Science magazine

Clinical Trials Results Databases:
Hasty legislation mandating posting will not beenough to ensure public safety.
Unanswered Questions
Celia B. Fisher
Public outcry over pharmaceutical compa- results or data contradictory to a report submit- long-term health consequences of an adverse nies’ failure to report safety data from ted for review. However, there has been no dis- event, (iv) evidence of a clear causal relation cussion about whether the availability of large between the event and the product under investi- inhibitors has exerted pressure on industry, bodies of data from studies that may or may not gation, and (v) statistical power necessary to draw researchers, and policy-makers to ensure trans- have scientific merit will improve or distract conclusions regarding causal relations. Public parent and unbiased reports of clinical trials from the peer-review process. Moreover, the databases that include constantly updated tables results (1–3). One response receiving interna- absence of guidelines for how to include a large or summaries of adverse events in the absence of tional attention is creation of clinical trials reg- body of ancillary data in peer review of submit- such scientific understanding risk raising unmer- istries and results databases (4). In general, clini- ted manuscripts could compromise actual or per- ited public or health provider confidence or con- cal trials registries provide a public record of the ceived fairness of review. Results databases are cern. Health and safety protection of current and nature and eligibility criteria of newly initiated, not a substitute for systematic scientific peer potential research participants can be strength- ongoing, and closed trials. Results databases are ened through new guidelines for streamlining public postings of all clinical trials findings, The Public Library of Science (PloS) recently current safety data monitoring procedures that including potentially adverse side effects.
introduced an open-access journal called PloS emphasize reporting of product-relevant antici- Although there has been widespread debate on the Clinical Trials that promises to provide peer- pated adverse events and more timely review of rationale and criteria for registries, much of the reviewed data not affected by “the direction of dialogue (as well as legislation introduced in the results, size, or significance” of the trial (15). To Results databases are also not substitutes for U.S. Congress and in more than 20 states) fails to defray the cost of peer review and open access, the safety monitoring of commercially available address key questions about results databases (5).
journal charges a fee of $2500 upon acceptance of products. Safety concerns may not be apparent to inform patients about clinical trials in “Results databases are not a substitute for systematic
which they might participate (6, 7).
However, their purpose has expanded.
scientific peer review and scientific rigor.”
The International Committee of MedicalJournal Editors (ICMJE) sought to guardagainst “positive results bias” in publication of the article for publication with a nonspecified until a commercially approved product is studied clinical trials by limiting acceptance of manu- sliding scale for those lacking sufficient funds. It in a new patient population, until practitioners scripts to studies that had been entered into a reg- is too early to tell what effect this will have on the have prescribed it to a wider heterogeneous pop- istry at their inception (8). The ICMJE announce- science establishment’s ability to maintain and to ulation, or until consequences of product misuse ment prompted U.S. state and federal legislation monitor high standards of research design, analy- come to light. Although reporting of adverse [Fair Access to Clinical Trials (FACT) Act has been sis, and dissemination. However, lack of emphasis events is mandatory for marketed products, cur- introduced in the House of Representatives and on the direction of results or size, elements critical rently there is no process for evaluating safety Senate], proposals from international bodies such to good scientific method, risks diluting scientific data of postmarket products across independ- as the World Health Organization, and creation of ently conducted trials or for a national communi- voluntary industry registries (9–14). cation channel to encourage and facilitate physi- Participant and Patient Protections
cian reporting—to companies and the FDA—of reporting the results of completed clinical trials In the United States, subject protections are cur- serious, unanticipated, and significant adverse are also available from Web sites posted by the rently instituted through Institutional Review events in everyday practice. An active postmar- U.S. National Institutes of Health (7) and the Board (IRB) review of protocols before imple- ket monitoring interface is essential to long-term Pharmaceutical Research and Manufacturers of mentation and by safety and data monitoring understanding of how medical products benefit America (PhRMA) (12), and a handful of indus- boards during the conduct of clinical trials.
try registries, but they are not required. However, Some have argued that access to safety data from the FACT Act and some state legislation propose previous studies will help potential research sub- Health-Care Practice and Cost
jects evaluate the risks of enrolling in new stud- Even in large-scale clinical trials, the validity of ies. It is also hoped that public databases will results rests on representative sampling, dropout Implications for Clinical Trials Science
improve prescribing and treatment by helping rates, and replication. Practitioners and their pro- The current legislative proposals call for posting health-care providers and patients keep pace fessional organizations rely on a system of peer and open access to all raw or summative clinical with rapid advances. However, public databases review and FDA approval to help filter multiple trials data from successful studies, as well as sources of information about health products and those that have failed or produced equivocal Ethical and scientific evaluation of the poten- to establish consensus on standards of care. These tial for and significance of adverse participant data-filtering mechanisms are used by physicians reactions in a clinical trial requires: (i) an under- and hospitals to make decisions about health The author is the director for the Center for Ethics standing of the health status of the participant product purchases and by health management Education, Fordham University, and Marie Ward DotyProfessor of Psychology, Bronx, NY 10458, USA. E-mail: population, (ii) the types of side effects that were organizations to establish criteria for coverage of or were not anticipated, (iii) the immediate and prescription drugs and medical procedures. als results databases must consider implications for harmonization across government and private sponsors, state and federal legislation, global and national studies, and products that are approved or commercially available in some but not all from useful treatment regimens orprescriptions for off-label use of a Conclusions
Timely and transparent reporting of clinical trials results is essential to effective health- care decision-making and public confidence.
However, policies hastily crafted to assuage public concerns may produce unanticipated problems. Clinical researchers and the pharma- ceutical industry must take a leadership role, showing greater willingness to engage with other players. But it is not their responsibility alone.
this problem may be compounded by proposed have to carry errors and omissions (E&O) insur- Government policies must take into account pro- government actions calling for nonpromotional ance to cover this type of exposure, and products tections for public health and industry sustain- language in database postings that prohibit liability premiums could be adversely affected.
ability. Doctors and hospitals must also provide sponsors from providing conclusions about the Some have argued that mandatory posting of timely information. Continued dialogue among implications of the data for product efficacy and clinical trials results databases could place com- stakeholders is necessary to ensure that steps treatment decisions. Public dissemination of panies at risk of violating Securities and taken will enhance scientific and social responsi- decontextualized results summaries may also Exchange Commission Rules against hyping a bility and will contribute to the vitality and sus- exert pressure on the FDA to approve or withdraw drug under FDA review through “forward-look- tainability of clinical trials research.
products prematurely. The establishment of pro- ing statements” as defined by the Private fessional guidelines for the application of data- Securities Litigation Reform Act of 1995 (16). If base information for prescribing may help address a company posts positive results from the first 1. K. Gilpin, New York Times, 2 June 2004; (www.antidepressantsfacts.com/2004-06-02-NYT-NY- study completed and then completes a second Administrative resources required to maintain one that does not support the first, the company 2. Center for Drug Evaluation and Research, Food and Drug and monitor results databases may increase the might well be accused of misleading investors. Administration, "Questions and answers: FDA regulatory costs of health-care products. Pressure to access Premature posting of data from unapproved actions for the COX-2 selective and non-selective non-steroidal anti-inflammatory drugs (NSAIDs)"; constantly changing results databases may also compounds or off-label usage studies could (www.fda.gov/cder/drug/infopage/COX2/COX2qa.htm).
create an unreasonable medical “standard of hamper competition. Posting results on unap- 3. A. Gardner, HealthDay News, 29 August 2005; care,” which, in turn, can trigger medical malprac- proved compounds or new applications of mar- (www.healthday.com/view.cfm?id=526217).
4. J. Couzin, Science 307, 189 (2005). tice cases and increase professional liability insur- keted products could erode intellectual property 5. "Clinical trials registries and results databases white pa- ance rates. The relation of clinical trials results protections. Posting of premarket product trial per," Proceedings from the Fordham University Summit databases to product use and purchase must also results could reveal competitively valuable on Bio-Pharmaceuticals in the 21st Century: Responsibil-ity, Sustainability, and Public Trust, New York, 10 to 11 be considered. What if, for example, preliminary analyses or end points derived from intensive January 2005 (Fordham Univ., New York, 2005); results reported in a database supporting less negotiation with FDA and international regula- (www.clinicaltrialsethics.org/White%20Paper.htm).
costly products discourage hospitals from pur- tory authorities. For example, there are rules in 6. A.T. McCray. Ann. Intern. Med. 133, 609 (2000). chasing a proven but more expensive device? the United States, Europe, and other countries 7. See also (www.clinicaltrials.gov/).
8. ICMJE, "Clinical trial registration: A statement from the that allow generic manufacturers and other International Committee of Medical Journal Editors." results databases? Will a single study indicating applicants to obtain approval through abridged N. Engl. J. Med. 351, 1250 (2004); (www.nejm.org).
a negative result of a postmarket product dis- procedures by referencing safety and efficacy 9. S. 470—The Fair Access to Clinical Trials (FACT) Act (February 2005, 109th Congress); (http://olpa.od.nih.
courage health-care plans from covering its use? data in the public domain. Public posting of raw gov/tracking/109/senate_bills/session1/s-470.asp).
Might health-care insurers pressure physicians data or full study reports could be used as a basis 10. World Health Organization International Clinical Trials to switch to less costly medications on the basis for such applications, thereby compromising Registry Platform; (www.who.int/ictrp/en/).
11. International Federation of Pharmaceutical Manufactur- of preliminary trials posted on a results data- regulatory exclusivity for marketing authoriza- ers and Associations, "Global industry position on disclo- base? To provide adequate answers to these tion holders, hurting investor return, and dis- sure of information about clinical trials" (6 January questions, health-product stakeholders need to push for cost-effectiveness studies and guide- To ensure scientific integrity, advance public 12. The Pharmaceutical Research and Manufacturers of lines for the use of databases in health-care prac- health, and sustain health-care innovation, some America (PhRMA) Clinical Study Results Database; U.S.-based and international organizations have proposed creating a “blind” data repository 13. American Medical Association offers guidelines for clini- cal trial registry; (www.ama-assn.org/ama/pub/ Industry Sustainability
linked to a clinical trials registry. In this model, Sponsors are concerned that failure to post results investigators and/or sponsors would be required 14. GeMCRIS: Genetic modification clinical research infor- could be construed as sponsor fraud or negli- to submit their data on project completion, but mation system; (www.gemcris.od.nih.gov/).
gence; product liability actions could become release into a public database would coincide 15. In 2006 (http://clinicaltrials.plosjournals.org); now avail- able through (www.plos.org/journals).
more frequent. At the same time, manufacturers with article submissions and/or approval by 16. "Posting trial data could run afoul of law against mis- that do publicize preliminary product findings on FDA or an international body (5, 11).
leading investors," FDA Week (12 November 2004); mandatory databases might be accused of fraud- Clinical research on drugs, biologics, and (http://insidehealthpolicy.com) (by subscription).
17. World Health Organization, International clinical trials ulently promoting an insufficiently tested prod- medical devices is a multisite, multistate, global registry platform (ICTRP); (www.who.int/ictrp/en/).
uct. Posting of results on databases may create enterprise that requires a solution that is national undue investor hype. Product manufacturers may and global. All legislated or voluntary clinical tri-

Source: ftp://quatramaran.ens.fr/pub/godfroy/articles/Fisher_2008.pdf