation in advancing scientific knowledge. Just as the establishment of most scientific fact is based on the development of a consensus view, so the appropriateness of regulatory action in relation to blood and blood products often operates by consensus. To the extent that the FDA can establish a collegial consensus on the state of scientific knowledge and on the appropriate actions to be taken to protect the public health given that knowledge, it can be assured both that its regulatory actions will be perceived to be well-considered and that they will be largely self-enforcing.
Effective, informal regulatory techniques also economize mightily on agency resources. Use of formal regulatory powers is often an extraordinarily costly and time-consuming process in the American legal system. That system's commitments to fairness, accuracy in the determination of facts, public participation, and political and legal accountability can make even the smallest, formal, legal task a bureaucratic nightmare.
Like any regulatory posture, the advantages of the collegial or cooperative approach to regulation that has characterized the FDA's regulation of the blood supply and blood products involve attendant disadvantages (Dubinsky, Meyer, Quinnan interviews). Collegiality can easily merge into coziness and lassitude. The perceived need to secure future cooperation can deter the timely exercise of power. The agency can become too dependent on regulated parties for voluntary submission of the information necessary for the agency's actions, as well as on industry self-regulation, as the first line of defense of the public health. Informal approaches may also have differential effects in a heterogeneous industry that includes large, centralized, and well-informed members as well as small, dispersed, and relatively under-informed participants.
Finally, consensus approaches may be inhibiting in the arena of regulating blood and blood products. These products are often lifesaving when administered, are critically necessary to certain groups of individuals (e.g., hemophiliacs), and are almost constantly in limited supply. Historically the "consensus" in the blood industry and those dependent upon its products has been to err in favor of maintaining product availability (FDA, BPAC 1983; Aledort, Rodell interviews).
Significant political will would be required, therefore, to take independent formal action given these long-standing informal understandings. In the primary period under study there is reason to believe that such political will would have been difficult to muster at the FDA. Not only was the general regulatory climate deregulatory in tone, the Reagan administration was slow to highlight AIDS as a public health crisis and the Centers for Disease Control and the FDA's Bureau of Biologics were still recovering from criticism that it had behaved much too aggressively in pursuing the swine flu vaccine.
Many of the questions that have been raised about the FDA's activities concerning the protection of the blood supply and blood products from contamination by HIV relate to the informality and tardiness of the FDA's regulatory actions (Donohue, Evatt, Feldman, Lipton, Rodell interviews). But even with 20/20 hindsight it is often difficult to determine whether the FDA utilized its powers—both formal and informal—appropriately. Evaluating whether actions were timely or appropriately formal depends upon recreating the historical context of those actions, including the knowledge base available to the FDA and the contemporaneous actions of private parties that would make public action appear either necessary or unnecessary under the circumstances. Because confident re-creation of the contemporaneous context is often impossible, and because interpretation of events in the light of subsequent knowledge might be misleading and unfair, the Committee has often been led to conclude that while actions may not have been optimal, they cannot be said to have been unreasonable.
Nevertheless, in retrospect we can learn lessons that suggest changes to the blood regulatory system that might produce better outcomes in relation to some
future threat. In other words, while the Committee is uncertain how many regulatory decisions would have been specifically altered by changes in regulatory approach, it is reasonably confident that certain process changes would have given the FDA a better chance in general of producing improved public health results.
The Committee developed operating hypotheses that attempt to take account of the legal authority of relevant actors, the information available in particular time periods, the countervailing public health considerations that bore on particular decisions, the resource limitations that constrained particular actors, the institutional culture and structure within which decisions were made, and the economic and political incentives that bore on decisionmaking. Our general working hypotheses are captured in the following five explanatory propositions that—at least in theory—can be "applied to" or "tested against" particular regulatory actions or inactions:
Available information had not reached the necessary threshold for action given the legal authority of the relevant actor(s).
Information was legally sufficient for a relevant action, but that action was not clearly warranted by the facts given countervailing public health considerations.
Action was legally warranted and justifiable in the interest of public health, but was delayed or aborted by resource limitations.
Action was legally warranted and considered to be justifiable in the interest of public health, but was delayed or aborted because of the existing institutional structures, relationships, or standard operating procedures within which action was taken.
Action was legally warranted and considered to be justifiable in the interest of public health, but was delayed or aborted because of fears of either the economic or the political consequences of taking the action in question.
Each one of these general hypotheses has been restructured in specific terms in relation to the critical events scrutinized. These hypotheses are not necessarily exhaustive and are not mutually exclusive. Moreover, because questions concerning legal authority and public health policy (Hypotheses 1 and 2) are more easily analyzed than the indirect effects of resource constraints, standard operating procedure, and economic or political concerns (Hypotheses 3, 4, and 5), these latter hypotheses are often somewhat underdeveloped. The Committee often concluded that the influence of the factors that Hypotheses 3 to 5 test
cannot be ruled out—indeed these factors are ever-present aspects of human decisionmaking, but that should not be confused with a finding that these factors have been shown to have had a determinate impact on the specific decisions analyzed.
FINDINGS AND CONCLUSIONS
FDA Letters of March 1983
On March 24, 1983, Joseph Petricciani, Director of the FDA Bureau of Biologics, issued three letters recommending steps to take to reduce the risk of transmission of AIDS through blood and blood products. These letters went to whole blood collection centers, plasma collection centers and plasma fractionators. Common to all these letters is a request that measures be taken to institute appropriate donor screening practices and procedures for individual donors and donor groups known to be at increased risk for transmitting AIDS (Petricciani, 1983a,b,c). (See also Chapter 3, Chapter 5, and Appendix D).
The letters sent by Dr. Petricciani (1983a,b,c) are interesting from three different perspectives. The letters recommend slightly different requirements. The requirements recognize the differing positions of whole blood collection centers, plasma collection centers and plasma fractionators in the collection and distribution of blood and blood products. The letter to plasma fractionators seems to have imposed a higher standard of donor education on source plasma collectors than for blood banks that collect blood for transfusion. For example, both blood banks and plasma collection centers were instructed to revise their standard operating procedures to include the quarantine and disposal of blood and/or plasma collected from a donor known or suspected to have AIDS (i.e., an individual with a history of night sweats, unexplained fevers, unexpected weight loss, or signs of lymphadenopathy or Kaposi's sarcoma). In addition to these requirements, plasma collection centers were required to label each unit of plasma if it was collected from a high-risk donor, to examine donors for lymphadenopathy, and to record the donor's weight prior to each donation. The American Blood Resources Association (ABRA) complained of this disparity and questioned its scientific and public health rationale (Reilly 1983). ABRA suggested that the plasma center protocol be applied to blood banks as well. Dr. Petricciani responded (on August 22, 1983) by saying only that there was concern that more demanding requirements might adversely affect the supply of whole blood.
The letters did not specify what was to be done about plasma or derivatives that had been collected from unscreened donors (i.e., individuals who had not been specifically questioned about early signs and symptoms of AIDS or whether
they were members of high-risk groups) and had not yet been used. None of the letters required, by its terms, that collection centers or plasma fractionators segregate or destroy products collected or manufactured prior to the date of the letters. The Committee believes, however, that collection centers or plasma fractionators could have interpreted these recommendations as requiring that they withdraw or at least segregate all products that they had collected from unscreened donors and were currently in their stocks. Because these letters asserted that they were based upon a consensus of the collection centers and the plasma fractionators, the Committee believes that the letters suggested that these practices were ongoing at most, if not all, blood collection centers and blood products manufacturers at the time that the letters were issued.
The Committee has not been able to determine definitively what the intended legal effect of these letters was or precisely how the letters were understood by their recipients. There are reports, both in testimony prepared for this Committee (McAuley interview; Gury 1982) and in the summary of the Public Health Service's Workshop to Identify Opportunities for Prevention of Acquired Immune Deficiency Syndrome (1983), suggesting that blood and plasma collection centers were engaging in some types of donor screening activities as early as December 1982 (Silvergleid interview; Gury 1982; see also Chapter 5). On December 17, 1982, Alpha Therapeutics began excluding the use of plasma from all donors who had been in Haiti, had used IV drugs, or, if male, had sexual contact with another man (Gury 1982). Alpha Therapeutics also adopted a policy effective December 26, 1982, that its affiliates should no longer send it any unscreened plasma (Gury 1982). There are, in addition, indications from the National Hemophilia Foundation's (NHF) Medical and Scientific Advisory Council (MASAC) (NHF 1983) and from recommendations of the ABRA issued on January 28, 1983, that most plasma fractionators and collection centers were moving toward donor screening and donor education programs in January 1983. Hence, it is impossible to know to what extent the stock of plasma or whole blood present in March 1983 contained blood or plasma that had not been subjected at least to some forms of donor screening.
On the other hand, donor screening practices may have varied substantially up to this point and may have been routine only in some organizations. Nor is there evidence that plasma fractionators were engaging in precisely the segregation and selective processing activities suggested by the Petricciani letters (1983a,b,c). Hence, it seems reasonable to suppose that there were substantial stocks of products in plasma fractionators' inventories and elsewhere that had not been screened in accordance with the FDA letters' directions.
The March 24, 1983, letters clearly did not impose a requirement that collection centers or plasma fractionators recall products that had not been collected or produced in accordance with the FDA's recommendations. The recall record that has been compiled by the Committee does not suggest that there was a rush to recall products following receipt of the FDA letters (Petricciani 1983a,b,c). Therefore, it seems unlikely that plasma fractionators interpreted these letters as requiring recalls. This in turn raises the question of why the FDA did not require a recall of unscreened and unsegregated (or selectively manufactured) products at this time.
The Committee failed to reject the first hypothesis; that the FDA lacked the legal authority or the resources to demand a recall seems unlikely. The FDA seems to take the position that it can request or require (by threat of seizure or of other action) a recall in any circumstance in which blood or a blood product fails to meet existing requirements or poses a risk to the public health [21 C.F.R. §7.11] (FDA 1988, 1994). The latter criterion seems clearly to fit this case and the former may as well. The Petricciani letters could be viewed as an official recognition of the existence of a good manufacturing practice rather than as a new requirement. And the FDA has taken the position (Falter 1994; Parkman, Quinnan interviews) that it can enforce good manufacturing practices that are or have become the industry standard even though they are not explicitly codified in any FDA issuances.
It seems unlikely that the FDA was influenced not to engage in recall activity because of resource constraints. Recalls put virtually all of the resource requirements on collectors or plasma fractionators. The FDA must monitor this process, but there is no suggestion that it did not have the resources to do that. (In the face of strong manufacturer resistance, the FDA would have had to engage in formal action, possibly a seizure. This could have been enormously resource intensive and probably could not have been carried through if attempted. But "massive resistance" is virtually unknown in this industry.)
On the basis of what the Committee currently knows, none of the other hypotheses can be rejected definitively. The March 21 and 24 letters (Petricciani 1983a,b,c) make clear that the FDA considered that contaminated blood or plasma (including AHF concentrate) might be a public health hazard. It is not clear, however, what the effect would have been on the availability of blood and blood products had the FDA required the recall of all stocks that had not been collected or manufactured in accordance with the requirements memorialized in the March 24 recommendations. There was grave concern in the whole blood collection and plasma fractionation industry, as well as in the medical profession and among patient groups, that sufficient blood and blood products remain available for necessary treatment. A recall of existing stocks could have seriously disrupted the availability of necessary therapeutic agents.
On the other hand, the Committee is puzzled why there is no evidence from any of the materials that it reviewed that a careful analysis was made of the
availability issue or that thought was given to some form of "staged" recall that would replace unscreened inventories as soon as new screened and segregated products became available in requisite quantities (Novitch 1983; Petricciani interview). Hence it is not possible for the Committee to conclude that the FDA appropriately balanced the two public health concerns: risks of infection and availability of blood and blood products.
Although the Committee has found no direct evidence on this point, the FDA's approach to the possibility of a recall in this instance may well have been influenced by its standard, consensus-based approach to regulation of blood and blood products. The letters of John Petricciani (1983a,b,c) are suggestive of this. For example, they did not announce themselves as falling within any particular category of FDA issuance, which, under the FDA's contemporaneous organizational regulations, included "regulations," which had the force of law; "guidelines," which represented approved practices; and ''recommendations," which were stated to be "matters authorized by but not involving direct regulatory action." The Petricciani letters (1983a,b,c) were clearly not regulations and they did not state that they were either guidelines or recommendations. Although the letters sometimes used language such as "we request" and asked manufacturers to "please advise the Office of Biologics" of the procedures that they had adopted, there were also sections that were written in the language of demand or requirement. The letters spoke, for example, of "compliance" with the notices and stated that if a producer was not following certain procedures that had been approved by the FDA's Center for Biologics Evaluation and Research for major organizations, its procedures must be submitted directly to the Office of Biologics for approval.
The ambiguity about whether the Office of Biologics was responding to consensus and officially recognizing good practices or was imposing requirements is emblematic of a consensual form of regulatory posture. The action was in the public health tradition of teamwork among multiple actors, both public and private. Moreover, because there was no consensus at this time, an appropriate trade-off between risk of infection and product availability with respect to blood and blood products that might possibly be contaminated with the AIDS virus, a recall may have seemed self-evidently premature.
The Committee simply has no record on the basis of which it could accept or reject the hypothesis that the FDA's failure to demand a recall at this point was motivated in part by considerations of economic loss to the manufacturers or political consequences. There is evidence in the general political and regulatory record, circa 1983, that might lend credence to the hypothesis that stronger action by the FDA would have run political risks. There was a strong deregulatory environment in the administration generally [Paperwork Reduction Act of 1980—P.L. 96-511; Executive Order on Federal Regulation, No. 12291, February 17, 1981] and at the FDA [DHHS response to Executive Order 12291—46 F.R. 26052] which, assuming industry opposition to recalls, might
have put the Office of Biologics in a difficult position had it wielded its mandatory powers (or threatened to do so in order to assure "voluntary" recalls).2 Moreover, as was mentioned earlier, these actions took place in a climate of White House skepticism concerning AIDS activism and the aftereffects of the swine flu incident.
Summary and Conclusions
The Committee is uncertain what public health consequence might have resulted from a more aggressive recall strategy circa March 1983. The uncertainties surrounding AIDS and its transmissibility may well have justified—certainly not rendered unreasonable—the modest approach of the Petriccianni letters (1983a,b,c). Nevertheless, with respect to transfused blood and nonpooled plasma products, small successes in screening out infective blood would almost necessarily have translated directly into lives saved. Hence, the FDA should have implemented its policy in a fashion designed to assure maximum efficacy. The Committee is concerned that the Petricciani letters failed to accomplish this goal in two respects.
First, the letters were difficult to interpret and did not reveal their own legal status. Such a tactic might be prudent from an agency operating at the margins of its regulatory authority, but that was hardly the situation here. Plasma fractionators and collection centers might thus easily have responded to them in quite different ways. Second, there is no indication that donor screening triggered serious analysis of what to do about a possible recall or withdrawal of already collected plasma or manufactured AHF concentrate. In the Committee's view, any action to increase the safety of new blood units or products should trigger systematic consideration of what to do concerning existing, and presumptively less safe, stocks. The Committee believes that this should be done even in situations such as this one, where the scientific uncertainties concerning AIDS and its transmission remained high.
The question of what to do concerning the recall of AHF concentrate that might be contaminated with the pathogen that was the causative agent in AIDS seems to have come to a head in midsummer 1983. Virtually the entire meeting on July 19, 1983, of the FDA's Blood Products Advisory Committee (BPAC) was devoted to the question of what was known about AIDS and its transmission through the blood supply, and what BPAC should recommend to the FDA concerning recalls (FDA, BPAC 1983; see also Chapter 3).
The National Hemophilia Foundation seems to have been the only participant at the BPAC meeting that supported the automatic recall of any product that was found to have been manufactured with plasma taken from a person subsequently determined to have AIDS or to have had characteristics strongly suggestive of AIDS. Indeed, the BPAC recall agenda seems to have been set by an NHF Medical and Scientific Advisory Council (MASAC) position favoring automatic recall (FDA, BPAC 1983). However, that position had been formulated by the NHF prior to the BPAC meeting. At the meeting itself, NHF medical director, Louis Aledort, first stated the NHF position, then followed by stating his personal view that the NHF position had been formulated prior to the consideration of the Pharmaceutical Manufacturers Association assertions concerning the possible impact of automatic recalls on the availability of AHF concentrate (FDA, BPAC 1983).
The BPAC did not accept the automatic recall proposal. Instead, it recommended that product recall be handled on a "case-by-case" basis. The BPAC position was based on countervailing considerations. Distribution and use of "lots" of a product incorporating plasma from a donor with a definite diagnosis of AIDS was clearly undesirable. However, the BPAC was sufficiently concerned about the possibilities of misdiagnosis based on particular signs and symptoms, and about the effect of automatic recalls on the availability of AHF concentrate, that it was unwilling to recommend to the FDA that it demand automatic recall whenever a product was thought to contain plasma from an individual confirmed as or suspected of having AIDS.
The views presented to the BPAC were reviewed within the FDA (Donohue 1983). Based on that review it was decided that the "working policy" of the Office of Biologics would be to evaluate the desirability of a recall on a case-by-case basis whenever a batch of AHF concentrate was found to have been produced from a plasma pool that included material from a donor later found to have AIDS or strongly suspected to have AIDS (Novitch 1983). In each case the Office of Biologics was to take into account its judgment of the accuracy of the diagnosis, the timing of the occurrence of symptoms of AIDS in relation to the time of donation of the plasma, and the impact of a recall based on this
particular donor's having had AIDS on the overall supply of AHF (Novitch 1983).
This case-by-case approach did not apply to whole blood or to plasma. Indeed, what to do concerning whole blood or nonfractionated plasma seems to have been missing from the discussion. Nor did the transcripts of the July 19, 1983, BPAC meeting or any of the internal FDA memoranda mention recall of any potentially contaminated units of cryoprecipitate (FDA, BPAC 1983).
The Committee questioned whether the FDA was correct in focusing the discussion and recommendations on AHF concentrate. Clearly, AHF concentrate was the greatest threat to hemophiliacs. It was in far greater use than cryoprecipitate and large numbers of donors contributed to each lot of AHF concentrate. Nonetheless, once everyone recognized the danger of material collected from an infected donor before the discovery that he or she had AIDS, that lesson seems equally applicable to all products that could be stored before use. Why, then, did the FDA not instruct blood banks to search their inventory for cryoprecipitate derived from a donor with AIDS and destroy it?
The question also remains, Why did the FDA not demand the automatic recall of AHF concentrate where it was determined that the product had been manufactured with plasma from a known or suspected AIDS carrier? Indeed, an automatic recall requirement could have gone further to include the recall of AHF concentrate manufactured through the use of plasma from any of the high-risk groups identified in the March 1983 Petricciani letters.
The legal hypothesis is once again easily rejected. No one seems to have doubted FDA's authority to establish an automatic recall policy. Other hypotheses are not nearly so easy to accept or reject.
On the face of the record of the BPAC meeting (FDA, BPAC 1983) and background material that has been made available to the Committee, public health considerations were at the forefront of the agency's concerns (Bove, Curran, Donohue, Epstein interviews). One should remember that at this point the causative agent for AIDS had not been isolated. There were still disputes in the scientific community about whether AIDS was a specific disease and, if so, whether it was or could be transmitted through blood and blood products (Bove 1983; FDA, BPAC 1983; Osborn interview). All shades of opinion seem to have been represented at the BPAC meeting of July 1983. Moreover, these scientific disputes were in evidence elsewhere on the public record. For example, in testimony before the House Intergovernmental Relations Subcommittee in early August 1983, Assistant Secretary for Health Edward Brandt stated that AIDS was believed to be transmitted through blood and blood products (U.S. House of Representatives 1983). He further testified that
evidence was strong that transmission was by a virus not yet identified and that the virus had a relatively long incubation period. Yet, at the same set of hearings Dr. Joseph Bove, then head of the Yale Blood Bank, chairman of BPAC, and a chair of the American Association of Blood Banks, testified that the scientific community did not know that AIDS could be transmitted by transfusion or through other blood products (U.S. House of Representatives 1983). Moreover, he emphasized that from what was known, the incidence of such transmission was likely to be less than one in one million transfusions (U.S. House of Representatives 1983).
The general tenor of the BPAC discussion on July 19, 1983, seems to have been that there was a strong possibility that a combination of inaccurate diagnosis, multiple donations by a suspected AIDS carrier, and the pooling process for the production of AHF concentrate could result in rapid elimination of all batches of Factor VIII should the FDA adopt an automatic recall policy. A worst-case scenario presented by Dr. Michael Rodell, who was representing the four member companies of the Pharmaceutical Manufacturers Association (PMA) involved with the manufacture of AHF concentrate, seems to have been particularly salient (FDA, BPAC 1983). In Dr. Rodell's scenario, under an automatic recall policy, the identification of a single donor as having or as suspected of having AIDS could result in the elimination of 25 million to 250 million units of AHF concentrate from existing inventories. Because the industry only produced 800 million units of AHF concentrate per year, three or four automatic recalls, based on identification of three or four suspect donors, who were each represented in 50 plasma pools in a given year, might completely eliminate the availability of AHF (FDA, BPAC 1983).
On the basis of discussion at the BPAC meeting, one view that could be taken of the contemporaneous public health situation might be something like this: The automatic recall of AHF concentrate whenever it was found to have been produced with plasma from a suspected AIDS carrier might easily have led to catastrophic shortages of a lifesaving and life-enhancing drug. Moreover, this policy would have been adopted in a situation in which the disease that it would prevent had not been firmly established to exist, the etiology of the disease was unknown, its transmissibility through blood and blood products was not established, and, if established, it suggested a risk of one infection in one million transfusions. Looked at in that light, public health clearly did not demand the adoption of an automatic recall policy. Indeed, it would seem to have demanded precisely the sort of cautious policy that the FDA adopted.
On the other hand, it is not obvious why the scientific information available to the BPAC or to the FDA should have been analyzed in precisely this way. The PMA's worst-case scenario involved a recall because of the discovery that plasma had been used from a very frequent plasma donor represented in as many as 50 plasma pools in a single year (FDA, BPAC 1983). No information (so far as the minutes reveal) was provided to suggest how often this worst-case
scenario might occur. Nor was there serious consideration of how collection and pooling might be managed to avoid the realization of this doomsday scenario in the future. Moreover, while there was still disagreement in the scientific community about the existence, etiology, and transmissibility of AIDS, those who disagreed with Assistant Secretary Brandt's testimony to the Intergovernmental Relations Subcommittee were in a distinct minority (Bove 1983; Osborn interview).
Indeed, the working hypothesis of most people in the field seems to have been that AIDS was a disease that had an epidemiologic pattern and perhaps an etiological pattern very similar to hepatitis (Curran, et al. 1984, Goedert and Gallo 1985). On that assumption one would have predicted with considerable confidence that AIDS would be a blood-borne disease and that it might rapidly become ubiquitous in a population, such as hemophiliacs, who took large amounts of AHF concentrate. If the disease also had a long latency period, as Assistant Secretary Brandt had suggested, then it would also be the case that computation of risk factors based on the current number of identified AIDS sufferers would vastly understate the risk. Hence, there was an equally plausible scientific scenario that suggested that without a very strong policy designed to eliminate all infective AHF concentrate from the market, AIDS would become widespread in the hemophiliac community. This, indeed, had been the position of some scientists at CDC and elsewhere for several months (Curran, Evatt, Foege, Francis interviews; Foege 1983).
From the public record it appears that the FDA, in a climate of uncertainty, adopted a middle position (FDA, BPAC 1983). FDA sought a policy that would screen out AHF concentrate potentially contaminated with HIV while maintaining availability of the product.
Nevertheless, there are two unanswered questions concerning this superficially prudent approach. The first has to do with the effectiveness of the approach as implemented. In retrospect, the Committee realizes that the approach was not particularly effective and that large numbers of hemophiliacs were infected with HIV between July 1983 and 1985 (the time at which a high percentage of available AHF concentrate was heat-treated) (Hammes pers. com. 1995; Leahy pers. com. 1995 McAuley pers. com. 1995; Persky pers. com. 1995). Given the large plasma pools from which AHF concentrate was made, it seems doubtful that anyone could have viewed donor screening as likely to be 100 percent effective for this product. Second, it is not clear why the loss of AHF concentrate, or shortage of it, for some period was considered intolerable. Alternative treatment modalities—admittedly less convenient and effective—were available for hemophiliacs (see Chapter 7). The second issue is addressed first.
There is no indication from the public record that the FDA did a careful calculation of the risks and benefits involved in a recall policy that might have been criticized as "overreactive" given the state of scientific knowledge. In a brief letter to Dr. Petricciani, Dr. Dennis Donohue, director of the FDA's
Division of Blood and Blood Products, reported his interpretation of the BPAC review (Donohue 1983). Dr. Donohue concluded that it was clear that the benefits of the availability of AHF concentrate as a protection against life-threatening or disabling hemorrhage far exceeded the risks of acquiring AIDS from the administration of AHF (Donohue 1983; Novitch 1983). What is not clear is why Dr. Donohue came so firmly to that conclusion given the evidence that was presented to BPAC—that is, why the prospective loss or shortage of AHF concentrate appeared to eclipse the threat of AIDS. Nor is it clear why the BPAC evidence and recommendations became the sole basis for FDA's recall policy. There may have been independent investigations and analyses done within FDA that supported these conclusions, but they do not appear in the record that has survived and been made available to the Committee.
The Committee believes it is not possible to conclude that the FDA made a decision that was clearly in the interest of public health given available information as of July 19, 1983. A close reading of the data suggests that a policy, not only of automatic recall, but of delicensing AHF concentrate until further information was available concerning its role in the transmission of AIDS might have been justified on public health grounds. This would have included, of course, a recall of all stocks of AHF then on the market and withdrawal of all AHF concentrate in the inventory of producers. On the other hand, the Committee would like to reiterate that the facts clearly did not compel such an aggressive approach.
To return to the first unanswered question, how effective could the policy that was adopted have been expected to be? It appears to have been extremely poorly conceived—indeed, virtually a nonpolicy. The case-by-case review was to use three criteria: (1) the reliability of the diagnosis, (2) the time lapse between the donor's donation and the appearance of AIDS symptoms, and (3) the effect on the availability of AHF concentrate (Donohue 1983). None of these criteria has any operational content, and in some cases they seem to ask for information that would be very difficult to acquire or process. There are no standards for what is a "good" diagnosis, no time period is established as an appropriate interval between donation and diagnosis, and no quantitative or qualitative guidance is provided concerning how large an effect on availability would be too great to tolerate. Indeed, no latency period had been established for AIDS and it is not obvious how anyone could rapidly determine how much AHF concentrate from a particular donor was still in stock or what proportion of existing stocks it represented.
Because there was no established system of tracking the health of blood donors, the Committee is puzzled by how plasma fractionators were to learn that a donor had AIDS. There is no suggestion that the CDC would or could legally share its lists of AIDS cases with the manufacturers. It was difficult to diagnose AIDS, and AIDS was not, at this time, a notifiable disease except in a few states
(CDC, MMWR, June 1983).3 Case-by-case recall thus seems to have been a nonsystematic approach that dealt with cases only if an individual was diagnosed with AIDS. However, there was certainly no assurance that most or even many of these diagnosed AIDS cases would come to the attention of the FDA, the CDC, or the plasma fractionators.
The hypothesis that the FDA lacked the resources to engage in stronger regulatory action cannot be rejected. On the one hand, the policy that the agency adopted was more resource-intensive from the agency's perspective than either an automatic recall policy or a blanket ban on further production and distribution of AHF concentrate. Indeed, the criteria established were so vague that the case-by-case approach required revisiting all the factual uncertainties present in the July BPAC meeting in every instance of a proposed recall. On the other hand, FDA may not have had the internal resources to develop an independent and more systematic recall policy. Hence, once BPAC made its recommendation, FDA had little choice but to follow it.
Given these circumstances it is not possible to reject hypotheses that feature the influence of institutional roles and of economic interests or political factors. The FDA's policy seems to have been very much influenced, if not wholly determined, by the advice that it received from the BPAC. That FDA committee included broad representation on this issue, but seems to have been strongly influenced by plasma fractionators. It was the manufacturers and their industry organization that had the important data related to availability and the technology of production. And those data seem to have been highly influential in molding the BPAC position. Moreover, this episode illustrates the potential weakness of consensus policymaking. In the absence of scientific consensus, FDA seemed to feel itself bound to craft a middle-of-the-road policy, which was remarkably vague and had unknown efficacy to protect the blood supply.
Because automatic recall might have virtually eliminated the market for AHF concentrate, the plasma fractionators had very strong economic interests in seeking to avoid that result. Their advice may well have been influenced by this economic interest. (Indeed, in their presentation they suggested that a
general recall would cause them to rethink the continuation of AHF concentrate production.) When the BPAC recommendation is combined with the deregulatory political climate of the times, including the micro-politics of the strong demand by hemophiliacs and the hemophilia treatment community that AHF concentrate remain available, it becomes plausible to imagine that the FDA resolved scientific uncertainty in favor of inaction—or, more precisely, very limited action—when a more protective policy would have been justified in the interests of public health.
It seems important to stress once again that these conclusions are being offered with the benefit of 20/20 hindsight focused on an incomplete record of the events. Had AIDS turned out not to be a viral disease capable of transmission through blood products, an FDA that had eliminated the availability of AHF concentrate in midsummer 1983 would have been at least as politically vulnerable as the FDA that had banned saccharin on the basis of good scientific evidence and a mandatory legal requirement only a few years earlier, or that had moved rapidly to develop a swine flu vaccine.
Summary and Conclusions
Once again this episode seems to suggest weaknesses in the regulatory process related to blood and blood products. The weakness took three forms:
A reactive agency posture permitted views of outside parties to shape both the recall agenda and the ultimate policy choice. Apparently, one of the reasons that the recall question was put on the BPAC agenda was the position the NHF's Medical and Scientific Advisory Council was taking on automatic recalls. Then when industry raised seemingly decisive objections to that policy and offered the case-by-case alternatives, there were no other options on the table for discussion. Moreover, blood and blood products of no interest to NHF did not receive consideration.
There was an apparent inability to rethink, or gather the necessary factual basis for rethinking, advice from outside parties once it was offered.
An acute decision dilemma failed to provoke a search for broader alternatives. If widespread contamination of pooled blood was a legitimate bar to regulatory action, why was there no mandate to search for alternative methods for pooling plasma that would minimize the risk of widespread contamination and avoid the doomsday scenario offered by the PMA? For example, the FDA could have encouraged the plasma fractionation industry to devise methods that would make it economically feasible to reduce the number of donors to a lot of pooled plasma. (There has apparently been no change in this situation. In December 1994, Dr. Rodell posed a similar scenario at the BPAC during a discussion on the possible recall of AHF thought to be contaminated with Creutzfeldt-Jakob Disease (CCBC 1994).
Heat-Treated AHF Concentrate and the FDA's Recall Policy
When Assistant Secretary Brandt testified before the U.S. Congress in August 1983, concerning FDA's decision to reject a blanket recall policy, he included as one of his reasons for that rejection the belief that safer AHF concentrate would soon be available (U.S. House of Representatives 1983). Leaving aside for a moment the question why this belief would not have justified a stronger rather than a weaker recall policy (e.g., a temporary ban or license suspension), it seems probable that Secretary Brandt's reference was to the emerging availability of heat-treated AHF concentrate.
By early 1984, all producers of AHF concentrate had been licensed to produce heat-treated AHF concentrate (Hammes pers. com. 1995; Leahy pers. com. 1995; McAuley pers. com. 1995; Persky pers. com. 1995). The FDA approved licenses of the heat-treated AHF on the basis of clinical trials demonstrating its biologic effectiveness to correct blood clotting problems (Fratatoni 1995). Testing revealed that heat treatment had substantial positive effects with respect to the elimination of hepatitis viruses, and it was conjectured by some that it would reduce the infectivity of AHF concentrate with respect to AIDS as well (Colombo, et al. 1985; McDougal, et al. 1985) (see Chapter 4 for complete discussion of the development of heat treatment methods for viral inactivation).
One must remember, of course, that these developments took place before the April 1984 isolation of the AIDS virus and at least 18 months prior to the development of a test (ELISA) that could detect the existence of HIV in blood and/or plasma prior to transfusion or the manufacture of AHF concentrate (Hammes pers. com. 1995; Leahy pers. com. 1995 McAuley pers. com. 1995; Persky pers. com. 1995). Hence, it was hoped that heat treatment would be useful against HIV, but it was not scientifically proven (Aronson interview). In fact, events later demonstrated that heat treatment was almost 100 percent effective with respect to the elimination of HIV (Levy, et al. 1984; McDougal, et al. 1985). Like most other strains of virus, the HIV virus was not resistant to the application of heat. In October 1984, the CDC announced that preliminary evidence concerning the effects of heat treatment on the viability of the AIDS virus was strongly supportive of the usefulness of heat treatment in reducing the potential for transmission of the AIDS virus in AHF concentrate and suggested that the use of non-heat-treated AHF concentrate should be limited (CDC, MMWR, October 1984). Most people accepted the model virus studies on
inactivation of the AIDS virus by heat treatment by 1985 (Hammes 1995; Leahy 1995; McAuley 1995; Persky 1995).
Heat-treated AHF became available in 1983 (Persky pers. com. 1995), and by 1985 heat treatment was accepted as effective in deactivating HIV (McDougal, et al. 1985). Even though there was a reasonable supply of heat-treated products, the FDA waited until 1989 to require the recall of all untreated AHF concentrate. By the time the FDA required recall of untreated units of AHF, the manufacturers had acted on their own (Epstein interview).
In this case, it is possible to construct a scenario in which the legal authority hypothesis cannot be so easily rejected. This may seem a strange conclusion given the argument above that the FDA might, in July 1983, have delicensed all AHF concentrate. However, the data necessary to delicense only untreated AHF concentrate, while leaving heat-treated AHF concentrate on the market, were not necessarily available for HIV. Removal of all AHF concentrate from the market might have been premised on a determination that, given the epidemiological evidence that was accumulating, AHF concentrate could no longer be determined to be safe under normal conditions of use. However, circa 1983, that same conclusion would have to be drawn with respect to heat-treated AHF concentrate as well. There was no test that would demonstrate definitively that heat-treated AHF concentrate was less infective with respect to HIV than non-heat-treated AHF concentrate until the mid-1984 studies by the CDC (CDC, MMWR, October 1984; McDougal, et. al. 1985). The CDC, in cooperation with Cutter Biological, conducted a study in which the AIDS virus was added to Factor VIII concentrate and was subsequently heat treated. The study showed that the virus was undetectable after 24 hours of heat treatment (CDC, MMWR, October 1984). Preliminary data were presented to NHF's Medical and Scientific Advisory Council in September 1984 and formed the scientific basis for NHF's subsequent October 1984 recommendations that heat-treated AHF concentrate be considered in the treatment of hemophilia.
On the other hand, the FDA could have decided that untreated AHF should be removed from the market because heat-treated AHF concentrate was safer with respect to the propagation of hepatitis. However, as is explained in Chapter 4, the FDA and others for many years had doubts about the utility of heat-treating AHF concentrate for hepatitis. Recipients of AHF concentrate were likely to be infected with hepatitis from exposure to numerous donors through multiple administrations of AHF concentrate (Aledort, Dietrich, Levine, Lusher interviews). This position itself may not have been well considered, but that seems to be the way that the FDA and others looked at the situation as of the end of 1983 (Aledort, Dietrich, Levine, Lusher interviews).
In this view, then, the situation from a legal and public health perspective might be summarized as follows: Because, prior to October 1984 at the earliest, the FDA could not demonstrate that heat-treated AHF concentrate was less infective for AIDS than untreated AHF concentrate—label inserts for heat-treated AHF concentrate could not say that it was a safer product [Baxter Product Insert, October 1983]—it had no legal basis upon which to demand that untreated AHF concentrate be recalled while leaving heat-treated AHF concentrate on the market. This is a situation in which the FDA's greater legal power (the probable authority to delicense all AHF concentrate owing to public health concerns) does not necessarily include the lesser (the power to eliminate untreated AHF concentrate on grounds of AIDS risk when it was not demonstrably less safe than heat-treated AHF concentrate). By contrast, the FDA might have required the expeditious elimination of untreated AHF concentrate on grounds of its demonstrated infectivity with respect to hepatitis by comparison with heat-treated AHF concentrate. Here, however, the public health benefits looked small (given the assumptions outlined above) in relation to the substantial economic costs involved in eliminating all untreated AHF concentrate from the market.
Moreover, there was some resistance in the medical community to the use of heat-treated AHF concentrate as it became available (Aledort, Dietrich, Levine, Lusher interviews). The reasons for this resistance included the problem of possible neoantigencity and subsequent inhibitor formation (which renders the patient untreatable with the AHF concentrates), the knowledge that heat treatment was developed to eliminate hepatitis virus, and the higher cost to the patient (Aledort, Dietrich, Lusher interviews). Physicians continued to prescribe non-heat-treated AHF to patients known to be infected with hepatitis (Aledort, Dietrich, Lusher interviews). Under these assumptions, the NHF and physicians treating hemophilia were reluctant to recommend the heat-treated product until there was more clinical evidence to prove that it was effective. Indeed, some manufacturers have suggested that the FDA's concerns about availability and resistance from the medical community probably influenced the FDA not to suspend licenses for untreated AHF concentrate when heat-treated AHF concentrate products were approved (Hammes pers. com. 1995; Leahy pers. com. 1995 McAuley pers. com. 1995; Persky pers. com. 1995).
It is worth noting, however, that there is another perspective on this story which might suggest that more aggressive action would have been warranted based wholly on the protection of individuals with hemophilia against the further spread of hepatitis. It is not clear, for example, exactly how great the economic or political costs would have been to plasma fractionators and to the FDA to remove untreated AHF concentrate from the market more rapidly. The Committee has only fragmentary information about how rapidly untreated AHF concentrate was removed through manufacturer's voluntary efforts or about what FDA action would have added to the rapidity of its removal. However, Alpha
Therapeutic reports that while licensed to produce heat-treated Factor VIII in February 1984, it was already processing 45 percent of its total Factor VIII output with the heat treatment step in anticipation of licensing and requested voluntary withdrawal of its license for untreated AHF concentrate in July 1985 (McAuley pers. com. 1995). Miles, Inc. reports that 40 percent of its shipped product (Factor VIII) was heat-treated in 1984 and 99 percent was treated during 1985 (Hammes pers. com. 1995). (The timing was somewhat later concerning Factor IX.) Moreover, by 1985 the unit prices to intermediate consumers for heat-treated Factor VIII were only slightly higher than the unit price of untreated Factor VIII in 1983 and 1984 (Hammes pers. com. 1995; Persky pers. com. 1995). However, it was reported to the Committee that the price to the final consumer was higher and this was one of the reasons why some in the medical community did not recommend heat-treated AHF concentrate earlier (Dietrich, Lusher interviews). From these data it would appear that supplies of treated AHF concentrate were rapidly becoming available and that the costs of eliminating untreated AHF concentrate in inventory would have been limited essentially to its production costs and the costs of retrieval.
Had the FDA demanded recall of all untreated AHF concentrate—based perhaps on an assumption that the AIDS virus would turn out to be more like the general run of viruses in its susceptibility to pasteurization—there might have been no legal contest from the industry. Once a retrovirus had been isolated as the causative agent in AIDS, most virologists would have presumed that it would be susceptible to heat inactivation. Moreover, by 1985 the medical and scientific community agreed that heat-treated AHF concentrate was less likely to infect humans with HIV (McDougal, et al. 1985). Hence, even if the FDA doubted its legal authority to press for a recall of untreated AHF concentrate, it might well have achieved that goal without any test of its legal authority. And, if tested, it might well have prevailed on one of several theories.
Summary and Conclusions
On the positive side, this series of events demonstrates some of the strengths of industry self-regulation. The plasma fractionation industry seems to have perceived clear advantages to developing a workable heat treatment process and to have rapidly secured approval for different processes (Hammes pers. com. 1995; Leahy pers. com. 1995; McAuley pers. com. 1995; Persky pers. com. 1995. Also, the FDA licensed heat-treated AHF concentrate with commendable speed using relaxed testing requirements.
On the other hand, the weaknesses of this approach are also evident. The industry is responsive to the demands of physicians and therefore was not particularly interested in the pursuit of heat treatment when the threat was
hepatitis because, at the time, the hemophilia treatment community considered the risk of hepatitis to be an acceptable price to pay for the benefits of AHF concentrate. This delayed development for several years. Then, when heat treatment was coming into use there seems to have been no particular effort by the FDA to help overcome treater resistance to the new product. Any regulatory actions would have required careful attention to availability problems, but there seems to be no reason to believe that the transition from untreated to treated AHF concentrate could not have been accomplished faster with active agency leadership (see also discussion in Chapter 4).
Lookback and Notification of Individuals Transfused with Contaminated Blood Products
By 1985, it was known that recipients of HIV-infected blood products became infective immediately but remained asymptomatic for up to five years (Curran, et al. 1984; Bove 1986), during which time those with whom recipients had intimate contacts were at risk of acquiring HIV. The FDA did provide interim procedures to blood establishments to apply when a repeat donor had a positive (repeated reactive) test for HTLV-II antibody, which included lookback of all prior donations and notification of all consignees of possibly contaminated products (Esber 1985). However, despite a 1988 presidential directive to the Department of Health and Human Services to formulate a lookback and notification policy for tracing the recipients of possibly infected blood products, the FDA did not issue recommendations until September 1991 (FDA, BPAC 1991a,b). Why did it take the FDA from 1985-1991 to recommend tracing recipients of transfusions from a donor who was later found to have AIDS?
One of the basic principles of infectious disease prevention is notifying infected individuals so that they can avoid infecting other people with whom they come into contact. Even as early as 1983, it was clear that AIDS had an incubation period of at least 18 months, and later it appeared that approximately 5 years was a common interval between exposure, infection, and developing the symptoms of AIDS (Bove 1986). In January 1983, there was evidence for transmission of AIDS by heterosexual contact (CDC, MMWR, January 1983). Thus, transfusion recipients are likely to transmit HIV infection without being aware that they are doing so. Neither the FDA recommendations about donor questioning and deferral nor the FDA working policy on recall strategy contained recommendations concerning recipients of blood products donated by
individuals subsequently found to have HIV infection. Failure to develop such a policy is peculiar.
In 1988, the Department of Health and Human Services received a presidential directive to implement actions in response to blood safety issues raised by the report of the Presidential Commission on the Human Immunodeficiency Virus Epidemic. One issue was prompt notification of individuals who are at increased risk of HIV infection but are unaware of their increased risk (Presidential Commission on the Human Immunodeficiency Virus Epidemic 1988). In 1991, Jay Epstein, acting director of the FDA's Division of Transfusion Services, reported progress to the BPAC and solicited their advice about the wording of regulations (FDA, BPAC 1991a,b). The issues in question included lookback4 and recipient notification.5
At the September 26–27, 1991, BPAC meeting, Dr. Epstein presented recommendations that were to be applied to product recalls and recipient tracing (lookback) related to components from prior collections. The BPAC unanimously approved the recommendations, which included the following: "When it is discovered that a donor is HIV-positive a blood center should search for recipients of any donations from that donor during the preceding five years and quarantine any suspect blood product. Blood centers must also notify the consignees of units for transfusion obtained from the donor's prior collections of the results of additional testing of the donor's current blood collection" (FDA, BPAC 1991b). Assuming that this policy is sufficient for the future, why did it take so long for the FDA to take action to protect the contacts of people who may have unknowingly become infected with HIV through a transfusion?
Did the FDA have insufficient information to justify legally a vigorous stance about lookback? To the contrary, there was an abundance of relevant biomedical evidence by 1985. In a memorandum from the FDA to blood establishments dated July 22, 1985, the FDA provided interim procedures to "apply when a repeat donor has a positive (repeated reactive) test for HTLV-II antibody," which included lookback of all prior donations with a maximum dating period of products and notifying all consignees of products so that products can be destroyed (Esber 1985). However, the Committee found no evidence on the record that the FDA formulated recommendations on lookback until January 18–19, 1991, when Dr. Epstein presented to the BPAC a draft of the recommendations the BPAC voted to adopt seven months later (FDA, BPAC
1991b). The FDA had both legal justification and sufficient information for taking action on lookback and notification. Aggressive action by the FDA could have been justified legally and on public health grounds, because notifying an individual that they had HIV could have reduced the risk of secondary transmission to a loved one. Clearly, public health considerations strongly favored a policy of lookback and notification.
The Committee has found no evidence to suggest that resource constraints prevented action. A lookback and notification policy would primarily burden local blood centers, which would have responsibility for tracing recipients of blood or blood products that had been transfused any time within five years of the donor becoming aware of his or her HIV infection. The FDA may have been sensitive to the resource limitations of blood centers, but the Committee has found no evidence to suggest that anyone sought to influence the FDA to delay instituting a policy of lookbacks.
Summary and Conclusions
The FDA's delayed approach to lookbacks is similar to its general approach to patient information detailed elsewhere in this report (see Chapter 7). Providing information in such circumstances as those attending the HIV invasion of the blood supply seems to have a particularly strong justification. Information would have permitted protective action concerning others as well as possible benefits for the infected, or potentially infected, individuals.
INFLUENCE AND RESPONSIBILITIES OF OTHER ORGANIZATIONS
While the FDA had primary regulatory responsibility for blood and blood products, it acted in an environment populated by other organizations having both responsibilities and influence. Other than the FDA, the primary governmental agencies were the Centers for Disease Control and the National Institutes of Health. The primary nongovernmental organizations were the not-for-profit blood bank community, the commercial collectors of plasma and producers of plasma fractionated products, and the National Hemophilia Foundation.
The National Institutes of Health has responsibilities for intramural and extramural biologic and biomedical research. The NIH has a scientific role in sponsoring the advance of biological and biomedical knowledge and in ensuring the integrity of research findings. It also has a public health responsibility that is discharged, in part, through its allocations of research funds in relation to public health needs (U.S. Government Manual 1994).
In connection with the AIDS contamination of the blood supply, there was little basic biomedical knowledge on HIV and AIDS and the rapid development of new knowledge was extremely valuable. Once the AIDS virus had been isolated, events moved very rapidly toward effective screening and treatment of blood and blood products. The Reagan administration and the NIH itself have been criticized for their failure to make AIDS research an urgent matter (Alter, Gallo, Perkins interviews). While it is clear that scientific uncertainty had a great impact on the regulatory process, we do not know what extent the devotion of greater research resources at earlier stages would have sped the resolution of some of the basic mysteries concerning AIDS as a disease and its transmission through the blood supply. It seems safe to assume that there would have been some effect; the magnitude is what is in doubt.
The Centers for Disease Control has responsibility for epidemiological research and monitoring, as well as for the dissemination of information concerning public health markers. In many ways the CDC seems to have played its early warning role quite well in this crisis. Certain CDC officials were extremely active very early in the critical period that we are discussing. They offered hypotheses about the etiology of the disease and its potential transmissions through the blood supply, which turned out to be quite accurate (Curran, Evatt, Foege, Francis interviews). Other governmental agencies seemed rather slow-moving by comparison with the CDC. However, it should be remembered that the CDC's role is to collect and disseminate information that it believes to have public health significance. It is not responsible for regulating other parties' behavior. Hence, it can responsibly use a much lower threshold of certainty in sounding alarms and pressing data on others. By contrast, it was sensible for other public health agencies to act cautiously considering that the CDC's concerns were based on a small number of cases and on hypotheses about those cases that appear more plausible in retrospect than they did at the time (Foege interview).
With respect to nongovernmental organizations, it seems clear that there was a presumption that the blood bank industry, and perhaps to a lesser degree the commercial plasma and plasma fractionation industry, would be substantially self-regulating. Moreover, there was considerable concern throughout the critical period reviewed to take into account the desires and knowledge of affected producer, patient, and physician constituencies, particularly when they were highly organized and active. Moreover, these same parties were given a substantial role in providing technical advice to the FDA concerning the exercise of its regulatory responsibilities (Aledort, Bove, Rodell interviews).
In normal times these close working relationships between the FDA and the nongovernmental sector have major advantages, as we point out at the beginning of this chapter. Indeed, it would appear that self-regulation and advice-giving played important positive roles in mitigating and ultimately limiting the tragedy of blood-borne HIV virus. Certain blood banks acted early to institute screening programs. Alpha Therapeutic acted with great dispatch in eliminating unscreened components from its AHF concentrate (Gury 1982). And the plasma fractionation industry adopted a strong screening policy that was urged on all its members (ABRA 1983) well before the March 1983 FDA letters that recommended screening (Petricciani 1983a,b,c). Ironically and tragically, plasma fractionators were not encouraged by government to aggressively produce and market heat-treated AHF concentrate a year before scientific proof that heat treatment killed HIV, but they moved rapidly to provide that product once it became plausible to believe that heat treatment was effective against HIV. Although many may now think that its advice was myopic, the National Hemophilia Foundation was very active in attempting to balance the risks of HIV infection against the risks of uncontrolled bleeding and in informing its member units concerning what it believed to be the proper course of action [e.g., NHF Chapter Advisories #4-12; NHF Medical Bulletins #3-11].
Nevertheless, we have noted many worrisome aspects of the FDA's reliance on both self-regulation and information development and dissemination in the nongovernmental sector. Because the FDA did not demand information and monitor action, it had no way of knowing whether either self-regulation or implementation of its recommendations would be effectively carried out by regulated parties. The FDA and others may have been overly influenced by the concerns of the NHF and other treaters that AHF concentrates remain fully available (in some cases, in a non-heat-treated form) (FDA, BPAC 1983). Information about availability and the effects of regulation on availability seems to have come almost exclusively from the industry and to have had major impacts on FDA decisionmaking (FDA, BPAC 1983). Actions were not taken in the absence of consensus among nongovernmental parties, when stronger
action to manage uncertainty in the interest of public health may have been warranted. Reliance seems to have been placed on the ordinary processes of the market in substituting higher for lower quality products, when it is uncertain whether those processes were effective or were as effective as would have been optimal.
The divided specialized responsibilities of governmental agencies and close association with and reliance upon nongovernmental organizations may work well in routine regulatory situations. In situations where uncertainty is great, there is a need to change regulatory routines to closely monitor potentially catastrophic health threats more closely. The ordinary routines of the regulatory system may too strongly favor the status quo. The difficult institutional design problem is how to develop a set of "triggers" that would force regulatory action, or the careful consideration of actions, without simultaneously overburdening the blood regulatory process as it deals with routine matters. Achieving this sort of balance in the system motivates the Committee's recommendations for restructuring the blood regulatory process (see Chapter 8).
THE ADVANTAGES OF MARGINAL THINKING
The FDA had several opportunities to take actions that appear to have had little or no risk for causing net harm and some possibility of reducing the number of people who contracted HIV infection through blood or blood products. None of these activities would necessarily have had large effects, which may explain why they were overlooked. The important strategy that these instances highlight, however, is a strategy of opportunistic, marginal improvements in safety in the absence of sufficient information or opportunities to make larger improvements. Consider the following examples that appear from the previous discussion.
Lookback and Notification
The FDA instituted the "lookback" recommendation in 1991, at least six years after it was clear that AIDS had a long incubation period during which the patient could transmit HIV through sexual contact or contact with blood (FDA, BPAC 1991b). Lookback required blood banks to contact recipients of blood from infected donors and notify them that they might be a HIV carrier and should be tested for HIV antibodies. To be sure, blood banks instituted their
own partial lookback systems. But there is no reason to believe that these practices were universal, uniform, or fully effective.
Removal of Untreated AHF Concentrate
When heat-treated AHF concentrate became available, there was an opportunity to reduce the amount of untreated AHF concentrate in the treatment centers' stocks. The number of potentially infectious vials would have decreased if, for example, the FDA had required a staged withdrawal of a vial of untreated AHF concentrate for every vial of treated AHF concentrate that was distributed. The success of this strategy presupposes that demand for treated AHF concentrate could be influenced by the FDA's policies. But clearly the agency could have been more aggressive in its removal efforts once more information was known about the efficacy and safety of heat-treated concentrate.
Use of Screened Whole Blood
Blood banks discard whole blood after about three weeks of storage. The Petricciani letters of March 1983 (Petricciani 1983a,b,c), among other things, required blood banks to screen potential donors for some high-risk behaviors and to quarantine or discard blood collected from these people. Blood collected from screened donors had a lower probability of being infectious. The Petricciani directive could have required blood banks to institute a policy of using blood from screened donors whenever possible. With this policy, the chance of receiving infected blood would have decreased, at no risk to the overall supply of blood. Once again the likely effects would have been small. Nevertheless, every substitution of a safe unit for a contaminated unit of blood would have prevented an infection and subsequent death.
Destruction of Potentially Contaminated Cryoprecipitate
Blood banks freeze cryoprecipitate from single donors and store it until the time of use. The FDA could have required blood banks to check their stores of frozen cryoprecipitate and destroy units obtained from donors subsequently found to have AIDS. This would have been an unpopular move, given the short supplies of cryoprecipitate, and of marginal benefit. Yet once again it would have been a low-cost option and would have provided a chance of saving some lives.
Innovative Techniques for Pooling Plasma
The smaller the number of donors to a lot of pooled plasma, the smaller the chance that the pool would contain HIV or other contaminants. The FDA could have encouraged the plasma fractionation industry to devise methods to reduce the number of donors to a lot of pooled plasma. For example, storing units obtained from repeat donors and using material from a donor only when a preset number of units had accumulated would reduce the number of donors to a given lot of pooled plasma. Other techniques might also have been available or developed.
Reducing risk by smaller pools would not eliminate risk. Indeed, a substantial pool is necessary to assure the efficacy of some plasma derivatives and reduce certain risks in others. But maintaining these levels could be accomplished while reducing pool sizes by a factor of 20. The critical point to this example is that because FDA promoted no changes in pooling practices, it faces in 1995 the same dilemma concerning Creutzfeldt-Jakob disease that it faced in 1983–1984 concerning AIDS (FDA, BPAC 1984; CCBC 1994). Recall of all tainted products might severely compromise the availability of AHF concentrate.
Testing Previously Untested Blood and Plasma for HIV
When blood and plasma collection centers first began to use the ELISA HIV test on blood donors (see Chapter 3), they began to face the problem of having two classes of blood and plasma: tested, which was presumably "safe," and untested, which had a higher risk of containing HIV. The FDA could have required blood and plasma collection centers as well as plasma fractionators to test each unused unit for HIV and discard positive units. The FDA did not issue such a directive, although there may have been testing carried out by some blood banks and others may have discarded untested units (Fratantoni 1995).
In summary, there were several opportunities for the FDA to specify practices that would have reduced the chance that blood or a blood product would transmit HIV, or that might have improved the FDA's chances to exercise regulatory control over future threats in the interests of public health. The Committee believes that there was enough knowledge at the time to be confident that these practices would have caused little or no harm. These practices were no guarantee, either individually or collectively, of a risk-free blood supply, but
each would have made a marginal improvement in reducing the chances for immediate or future harm from blood or blood products.
The analysis of these four critical events led the Committee to identify several weaknesses in the FDA's regulatory approach to blood safety issues. Most importantly, they reveal the lack of a systematic approach to safety in which events serve as triggers for careful consideration of options, including safer, but not zero risk, actions. In the absence of such an approach, uncertainty strongly reinforces the status quo as the agency waits for a consensus to crystallize and for the normal processes of FDA informal action and industry self-regulation to solve new problems.
American Blood Resources Association. Recommendations on AIDS and Plasma Donor Deferral; January 28, 1983.
Bove, Joseph. Statement before the U.S. Congress, Committee on Government Operations; August 1, 1983.
Bove, Joseph. Transfusion-Transmitted Diseases: Current Problems and Challenges. Progress in Hematology; 1986.
Centers for Disease Control, Morbidity and Mortality Weekly Report , June 24, 1983.
Centers for Disease Control, Morbidity and Mortality Weekly Report , October 26, 1984.
Centers for Disease Control, Morbidity and Mortality Weekly Report , January 7, 1983.
Colombo, et al. Transmission of non-A, non-B Hepatitis by Heat Treated Factor VIII Concentrate. Lancet; July 6, 1985.
Council of Community Blood Centers, FDA Committee Recommends Some Blood Recalls for Rare Neurological Disease. CCBC Newsletter; December 23/30, 1994.
Curran, et al. Acquired Immune Deficiency Syndrome (AIDS) Associated with Transfusion. New England Journal of Medicine, Volume 310; January 12, 1984.
Donohue, Dennis. Memorandum to Petricciani, John, FDA, Director, Office of Biologics; July 21, 1983.
Esber, Elaine, M.D., Director, Office of Biologics and Review, Food and Drug Administration. Memorandum to all registered blood establishments; July 22, 1985.
Falter, Steven, Director, Division of Bioresearch Monitoring and Regulations, Food and Drug Administration. Memorandum to Don Fowler; October 19, 1994.
Foege, William. Summary Report of Workgroup to Identify Opportunities for Prevention of Acquired Immune Deficiency Syndrome (AIDS); January 4, 1983.
Food and Drug Administration. Blood Products Advisory Committee, July 19, 1983.
Food and Drug Administration. Blood Products Advisory Committee, January 18, 1991a.
Food and Drug Administration. Blood Products Advisory Committee, September 26–27, 1991b.
Food and Drug Administration. Blood Products Advisory Committee, December 13, 1984.
Food and Drug Administration. Memorandum from Chief, Recalls and Administrative Actions Branch, to Don Fowler, Chief Freedom of Information Branch; November 8, 1994.
Food and Drug Administration. Regulatory Procedures Manual. U.S. Government Printing Office, Washington, D.C.; 1988.
Fratatoni, Joseph, M.D., Director, Division of Hematology, Food and Drug Administration. Letter to the Institute of Medicine; May 8, 1995.
Goedert and Gallo. Epidemiological Evidence that HTLV-III is the AIDS Agent. European Journal of Epidemiology, Volume 1; September 1985.
Gury, David, Vice President, Plasma Supply, Alpha Therapeutics. Letter to all source affiliates; December 17, 1982.
Hammes, William J., Associate Counsel, Miles, Inc. Letter to Institute of Medicine; February 13, 1995.
Hutt, Peter and Merrill, Richard. Food and Drug Law: Cases and Materials . The Foundation Press, Inc. Westbury, New York; 1991.
Leahy, Beth, Director Public Affairs, Armour Pharmaceutical. Letter to Institute of Medicine; March 17, 1995.
Levy, et al. Recovery and Inactivation of Infectious Retroviruses froFactor VIII Concentration. Lancet; September 29, 1984.
McAuley, Clyde, M.D., Medical Director, Alpha Therapeutic. Letter to Institute of Medicine; February 8, 1995.
McDougal, J.S., Martin, L.W., Cort, S.P., Mozen, M., Heldebrant, C.M., and Evatt, B.L. "Thermal Inactivation of the Acquired Immunodeficiency Syndrome Virus, Human T Lymphotropic Virus-III/Lymphadenopathy-associated Virus, with Special Reference to Antihemophilic Factor." The Journal of Clinical Investigation, Vol. 76; August 1985
National Hemophilia Foundation. Recommendations on the Prevention of AIDS in Patients with Hemophilia. January 14, 1983.
Novitch, FDA, Deputy Commissioner, Food and Drugs. Memorandum to Assistant Secretary of Health; July 23, 1983.
Persky, Marla, Associate General Counsel, Biotech Group Baxter Healthcare, letter to Institute of Medicine; February 24, 1995.
Petricciani, John, M.D., Food and Drug Administration, Office of Biologics. Letter to All establishments collecting human blood for transfusion. March 21, 1983a.
Petricciani, John, M.D., Director, Food and Drug Administration, Office of Biologics. Letter to All establishments collecting source plasma. March 24, 1983b.
Petricciani, John, M.D., Food and Drug Administration, Office of Biologics. Letter to All licensed manufacturers of plasma derivatives. March 24, 1983c.
Presidential Commission on the Human Immunodeficiency Virus Epidemic. Report of the Presidential Commission on the Human Immunodeficiency Virus Epidemic. U.S. Government Printing Office, Washington, D.C.; 1988.
Reilly, Robert, American Blood Resources Association. Letter to Petricciani, John, Food and Drug Administration; July 27, 1983.
U.S. Government Printing Office. U.S. Government Manual. Washington, D.C.; 1993–1994.
U.S. House of Representatives. Committee on Government Operations, Subcommittee on Intergovernmental Relations Hearing: The Federal Response to AIDS. U.S. Government Printing Office, Washington, D.C.; August 1, 1983.