Postapproval Risks 1976-85

PUnitedStatesGeneralAccountingOfficeGAOReport to the Chairman, Subcommittee .on Human Resources andIntergovernmental Relations,Committee on Government Operations,House of RepresentativesAprilFDA DRUG REVIEW1990Postapproval Risks1976-85Not to be released outside theBESTRICTEDGeneral Accounting Office def38 SPeCmeGAO/PEMD-90-15

United StatesGeneral Accounting OfficeWashington, D.C. 20648Program Evaluation andMethodology DivisionB-235944April 26,1990The Honorable Ted WeissChairman, Subcommittee on Human Resourcesand Intergovernmental RelationsCommittee on Government OperationsHouse of RepresentativesDear Mr. Chairman:In response to your request, we are submitting this report describing postapproval risks fordrugs approved by the Food and Drug Administration between 1976 and 1985. The reportidentifies drugs for which serious risks arose after approval for marketing, and itinvestigates the relationship of these risks to some attributes of the drugs and the reviewprocess.As we arranged with your office, unless you publicly announce the contents of this reportearlier, we plan no further distribution of it until 30 days from the date of the report. At thattime, copies of the report will be sent to the Secretary of the Department of Health andHuman Services and the Commissioner of the Food and Drug Administration, and we willmake copies available to others upon request.If you have any questions or would like additional information, please call me at (202) 2751854 or Dr. Michael J. Wargo, Director of Program Evaluation in Physical Systems Areas, at(202) 275-3092. Other major contributors to this report are listed in appendix VI.Sincerely,Eleanor ChelimskyAssistant Comptroller General

cExecutive SummaryAssessing the efficacy and safety of a drug to obtain Food and DrugAdministration (FDA) approval is a lengthy and complex process. Buteven after approval, many additional risks may surface when the genera1 population is exposed to a drug. These risks, which range from relatively minor (such as nausea and headache) to serious (such ashospitalization and death) arise from the fact that preapproval drugtesting is inherently limited. The extent of postapproval risks and thereasons they go undetected during preapproval testing, however, havenot been analyzed.PurposeThe Chairman of the Subcommittee on Human Resources and Intergovernmental Relations of the House Committee on Government Operationsasked GAO to study the frequency and seriousness of drug risks identified after FDA approval for marketing and to examine some of the characteristics of these drugs as a first step in understanding why theseadditional risks occur.BackgroundIH.2k.\ x-.---1-The drug approval process begins with the submission of an “investigational” application, when a drug company applies to FDA for permissionto test the drug in humans. Then, when the clinical studies involvinghumans provide evidence of a particular drug’s beneficial effect at anacceptable level of safety, the company submits a new drug application(120 were submitted in 1986) to FDA for approval of the drug for widespread use. The agency subsequently reviews all evidence pertaining tothe drug’s efficacy and safety. If it finds the cumulative evidenceacceptable, FDA approves the drug for marketing (after, on the average,29 months of review).The preapproval human clinical trials for a drug involve testing with arelatively small sample of the potential user population under controlledconditions that limit the extent of risk assessments. However, whentherapeutic benefits appear to outweigh the estimated potential risks,the new drug is approved as soon as possible for the benefit of thosewho can use it. After FDA approves the drug for marketing, it is thenused by patients under conditions much less controlled than those thatprevailed during testing.When a company markets an approved drug, it is required by law toinclude directions for its use-as well as warnings, precautions, andadverse reactions-on the drug’s label. Postmarketing surveillance thenidentifies potential adverse reactions not included on the original labelthat are discovered after marketing is begun. If an adverse reaction isPage 2GAO/PEMD-90-16FDA Drug Review: PostapprovalRisks 1976-86I1I1

75sExecutive SummaryIfound to be linked to the use of a drug, its label is changed to include theadditional risk. GAO reviewed the label changes for all 209 new drugs FDAapproved between 1976 and 1985 in order to determine the frequencyand seriousness of the additional risks linked to these drugs after theirinitial approval. GAO plans to continue its examination of the factors underlying the occurrence of serious postapproval risks in FDA-approveddrugs.Results in BriefIn studying the frequency and seriousness of risks identified afterapproval, GAO found that of the 198 drugs approved by FDA between1976 and 1985 for which data were available, 102 (or 51.5 percent) hadserious postapproval risks, as evidenced by labeling changes or withdrawal from the market. All but six of these drugs were currently marketed as of September 1989 and are deemed by FDA to have benefits thatoutweigh their risks. The serious postapproval risks are adverse reactions that could lead to hospitalization, increases in the length of hospitalization, severe or permanent disability, or death. These adversereactions resulted in a substantial change in the labeling of the drugs,typically either limiting the population for which they are intended or 'adding major wamings or precautions for their use.The number of serious postapproval risks is small when compared to thenumber of adverse reactions that had been identified at the time ofapproval. GAO did not attempt to determine the reasons why the risksemerged (or whether they could have been identified during preapproval testing), the extent to which patients were exposed to thembefore they were identified, or the number of patients affected. However, these findings make clear that further understanding is needed ofthe characteristics associated with these additional risks and why theyarose.Principal Findings II,!The serious postapproval risks identified in studying their frequencyand seriousness involved a wide variety of adverse reactions, includingheart failure, myocardial infarction, anaphylaxis, respiratory depression and arrest, convulsions, seizures, kidney and liver failure, severeblood disorders, birth defects and fetal toxicity, and blindness. Theseadverse reactions occurred over many drug classes. GAO found that in 12of 22 classes more than 50 percent of the drugs approved had seriouspostapproval risks. The 1 2 classes are cardiac drugs, psychopharmacologic drugs, drugs to combat drug abuse, antibiotics, fertility andantifertility drugs, metabolic and endocrine drugs, ophthalmic drugs,GAO/PEMDSO-lS FDA Drug Review: Postapproval Risks 1976-86

4’Iantiparasitic drugs, oncology drugs, anti-inflammatory drugs, anesthesiadrugs, and surgical drugs. The degree of additional risk varied: It wasmore serious and occurred more often for some drugs than for others,with approximately one quarter having serious postapproval risksaffecting three to five body systems (including cardiovascular, respiratory, central nervous system, renal, and gastrointestinal). Additionalrisks in a body system fall along a spectrum that includes other lesssevere adverse reactions. (See pages 24-41.)In examining some characteristics of drugs approved by FDA, GAO foundseveral that are associated with the presence of serious postapprovalrisks. Drugs that appeared on FDA’S postmarketing surveillance listbecause of adverse reactions not present on the current label were over10 times as likely to have serious postapproval risks as those that didnot. Drugs that were reviewed for use in children were more than twiceas likely to have serious postapproval risks. Drugs approved between1976 and 1980 were almost three times as likely to have serious postapproval risks as those approved between 1981 and 1985, although someevidence suggests that GAO’S assessment for the latter period may be low(see pages 51-53). GAO also found that among drugs approved in fewerthan 4 years, those that turned out to have serious postapproval riskshad generally been approved by FDA in a shorter time than those without such risks. The class of a drug may also be related to the emergenceof serious postapproval risks. (See pages 42-56.)RecommendationGAO recommends thatthe Commissioner of FDA establish formal procedures to evaluate postapproval risks for new drugs and use this information in premarketing review and postmarketing surveillance. GAObelieves that such procedures would contribute to better and moretimely drug labeling. GAO believes that FDA, in implementing this recommendation, should build upon the steps followed in this report: (1) identifying serious and nonserious postapproval risks, (2) enumerating therisks by drug class, (3) identifying the body systems affected by therisks, and (4) comparing the additional risks with those identified at thetime of approval. The Commissioner should also try to estimate the population exposed to the additional risks and assess their significance interms of expected fatalities and morbidity.Page 4GAO/PEMDSO-lB FDA Drug Review: Postapproval Risks 1976-85I

Executive SummaryAgency CommentsWhile agreeing that it is important to know whether the current drugreview process may overlook serious risks, HHS objected to GAO’S methodology. The department was also concerned that this report will unnecessarily alarm consumers, causing some to reject the use of lifesavingdrugs out of fear of adverse events that might occur only in extremelyrare instances, and that it could create a misleading impression of thedrug review process, by identifying drugs as having serious postapproval risks that could not have been so identified in the drug reviewprocess. Also, HHS did not concur with GAO’S recommendation, indicatingthat it was vague and that it assumed FDA could somehow anticipate theunknown.believes the study methodology was sound and, consequently, nosubstantial changes have been made in the findings, conclusions, or recommendation. However, GAO has modified the presentation in certainparts of the report to clarify the language and to avoid misunderstandings about the research performed (that is, the questions posed and themethodology for answering them).GAOWith respect to HHS’S view that consumers will be unnecessarily alarmedby this report, GAO has noted that 96 of the 102 drugs identified as having serious postapproval risks were currently marketed as of September1989 and thus are deemed by FDA to have benefits that outweigh theirrisks, despite the additional risks that have been identified. GAO does notdispute FDA’Sassessment of benefits and risks. What is in question hereis the degree to wlhich careful analysis can help in reducing predictablerisk to the consumer.HHS’Scomments are reproduced as appendix V to this report. GAO’Sresponse to each point is contained in appendix V and presented at otherappropriate places in the body of the report.Page 5GAO/PEMD-90-15 FDA Drug Review: Postapproval Risks 1976-85