Editor's Choice | This Week in BMJ | Press releases



BMJ No 7102 Volume 315

Letters Saturday 26 July 1997


Informed consent

Informed consent is not always obtained in United States

Editor,
A code of silence and a spirit of denial surround one of the oldest and most perplexing conundrums in medical research: how to recruit large numbers of fully consenting subjects. So it was refreshing to note the BMJ's pioneering willingness to devote much of the issue of 12 April to questions of informed consent. 'Rather than restrict the debate to ourselves,' as the editor, Richard Smith, put it, the BMJ dared to display publicly what institutional review boards and healthcare professionals usually handle with discretion or denial.

Our own analysis, published in The (Cleveland, Ohio) Plain Dealer and other American newspapers late last year, suggests the extent to which questions of consent persist.(1) This is despite the American government having apologised formally recently to survivors of a study of the natural course of syphilis in black men in Tuskegee, Alabama.(2) President Clinton has diverted the mandate of the National Bioethics Advisory Commission (which was formed to grapple with issues such as informed consent) to take up the more sensational if speculative topic of human cloning. Headlines have overtaken the issue before; in 1995, on the day that the Advisory Committee on Human Radiation Experiments warned of questions of consent in contemporary medical research, public interest focused on the verdict in the case of O J Simpson. In the past few months, disclosures in Augusta, Georgia, and Orange County, California, raised questions of consent, illustrating that the topic is more than a matter of distant history.(3, 4)

Since 1977 the US Food and Drug Administration (FDA) has conducted 4154 inspections of clinical trials. Our analysis of records of those inspections showed that 53% of the investigators were cited by the FDA for failing clearly to disclose the experimental nature of their work. In 46 trials involving at least 1000 men, women, and children, drugs were tested without any written evidence that subjects had consented.

We also found evidence that the US government, which makes annual payments to survivors of the study in Tuskegee, has sponsored experiments on unsuspecting subjects well into the 1990s. Among our case studies were tests by the Centers for Disease Control and Prevention, begun in 1990 in Los Angeles, of the immunogenicity and efficacy of the Edmonston-Zagreb measles vaccine, which the centres knew had earlier caused excess mortality in Africa; and a study in 1991 of hepatitis A vaccine, conducted on a Sioux reservation, in which the letterhead on the consent form implied an established prevention programme rather than the safety and efficacy trial it was.

Nor does the record overseas appear any better. Foreign and internal FDA documents that we reviewed contained accounts of fraud, concealed side effects, experiments diverging from protocols, and questions of consent. Consent forms were incomplete or inadequate in 65 of 137 inspections by the FDA of trials conducted in countries other than the US. In Canada, consent forms were inadequate in 21 of 36 inspections. Verifiable scientific data were missing from 53% of the international research submitted to support a US new drug application.

Our analysis of FDA data also showed that internal review boards, the front line in protecting test subjects, cannot be counted on to ensure that people know they are being used in medical research. At 942 internal review boards between 1990 and 1996, FDA inspectors found multiple violations: no evidence of continued safety monitoring (at 20% of the review boards); no copies of consent forms, injury reports, or protocols (19%); and that patients were not clearly told when procedures were experimental (16%), not offered proved alternative treatments (13%), not informed of expected risks and pain (10%), and not told of likely benefits (6%).

Some people-among them US senator John Glenn and Gary Ellis, director of the office for protection from research risks at the National Institutes of Health-have argued for stronger controls. Glenn has proposed legislation to close many regulatory gaps. But until more researchers and their publications are willing to come out publicly about what they know and what they believe should be done to better advance medical research without sacrificing-and documenting-consent of the fully informed test subject, prospects appear dim for meaningful reform. And questions of consent can be expected to recur with disturbing frequency.

Keith Epstein
Investigative reporter

Bill Sloat
Investigative reporter

The (Cleveland) Plain Dealer,
Washington Bureau,
Washington, DC,
USA

References

1 Epstein K, Sloat B. Drug trials: do people know the truth about experiments? The (Cleveland) Plain Dealer 1996 Dec 15-18:a1. [Articles and charts are posted at http://www.spj.org/sdxawards/03invrptg/index.htm#top or can be requested from: epstein@dgsys.com]

2 Harris J F, Fletcher M A. Six decades later, an apology. Washington Post 1997 May 17:a1.

3 Teegardin C, Whitt R. FDA inspection finds violations in Augusta studies. Atlanta Constitution 1997 May 11:a3.

4 Marsh B, Romney L. Hospital accused of violating consent rules. Los Angeles Times 1997 May 30:a3.


Home | Current issue | Past issues | Classified ads | Career Focus | Feedback
Collections | About this site | About the BMJ | BMA | Medline