The American Journal of Medicine
Volume 117, Issue 11 , Pages 842-845, 1 December 2004

Association of funding and findings of pharmaceutical research at a meeting of a medical professional society

  • Thomas E. Finucane, MD

      Affiliations

    • Division of Geriatric Medicine and Gerontology, Department of Medicine, Johns Hopkins School of Medicine, Baltimore, Maryland
    • Corresponding Author InformationRequests for reprints should be addressed to Thomas E. Finucane, MD, Division of Geriatric Medicine and Gerontology, John R. Burton Pavilion, 5505 Hopkins Bayview Circle, Baltimore, Maryland 21224
    • ,
  • Chad E. Boult, MD, MPH, MBA

      Affiliations

    • Division of Health Services Research, Department of Health Policy and Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland.

Received 2 March 2004; accepted 20 May 2004.

Article Outline

Purpose

To evaluate the association between funding and findings of pharmaceutical research presented at an annual meeting of a medical professional society.

Methods

We reviewed the abstracts of all papers and posters presented at the annual meeting of a medical professional society. Two independent raters classified each study of a drug (n = 48) as either positive (favoring the drug studied) or negative, and as either funded by a pharmaceutical company or not. We computed κ and chi-squared statistics to evaluate the agreement between the raters, as well as the association between the results and the sponsorship of the study.

Results

Thirty studies of drugs (63%) were supported by pharmaceutical companies, all of which reported positive results. Of the 18 studies (37%) not supported by pharmaceutical companies, 67% reported positive results. The association between pharmaceutical funding and positive findings was statistically significant (P = 0.0007).

Conclusion

At this scientific meeting, research funding from pharmaceutical companies was associated with study findings that supported the use of drugs marketed by pharmaceutical sponsors. We emphasize further study of this relation and suggest three principles—full disclosure, policies against “outcome bias,” and educational opportunities—that may help manage industry-academia conflicts of interest that could otherwise jeopardize the credibility of pharmaceutical research presented at scientific meetings.

 

Many studies of commercial products are submitted for possible presentation at the annual meetings of medical professional societies. Those that are accepted and presented earn scientific credibility and are often cited in commercial promotional materials designed to influence the prescribing behavior of physicians (1). The results of one to two thirds of such presentations are later published in journals. Similar to articles published in journals, presentations at meetings may be vulnerable to scientific bias resulting from conflict between scientific impartiality and the financial interests of the study’s funders.

Intense debate in the past decade has helped to define several potential conflicts of interest confronting academic institutions (2, 3) and scientific investigators (4, 5, 6, 7, 8, 9) who have financial relationships with companies that market the products that they study. A systematic review of 1664 English-language articles and a meta-analysis of data from 37 articles described the extent of such financial relationships as widespread, the effects on the quality as slight, and the outcomes of company-sponsored research as substantial (10, 11). Peer-reviewed articles about company-sponsored studies are more likely to report outcomes favorable to the commercial sponsors than are articles about non–company-sponsored studies (odds ratio [OR] = 3.6; 95% confidence interval [CI]: 2.6 to 4.9) (10). Similarly, a review of 3351 articles not restricted to the English language reported that studies sponsored by pharmaceutical companies were more likely to report outcomes favoring the sponsor than were studies with other types of sponsors (OR = 4.1; 95% CI: 3.0 to 5.5) (11).

Although there is extensive literature dealing with conflict of interest and potential scientific bias in journal articles, there is a dearth of quantitative studies addressing the extent of industry support for research presented at scientific meetings of medical professional societies, as well as examining the association of such support with the quality or outcomes of the studies presented. This lack may be attributable, at least in part, to the paucity of available data on such presentations. Nevertheless, thousands of studies, many of which are industry sponsored, are presented annually as lectures and posters at scores of medical professional society meetings. We therefore undertook the present study as a preliminary investigation of the possible association between funding by pharmaceutical companies and the outcomes of drug studies presented at the annual scientific meeting of one medical professional society.

Back to Article Outline

Methods 

We reviewed the abstracts of all the reports presented as posters or lectures at the annual scientific meeting of a clinically oriented U.S. medical professional society. We selected for analysis all abstracts that reported results about the effectiveness or safety of drugs. Two experienced raters independently classified each abstract according to its findings. A third experienced rater was available to adjudicate any discordant classifications. The study was classified as “positive” if the abstract concluded that a drug was safe, effective, or superior to another drug in safety or effectiveness, or “negative” if the conclusion was that a drug caused unacceptable adverse effects, failed to produce hypothesized therapeutic effects, or led to adverse events or therapeutic effects that were statistically indistinguishable from or less desirable than an alternative drug.

To classify each study with regard to industry support, the raters classified each abstract according to two criteria. They classified a study as industry supported if there was acknowledgment of receipt of a grant from a pharmaceutical company, or if at least one author was affiliated with a pharmaceutical company or a private research corporation that was supported by the pharmaceutical industry. To clarify the source of support for the 17 studies that met neither criterion, their authors were contacted by telephone or e-mail and asked “What organization funded the study?” If the authors denied having received pharmaceutical support (n = 9) or did not respond (n = 4), the studies were classified as not industry supported.

This study was “exempt” from detailed review by the Institutional Review Board of the Johns Hopkins Bloomberg School of Public Health.

Data analysis 

Agreement between the raters was assessed with the κ statistic. The chi-square test was used to evaluate the association between study results and sponsorship.

Back to Article Outline

Results 

Of the 619 abstracts submitted for possible presentation at the meeting, the conference reviewers rated 526 as scientifically suitable for presentation: 481 as posters and 45 as lectures. Of the 520 studies presented (six were withdrawn by their authors for unspecified reasons), 48 reported results of studies on the safety or effectiveness of medications. Most were nonrandomized observational studies (n = 20) or randomized clinical trials (n = 20).

The level of agreement between the two raters in classifying the findings (positive or negative) and source of funding (industry supported or not) for the 48 studies was complete (κ = 1.0). Thirty (63%) of the 48 studies were industry supported, and all 30 reported results that were favorable to a drug marketed by the pharmaceutical sponsor. Eighteen (37%) of the studies were not industry supported; of these, 12 (67%) had positive results and six (33%) had negative results. The association between industry support and the reporting of results favorable to a product of the sponsor was statistically significant (P = 0.0007). The same association was evident among subgroups composed of randomized clinical trials (P <0.05) and observational studies (P <0.05).

Of the 30 industry-supported studies, three (10%) acknowledged grant support in their abstracts. Another 23 studies (77%) named a pharmaceutical company as the institutional affiliation of one or more of the authors. Four studies (13%) named a private research corporation or university as the institutional affiliation of one or more authors, which, according to the authors, received study support from the manufacturer of the drug studied.

Back to Article Outline

Discussion 

These data show a strong association between pharmaceutical funding and the findings of studies of drugs presented at the annual scientific meeting of a medical professional society. All of the industry-supported studies reported results that were favorable to the sponsor; many will probably be published later as articles in scientific journals.

Three well-known forms of bias may have contributed to the findings reported here. “Positive outcome” bias is the tendency for reviewers and editors to select for presentation or publication studies that report positive findings. This bias can affect both the selection of abstracts for presentation at scientific meetings (12) and the selection of manuscripts for subsequent publication in scientific journals (13).

A second set of biases may result from the efforts of pharmaceutical companies to promote their products, which may be in conflict with the scientific impartiality among institutional decision-makers (3) and individual investigators (10). These efforts may include supporting biased study designs (e.g., those that include nonequivalent comparison interventions) (14, 15), controlling access to and analysis of study data, and censoring or “spinning” the results (16, 17, 18).

A third potential bias may stem from conflicts between investigators’ duty to conduct research objectively and their desire for professional recognition (e.g., by granting agencies and academic promotion committees) and financial reward. In some cases, the scientific objectivity of investigators may be threatened by their opportunities to publish “positive studies” and to receive industry funding, equipment, supplies, stock options, consultancies, and membership in speakers’ bureaus and advisory boards.

An alternate explanation for the observed association between funding and findings involves the sequence in which drugs are tested. For instance, the findings presented at meetings of professional societies may come from studies of drugs that have shown effectiveness and safety in premarketing trials, while further studies of less successful drugs may have been curtailed.

Factors that limit the inferences that can be drawn from our study include the sample size (data from only one meeting of one professional society), the inability to rule out definitively pharmaceutical support of six negative studies, and incomplete information on the quality of the studies that were presented and on the abstracts that were rejected or withdrawn from presentation at the meeting.

A strong relation between pharmaceutical support and the outcomes of drug studies pervades the printed scientific literature (10). Therefore, our finding of an analogous relation at the annual meeting of one medical professional society suggests the possibility that the same association may exist at other meetings. To help manage the conflicts of interest inherent in research conducted collaboratively by industry and academia, professional societies (19, 20), academic organizations (21), governmental agencies (22), journal editors (23), and industry representatives (24) have published general guidelines. For example, some journals refuse to publish such collaborative studies if the investigators lacked complete access to all of the data. Similarly, the American College of Physicians–American Society of Internal Medicine’s 2002 position paper on organizational issues related to physician-industry relations recommends that medical professional societies “develop and enforce explicit policies that preserve the independent judgment and professionalism of their members and maintain the ethical standards and credibility of the society” and “conduct professional meetings in a highly principled manner” and that product promotion activities (e.g., displays and sponsored lectures) be located and identified as “independently organized and separate from official scientific sessions” (20).

A recent review revealed, however, that fewer than half of U.S. academic institutions, federal sponsors of research, and peer-reviewed journals enforce the guidelines promulgated to help manage these complex relationships (10). The degree to which medical professional societies adhere to such guidelines in selecting research for presentation at their annual meetings has not been reported.

If the association we observed between industry funding and the reporting of positive findings at scientific meetings is widespread, it could have adverse consequences for consumers, physicians, insurers, and professional societies. We propose three principles that medical professional societies may wish to consider adopting to maintain the integrity of presentations about commercial products at their scientific meetings.

Full disclosure 

In the review process, the authors, reviewers, and judges of studies of commercial products would disclose their financial relationships with companies that market the products studied (25). Abstracts would include addenda stating who designed the study, collected and managed the data, conducted the analyses, and reported the results. Reviewers and judges would acknowledge actual and potential conflicts of interest or recuse themselves from the selection process when appropriate. Studies selected for presentation as lectures or posters would disclose the authors’ financial and operational relationships with the companies that market the products studied. Medical professional societies would disclose their financial relationships with private companies as well. Such disclosure may help to manage conflicts of interest, although by itself such disclosure would be insufficient. For maximal effect, disclosure would be conducted in the context of well-publicized high ethical standards, clear distinctions between acceptable and unacceptable behaviors, systematic monitoring of adherence, and sanctions to ensure compliance (26).

Policies against “outcome bias” 

Professional societies would adopt and communicate to their members, reviewers, judges, and editors clear policies promoting acceptance of abstracts on the basis of scientific merit, without regard for whether the study outcomes were positive or negative.

Educational opportunities 

Professional societies would offer lectures, panel discussions, and debates at their annual meeting to acquaint their members with the challenges of recognizing and understanding conflict of interest and scientific bias in research findings that may influence their teaching and patient care.

Back to Article Outline

Acknowledgment 

The authors gratefully acknowledge the assistance of David McCaffrey in collecting, classifying, and managing the data.

Back to Article Outline

References 

  1. Wyatt J . Use and sources of medical knowledge . Lancet . 1991;338:1368–1373
  2. Angell M . Is academic medicine for sale? . N Engl J Med . 2000;342:1516–1518
  3. Johns MM , Barnes M , Florencio PS . Restoring balance to industry-academia relationships in an era of institutional financial conflicts of interest (promoting research while maintaining trust) . JAMA . 2003;289:741–746
  4. Thompson DF . Understanding financial conflicts of interest . N Engl J Med . 1993;329:573–576
  5. Stelfox HT , Chua G , O’Rourke K , Detsky AS . Conflict of interest in the debate over calcium-channel antagonists . N Engl J Med . 1998;338:101–106
  6. Korn D . Conflicts of interest in biomedical research . JAMA . 2000;284:2234–2237
  7. DeAngelis CD , Fontanarosa PB , Flanagin A . Reporting financial conflicts of interest and relationships between investigators and research sponsors . JAMA . 2001;286:89–91
  8. Friedberg M , Saffran B , Stinson TJ , et al.   Evaluation of conflict of interest in economic analyses of new drugs used in oncology . JAMA . 1999;282:1453–1457
  9. Bodenheimer T . Uneasy alliance—clinical investigators and the pharmaceutical industry . N Engl J Med . 2000;342:1539–1544
  10. Bekelman JE , Li Y , Gross CP . Scope and impact of financial conflicts of interest in biomedical research (a systematic review) . JAMA . 2003;289:454–465
  11. Lexchin J , Bero LA , Djulbegovic B , Clark O . Pharmaceutical industry sponsorship and research outcome and quality (systematic review) . BMJ . 2003;326:1167–1170
  12. Callaham ML , Wears RL , Weber EJ , et al.   Positive-outcome bias and other limitations in the outcome of research abstracts submitted to a scientific meeting . JAMA . 1998;280:254–257
  13. Scherer RW , Dickersin K , Langenberg P . Full publication of results initially presented in abstracts. A meta-analysis . JAMA . 1994;272:158–162
  14. Safer DJ . Design and reporting modifications in industry-sponsored comparative psychopharmacology trials . J Nerv Ment Dis . 2002;190:583–592
  15. Johansen HK , Gotzsche PC . Problems in the design and reporting of trials of antifungal agents encountered during meta-analysis . JAMA . 1999;282:1752–1759
  16. Rennie D . Thyroid storm . JAMA . 1997;277:1238–1243
  17. Nathan DG , Weatherall DJ . Academia and industry (lessons from the unfortunate events in Toronto) . Lancet . 1999;353:771–772
  18. McCarthy M . Company sought to block paper’s publication . Lancet . 2000;356:1659
  19. Coyle SL . Physician-industry relations. Part 1: individual physicians . Ann Intern Med . 2002;136:396–402
  20. Coyle SL . Physician-industry relations. Part 2: organizational issues . Ann Intern Med . 2002;136:403–406
  21. Association of American Medical Colleges, Task Force on Financial Conflicts of Interest in Clinical Research. Protecting subjects, preserving trust, promoting progress, II: principles and recommendations for oversight of an institution’s financial interests in human subjects research. Available at: http://www.aamc.org/members/coitf/2002coireport.pdf. Accessed May 12, 2004.
  22. Office of Human Research Protections. Financial relationships in clinical research: issues for institutions, clinical investigators, and IRBs to consider when dealing with issues of financial interests and human subject protection. Available at: http://ohrp.osophs.dhhs.gov/nhrpac/mtg12-00/finguid.htm. Accessed May 12, 2004.
  23. International Committee of Medical Journal Editors. Uniform requirements for manuscripts submitted to biomedical journals. Available at: http://www.icmje.org. Accessed May 12, 2004.
  24. Pharmaceutical Research and Manufacturers of America. PhRMA principles on conduct of clinical trials and communication of clinical trial results. Available at: http://www.phrma.org/publications/policy//2003-11-20.871.pdf. Accessed May 12, 2004.
  25. Alpert JS , Furman S , Smaha L . Conflicts of interest (science, money, and health) . Arch Intern Med . 2002;162:635–637
  26. Rodwin MA . Physicians’ conflicts of interest. The limitations of disclosure . N Engl J Med . 1989;321:1405–1408

PII: S0002-9343(04)00583-2

doi:10.1016/j.amjmed.2004.05.029

The American Journal of Medicine
Volume 117, Issue 11 , Pages 842-845, 1 December 2004

Article Tools