Ethical considerations in studying drug safety — the institute of medicine report

T h e n e w e ng l a n d j o u r na l o f m e dic i n e H e a l t h L aw, E t h i c s , a n d H u m a n R i g h t s Ethical Considerations in Studying Drug Safety —
The Institute of Medicine Report
Michelle M. Mello, J.D., Ph.D., Steven N. Goodman, M.D., M.H.S., Ph.D., The tumult arising from revelations of serious ment began, some argued that the evidence of the safety risks associated with widely prescribed inferior safety of rosiglitazone was strong enough drugs, including rosiglitazone (Avandia, Glaxo­ to make the trial ethically unjustifiable. Two FDA SmithKline), rofecoxib (Vioxx, Merck), and cele­ epidemiologists wrote in a 2008 memorandum coxib (Celebrex, Pfizer), has led to widespread that a head­to­head trial “would be unethical and recognition that improvement is needed in our exploitative” and that even a robust informed­ national system of ensuring drug safety. Notwith­ consent process could not overcome the prob­ standing federal legislation in 2007 that strength­ lem.9 This was not the consensus FDA view, ened the authority of the Food and Drug Admin­ which was that the uncertainty regarding the istration (FDA) in the postmarketing period,1 cardiovascular risks associated with rosiglitazone, critical weaknesses in the national system persist. as well as those associated with pioglitazone, Central to these weaknesses are dilemmas was sufficient to justify a trial.10 surrounding not only the science but also the These concerns triggered a February 2010 ethics of drug­safety research,2 many of which letter from members of Congress to the FDA de­ came to the fore in the heated public debate manding a justification for the trial and alleg­ about the Thiazolidinedione Intervention with ing that the consent form did not provide ade­ Vitamin D Evaluation (TIDE) trial, which com­ quate risk information.11 In response, FDA pared the cardiovascular outcomes of long­term Commissioner Margaret Hamburg expanded the treatment with rosiglitazone with those of pio­ FDA investigation of the safety of rosiglitazone, glitazone (Actos, Takeda) in patients with type 2 obtained advice from an FDA advisory commit­ diabetes.3 At the request of the FDA, an Institute tee, and asked the IOM to convene our commit­ of Medicine (IOM) committee, on which we tee.6 Although the FDA advisory committee rec­ served, was convened to examine the ethics and ommended that the TIDE trial be continued if science of FDA­required postmarketing safety re­ rosiglitazone was permitted to remain on the search. In this article, we review the key ethics market, in September 2010, the FDA halted the findings from the committee’s May 1, 2012, re­ trial and placed stringent new restrictions on port4 and offer some reflections on the chal­ the availability of rosiglitazone.12,13 The TIDE experience made the FDA appreci­ ate the need for greater attention to the ethics of postmarketing research. First, it posed ques­ tions about what standard of evidence about drug In May 2008, the FDA ordered the manufacturer risk justifies a decision by the FDA to require of rosiglitazone, GlaxoSmithKline, to conduct a postmarketing research, particularly randomized trial in response to evidence from meta­analyses trials, as well as what evidence could render that rosiglitazone was associated with a higher such trials unacceptable. Second, it raised ques­ risk of myocardial infarction and death from car­ tions about what ethical obligations the FDA diovascular causes than placebo or medications has to patients who participate in these studies. that were not based on nonthiazolidinedione Finally, it highlighted a potential FDA role in comparators.5,6 Other studies suggested that pio­ ensuring that institutional review boards (IRBs) glitazone, an alternative thiazolidinedione, was are completely informed in their efforts to pro­ not associated with such risks.7,8 Before enroll­ tect study participants. Although major deficien­ Downloaded from nejm.org on August 24, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. T h e n e w e ng l a n d j o u r na l o f m e dic i n e cies with the TIDE consent form were identified The committee argued that when the FDA re­ by some FDA scientists and, later, by the IOM quires a postmarketing study, it assumes a mea­ committee (Table 1),4,9 the TIDE investigators sure of ethical responsibility for the welfare of countered that it had been approved by “480 the study participants; exercise of that responsi­ ethics committees and IRBs.”14 However, the bility cannot be handed off to contractors or the language of the consent form, the trial design, industry sponsor. The responsibility is particu­ and the materials supporting the justification of larly strong when the patients’ treatment is de­ the trial raised a question for the IOM commit­ termined by the study, such as in a randomized tee about whether these bodies adequately under­ trial, linking any adverse outcomes directly to a stood the nature of the evidence that gave rise regulatory decision to require a study of that to the trial. The IOM committee proposed a type. This determination led to one of the most framework for evaluating the ethics of FDA­ important recommendations from the IOM com­ required postmarketing research15 and made a mittee: the responsibilities of the FDA to research number of ethics findings and recommendations.4 participants mean that it should mandate a ran­ domized design only if the FDA “has concluded ETHICAL RESPONSIBILITIES OF THE FDA
that an observational study could not provide The IOM committee began by noting that the the necessary information [to help answer the public health mission of the FDA gives rise to important public health question at issue], that potentially competing ethical obligations “to pro­ an RCT [randomized, controlled trial] is likely to tect the public’s health by having strong science generate the information within the necessary on which to base regulatory decisions” and “to timeframe, and that the necessary RCT is ethi­ protect participants in research that it requires.”4 cally acceptable.” This recommendation com­ Requiring a postmarketing study is an ethical ports with but adds some further conditions to decision, reflecting a weighing of these values.
the current legal authority of the FDA under the The committee described the conditions that FDA Amendments Act of 2007, which empowers must be present to justify a decision to require a the agency to require a randomized trial if it postmarketing study. The FDA should require cannot obtain the data it needs from an obser­ postmarketing research only when, first, the un­ vational study.1 certainty about the benefit–risk balance of a In light of the critiques of the TIDE trial as drug is so great that a responsible decision inherently unethical, the committee addressed about its regulatory status cannot be made on the justifiability of trials in which participants the basis of existing evidence; second, the re­ may encounter a net increase in risk, as com­ search will reduce this uncertainty; third, the pared with ordinary clinical care, but no realis­ FDA will use the research results expeditiously tic prospect of personal benefit. It argued that to make a regulatory decision; and fourth, suffi­ such trials can be justified only if they are nec­ cient protections for research participants can be essary to answer a critically important public health question, if the potential risk is accept­ Table 1. Major Deficiencies in the Informed-Consent Form for the TIDE Trial.*
Consent Element
Specific Deficiencies
Did not clearly explain that the purpose of the study was to definitively establish that rosiglitazone in- volved a significantly higher risk of serious harms than pioglitazone Title included minor intervention (vitamin D) whose effects were not a subject of FDA concern Did not convey that the FDA was concerned primarily about the safety of rosiglitazone and was requiring the sponsor to conduct the TIDE trial to investigate safety signals Submerged cardiovascular risks associated with rosiglitazone in a list of outcomes related to potential effects of vitamin D on cancer and bone fractures Characterized previous studies as having provided conflicting findings concerning the cardiovascular safety of rosiglitazone, when the weight of the evidence was against rosiglitazone Inadequate explanation of alterna- Did not disclose that the current clinical standard of care had moved away from the use of rosiglitazone Did not mention that the American Diabetes Association had recommended that rosiglitazone not be used * Adapted from the Committee on Ethical and Scientific Issues in Studying the Safety of Approved Drugs, Institute of Medicine,4 and Graham and Gelperin.9 FDA denotes Food and Drug Administration, and TIDE Thiazolidinedione Intervention with Vitamin D Evaluation.
Downloaded from nejm.org on August 24, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. able and minimized, and if special safeguards are in place, including a highly explicit informed­ sures to strengthen the consent process in order consent process to ensure that patients under­ to maximize patients’ understanding of the con­ stand that they are potentially shouldering addi­ text in which the trial is being conducted, includ­ tional risk solely to contribute to the public good. ing what is already known about the risks asso­ Specific actions that the FDA should take to ciated with the drug. The report discussed both meet its ethical obligations include specifying the specific disclosures in the informed­consent study design, title, end points, and primary analy­ form and special efforts that could be made to ses; identifying design features that it views as ensure adequate comprehension of complex in­ ethically and scientifically indispensable; and, formation regarding risks (Table 2). To assist for clinical trials, specifying a safety­monitoring IRBs, the committee recommended that the FDA scheme. The committee recommended that the issue guidance interpreting current informed­ FDA routinely communicate with IRBs about re­ consent regulatory requirements in the context quired postmarketing studies — for example, by of required postmarketing studies.
issuing a letter to accompany IRB applications that conveys information that is material to the IRB’s determination of the ethics of the research, as well as providing additional communications over the life of the study as warranted by new Because a true picture of the benefit–risk profile information about the drug or by changes in pro­ of a drug only emerges over time, two different fessional practice. The committee also believed IOM committees have stressed the need for the that the FDA was ethically obligated to actually FDA to fully embrace a “life­cycle approach” to use the findings from required studies to make drug regulation, in which its obligations to pro­ tect public health are taken as seriously once a drug is on the market as they are before approval INFORMED-CONSENT PROCESS FOR REQUIRED
is granted.4,16 Postmarketing regulatory oversight POSTMARKETING TRIALS
is assuming heightened importance as the FDA The IOM committee emphasized that the ade­ accrues additional authority to fast­track drugs quacy of the informed­consent process is only for approval on the basis of more limited evi­ one element in the ethics of FDA­required post­ dence than was previously required in order to marketing research. Other central, and indeed address unmet medical needs and accelerate in­ prior, features include ensuring that the selec­ novation.17-19 This changing landscape raises tion of participants is equitable and that the level several challenges for ensuring the ethical con­ of risk to which they are exposed is acceptable. duct of research with approved drugs and balanc­ The committee also recognized, however, that ing societal interests in drug innovation and drug there are challenges to achieving meaningful in­ safety. We highlight two of these challenges here.
formed consent in postmarketing trials of drugs First, not all postmarketing research is ethi­ for which there is a signal indicating the possi­ cally equivalent. The TIDE trial represented an bility of drug­related harm. In such cases, there iconic kind of postmarketing study: an FDA­ is a suspicion that the benefits of the drug may required randomized trial to study a drug whose not justify its risks and often that it may have a benefit–risk profile was under a cloud of suspi­ worse benefit–risk profile than alternative drugs cion and at a time when alternative treatments available to treat the same condition. The com­ were available, albeit not all well studied. The mittee concluded that for postmarketing trials risks to patients of participating in the trial of such drugs, there are “heightened obligations probably outweighed the prospect of direct ben­ to ensure that potential research participants un­ efit. By contrast, when the FDA requires an ob­ derstand the risks posed by study enrollment.” 4 servational study that uses previously collected This was of particular importance for rosiglita­ data, the clinical experience of the participants zone, because the cardiovascular problem it ap­ is unaffected, the risks incurred are not at the peared to cause was the same outcome that behest of the FDA, and ethical concerns are good diabetic control was supposed to improve largely confined to confidentiality and the right — in other words, if this elevation in risk were to control one’s medical information.
real, there could be little offsetting benefit.
Both of these scenarios can be distinguished Downloaded from nejm.org on August 24, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. T h e n e w e ng l a n d j o u r na l o f m e dic i n e Table 2. Mechanisms for Strengthening the Informed-Consent Process for Postmarketing Drug-Safety Studies.*
Consent Element
Information to Be Conveyed
Clear explanation of why a new study is required even though the FDA already found the drug to have an acceptable benefit–risk profile Meaningful, balanced summary of existing evidence about the safety of the study drug Explanation of why experts deem it reasonable to ask patients to participate in the study, If applicable, a statement that participants are being asked to assume a higher risk than they would be likely to encounter under the current clinical standard of care solely for the purpose of answering an important public health question Drug-risk information should go beyond a bul et-point list of possible side effects, indicating where the weight of evidence falls concerning the likelihood that serious harms will oc- cur and the extent of current scientific uncertainty At a minimum, disclosures should include boxed warnings, a “major statement” in direct- to-consumer advertisements, salient findings from FDA advisory committees, and a summary of findings from peer-reviewed studies Clear explanation of how the care that patients receive in the study may differ from the care they currently receive or would be likely to receive if they did not enroll If applicable, a statement that the current clinical standard of care has moved away from use of the study drug, and an explanation of why Ongoing communication of Timely, comprehensible communications if safety signals intensify over the course of the study relevant new information Recommendation that study participants consult with their treating physician to determine whether continued study participation is prudent in light of new information * Adapted from the Committee on Ethical and Scientific Issues in Studying the Safety of Approved Drugs, Institute of from the context in which a phase 4 trial is re­ ethical difference between FDA­required post­ quired as a condition of an accelerated drug ap­ marketing research and comparative­effective­ proval and is initiated soon thereafter. Here, the ness research initiated by academic investiga­ trial requirement is not imposed because of a tors. By contrast, a comparative­effectiveness newly emerging concern about a drug already in study of two widely used drugs that is not oc­ clinical use but because additional evidence is casioned by heightened concern about the risks needed to confirm the initial judgment that the of one drug relative to the other is markedly dif­ benefits of a new drug are likely to outweigh its ferent, ethically, from a study required by the risks. Often, this initial judgment is based on FDA to pursue a safety signal that is already of the use of a surrogate end point for drug bene­ such concern that practice patterns are shifting, fit, not on the clinical outcomes that matter even if both studies use randomized designs.
most. Especially when the new drug targets an These differences highlight the need for IRBs unmet medical need, it may be in the patients’ to be sensitive to the place where a study falls best interest to take it, pending further timely within the life cycle of a drug and to the reason research. The ensuing trial is undertaken to for the research. Depending on who is initiating confirm the improvement in clinical outcomes the research, for what reasons, and when, the predicted by the surrogate — a different epis­ same study design may have very different rami­ temic and ethical situation than that in which fications for the benefit–risk balance of the substantial evidence suggests that the surrogate study and what patients need to know in order is misleading or that other harms might offset a to provide meaningful informed consent. Trials that may be regarded as unethical late in the The volume of phase 4 and other research life cycle because of accumulated evidence can with FDA­approved drugs is increasing, not only be much easier to initiate earlier if the need for because of the expanded authority of the FDA to additional research is anticipated and planned require such research but also because of the at the time of initial approval. In the case of growing volume of comparative­effectiveness re­ rosiglitazone, this need could have been antici­ search. In some cases, there may be no or little pated from preapproval data showing an adverse Downloaded from nejm.org on August 24, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. effect on serum lipids as well as the use of a late a persuasive reason why it would result in surrogate end point (glycemic control) for a competitive harm and the FDA determines that this harm outweighs the public health benefits Second, the experience with rosiglitazone of releasing the information.
underscored the fragility of our current system of discovering risks associated with drugs. This system relies heavily on drug sponsors and FDA scientists to conduct safety analyses on the ba­ The experience with rosiglitazone and the TIDE sis of data from clinical trials, some or all of trial offers a lesson in how our current approach which are not publicly available, and to release to the oversight of drug­safety and postmarket­ findings to the public. It has been shown repeat­ ing research can fail both the public and the re­ edly that the published record can misrepresent search participants. Although terminating the evidence known to the FDA.21,22 In the case of TIDE trial was justifiable, it left regulators with rosiglitazone, scientists from GlaxoSmithKline highly suggestive but nondefinitive data on the and the FDA had information from 42 clinical relative safety of rosiglitazone and the closest trials, of which only 7 were published and the clinical alternative, pioglitazone.32 others were inaccessible. Triggered by concerns Reactive policymaking is tempting but prob­ expressed by the World Health Organization in lematic. The history of regulation of human sub­ 2006, GlaxoSmithKline conducted and shared jects research suggests that rules that are “born with the FDA a meta­analysis of the safety of in scandal and reared in protectionism”33 often rosiglitazone that used these data, confirming a fall short of providing meaningful protections possibly elevated risk of ischemic events, but to research participants and that, once adopted, neither these results nor the primary trial results regulations can ossify and become difficult to were shared with the public until an unrelated dislodge. Nevertheless, the IOM committee’s re­ court settlement forced GlaxoSmithKline to re­ port makes a number of actionable recommen­ lease its complete clinical­trial data.23 This ac­ dations that the FDA can implement under its cess led to the published meta­analysis by inde­ existing authority.34 In addition, appointment of pendent researchers that made these data and an independent ethics advisory board would strengthen the decision making of the FDA as it It is often the work of independent scientists confronts emerging ethical challenges — both that has highlighted critical safety problems with those arising from required postmarketing tri­ approved drugs.5,24-29 Yet currently, data from pre­ als and those stemming from powerful new marketing studies that are submitted as part of drug surveillance systems, such as the FDA’s a new drug application or a supplemental new Sentinel Initiative. As the pace of the translation drug application are largely shielded from release of discoveries from bench to bedside continues to external scientists and the public owing to to intensify, so too does the imperative for concerns about a competitive disadvantage to thoughtful ethical governance throughout the drug sponsors.30,31 The IOM committee stopped life cycle of a drug.
short of calling on the FDA to increase public The views expressed in this article are those of the authors access to such data but recommended that the and, except where noted, do not represent the official position of the Institute of Medicine or of the committee that produced the agency initiate a process to determine ways to report discussed in this article.
“appropriately balance public health, privacy, and Drs. Faden and Goodman chaired, and Dr. Mello was a mem­ proprietary interests to facilitate disclosure” of ber of, the Institute of Medicine committee that produced the relevant data.4 Greater transparency would better Disclosure forms provided by the authors are available with equip independent scientists to investigate early the full text of this article at NEJM.org.
safety signals.31 Consideration should be given We thank the other IOM committee members for contribut­ to making drug­safety data from clinical trials From the Department of Health Policy and Management, Har- available to the public on request once the FDA vard School of Public Health, Boston (M.M.M.); the Depart- has reached a decision regarding a new drug ap­ ments of Medicine and Health Research and Policy, Stanford plication or a supplemental new drug application University School of Medicine, Stanford, CA (S.N.G.); and the Berman Institute of Bioethics, Johns Hopkins University, Balti- or once the manufacturer has abandoned the more (R.R.F.). application, unless the manufacturer can articu­ This article was published on August 22, 2012, at NEJM.org.
Downloaded from nejm.org on August 24, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved. 1. Food and Drug Amendments Act of 2007, Pub. L. No. 110­85, and protecting the health of the public. Washington, DC: Na­
tional Academies Press, 2007 (http:/ www.iom.edu/Reports/2006/ 2. London AJ, Kimmelman J, Carlisle B. Research ethics: re­
The­Future­of­Drug­Safety­Promoting­and­Protecting­the­Health­ thinking research ethics: the case of postmarketing trials. Sci­ of­the­Public.aspx).
17. U.S. Senate. Health, Education, Labor and Pension Commit­
3. ClinicalTrials.gov. Thiazolidinedione Intervention with Vita­
tee. Food and Drug Administration Safety and Innovation Act min D Evaluation (TIDE). 2011 (http:/ clinicaltrials.gov/ct2/show/ (discussion draft), 112th Cong., 2d Sess. 2012 (http://www.help NCT00879970?term=NCT00879970&rank=1).
.senate.gov/imo/media/audio/KER12172.pdf).
4. Committee on Ethical and Scientific Issues in Studying the 18. Chen JY, Carter M. Bioethics and post­approval research in
Safety of Approved Drugs, Institute of Medicine. Ethical and translational science. Am J Bioeth 2010;10:35­7.
scientific issues in studying the safety of approved drugs. Wash­ 19. Psaty BM, Meslin EM, Breckenridge A. A lifecycle approach
ington, DC: National Academies Press, 2012 (http:/ www.iom.edu/ to the evaluation of FDA approval methods and regulatory ac­ Reports/2012/Ethical­and­Scientific­Issues­in­Studying­the­Safety­ tions: opportunities provided by a new IOM report. JAMA 2012 5. Nissen SE, Wolski K. Effect of rosiglitazone on the risk 20. Misbin RI. Lessons from the Avandia controversy: a new
of myocardial infarction and death from cardiovascular causes. paradigm for the development of drugs to treat type 2 diabetes. N Engl J Med 2007;356:2457­71. [Erratum, N Engl J Med 2007;357: Diabetes Care 2007;30:3141­4.
21. Doshi P, Jefferson T, Del Mar C. The imperative to share
6. Hamburg M. Letter to the honorable Charles E. Grassley. clinical study reports: recommendations from the Tamiflu expe­
March 30, 2010 (http:/ online.wsj.com/public/resources/documents/ rience. PLoS Med 2012;9(4):e1001201.
22. Rising K, Bacchetti P, Bero L. Reporting bias in drug trials
7. Lincoff AM, Wolski K, Nicholls SJ, Nissen SE. Pioglitazone submitted to the Food and Drug Administration: review of pub­
and risk of cardiovascular events in patients with type 2 diabetes lication and presentation. PLoS Med 2008;5(11):e217. [Erratum, mellitus: a meta­analysis of randomized trials. JAMA 2007;298: PloS Med 2009;6(1):e17.] 23. Nissen SE. The rise and fall of rosiglitazone. Eur Heart J
8. Dormandy JA, Charbonnel B, Eckland DJ, et al. Secondary 2010;31:773­6.
prevention of macrovascular events in patients with type 2 dia­ 24. Sackner­Bernstein JD, Kowalski M, Fox M, Aaronson K.
betes in the PROactive Study (PROspective pioglitAzone Clinical Short­term risk of death after treatment with nesiritide for de­ Trial In macroVascular Events): a randomised controlled trial. compensated heart failure: a pooled analysis of randomized controlled trials. JAMA 2005;293:1900­5.
9. Graham D, Gelperin K. Memorandum to Mary Parks, re: bene­
25. Sackner­Bernstein JD, Skopicki HA, Aaronson KD. Risk of
fit­risk assessment of rosiglitazone vs. pioglitazone. October 7, worsening renal function with nesiritide in patients with acutely 2008 (http:/ www.finance.senate.gov/newsroom/chairman/release/ decompensated heart failure. Circulation 2005;111:1487­91. [Erra­ ?id=bc56b552­efc5­4706­968d­f7032d5cd2e4).
10. Mahoney K. Advisory committee clinical briefing document: 26. Nissen SE, Wolski K, Topol EJ. Effect of muraglitazar on
preliminary endocrine medical officer review of the RECORD death and major adverse cardiovascular events in patients with (Rosiglitazone Evaluated for Cardiac Outcomes and Regulation type 2 diabetes mellitus. JAMA 2005;294:2581­6.
of Glycemia in Diabetes) Trial, and update on cardiovascular safety 27. Solomon DH. Selective cyclooxygenase 2 inhibitors and car­
information from large clinical trials of rosiglitazone. 2010 diovascular events. Arthritis Rheum 2005;52:1968­78.
(http:/ www.fda.gov/downloads/AdvisoryCommittees/Committees 28. Curfman GD, Morrissey S, Drazen JM. Expression of con­
MeetingMaterials/Drugs/EndocrinologicandMetabolicDrugs cern: Bombardier et al., “Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid 11. Baucus M, Grassley C. Letter to the honorable Margaret A. arthritis,” N Engl J Med 2000;343:1520­8. N Engl J Med 2005;353:
Hamburg, MD. February 18, 2010 (http:/ www.finance.senate.gov/ 2813­4.
newsroom/chairman/release/?id=bc56b552­efc5­4706­968d­ 29. Nissen SE, Wolski K. Rosiglitazone revisited: an updated
meta­analysis of risk for myocardial infarction and cardiovascu­ 12. Woodcock J. Decision on continued marketing of rosigli­
lar mortality. Arch Intern Med 2010;170:1191­201.
tazone (Avandia, Avandamet, Avandaryl). 2010 (memorandum) 30. Kesselheim AS, Mello MM. Confidentiality laws and secrecy
(http://www.fda.gov/downloads/Drugs/DrugSafety/Postmarket in medical research: improving public access to data on drug DrugSafetyInformationforPatientsandProviders/UCM226959.pdf).
safety. Health Aff (Millwood) 2007;26:483­91.
13. Woodcock J, Sharfstein JM, Hamburg M. Regulatory action 31. Krumholz H. The FDA is faster: now let’s make it safer.
on rosiglitazone by the U.S. Food and Drug Administration. N Engl Outcomes. Forbes.com. May 16, 2012 (http://blogs.forbes.com/ 14. Food and Drug Administration, Center for Drug Evaluation 32. Punthakee Z, Bosch J, Dagenais G, et al. Design, history and
and Research. Joint meeting of the Endocrinologic and Meta­ results of the Thiazolidinedione Intervention with vitamin D bolic Drugs Advisory Committee and Drug Safety and Risk Evaluation (TIDE) randomised controlled trial. Diabetologia 2012; Management Advisory Committee. July 14, 2010 (transcript) 55:36­45.
(http:/ www.fda.gov/downloads/AdvisoryCommittees/Committees 33. Levine C. Has AIDS changed the ethics of human subjects
MeetingMaterials/Drugs/EndocrinologicandMetabolicDrugs research? Law Med Health Care 1988;16:167­73.
34. Food and Drug Administration. Guidance for industry: post­
15. Committee on Ethical and Scientific Issues in Studying the marketing studies and clinical trials — implementation of sec­
Safety of Approved Drugs, Institute of Medicine. Ethical issues in tion 505(o)(3) of the federal Food, Drug, and Cosmetic Act. April studying the safety of approved drugs: a letter report. Washing­ 2011 (http:/ www.fda.gov/downloads/Drugs/GuidanceCompliance ton, DC: National Academies Press, July 2010 (http://www.nap RegulatoryInformation/Guidances/UCM172001.pdf).
.edu/catalog.php?record_id=12948).
16. Committee on the Assessment of the U.S. Drug Safety Sys­
DOI: 10.1056/NEJMhle1207160
tem, Institute of Medicine. The future of drug safety: promoting Copyright 2012 Massachusetts Medical Society. Downloaded from nejm.org on August 24, 2012. For personal use only. No other uses without permission. Copyright 2012 Massachusetts Medical Society. All rights reserved.

Source: http://genomicaybioeconomia.org/images/Ethical%20Considerations.pdf