Tuesday, July 16, 2002
By Merritt McKinney
Clinical trials that are too small to reliably prove whether or not a treatment works are usually unethical and should only be conducted under special circumstances, according to a new report.
"The marginal value of the information that may be gained by conducting small clinical trials is insufficient to justify exposing patients to the common risks and burdens of research," Scott D. Halpern, of the University of Pennsylvania in Philadelphia, told Reuters Health.
"Ethics review boards, research funding agencies and medical journal editors should maintain strict requirements for adequate research methodology, including appropriate statistical power, in order for any clinical trial to be approved, funded or published," he said.
But Halpern stressed that the public "should not be deterred from participating in clinical trials, because they remain critical to medical progress."
Still, people who are thinking about enrolling in a trial may want to ask the investigators "whether the trial will directly, or only indirectly, contribute to such progress," he said.
In a report in the July 17th issue of The Journal of the American Medical Association, Halpern and his colleagues state that clinical trials should be large enough to have an 80% likelihood of detecting a clinically significant effect.
Since individuals often participate in clinical trials out of a sense of altruism, if a trial does not have this level of statistical power, then it is unethical to conduct such a study without first informing participants of the limitations of the trial, according to the researchers.
Halpern and his associates allow that there are two situations when an underpowered study is ethically acceptable. They note that in the case of treatments for rare diseases, recruiting an adequate number of participants may be impossible. Such an underpowered trial is acceptable, Halpern's team concludes, as long as investigators design the study in such a way that the findings can later be pooled with the results of other similar studies.
Another exception, according to the researchers, is an early-phase trial of a drug, medical device or other treatment. In both cases, however, investigators should inform participants of the limitations of the study, Halpern and his colleagues advise.
How large does a study need to be? When it comes to clinical trials, one size does not fit all. The number of participants that constitute an adequate number varies from study to study, Halpern explained. Still, he said that investigators should be able to calculate the minimum number of patients needed to achieve clinically important findings.
In a second report published in the same journal, Drs. Ann H. Partridge and Eric P. Winer of Dana-Farber Cancer Institute in Boston, Massachusetts discuss the implications of informing study participants of the results of clinical trials.
It may seem common sense that people who participate in medical studies would be informed of the results of the trials once they are completed. That is often not the case, Partridge told Reuters Health in an interview.
Partridge emphasized that patients are routinely informed of results when the findings will affect their care. For instance, all women participating in a recent trial of hormone replacement therapy were notified once the investigators discovered the increased risks associated with the hormones.
But in most trials, "as a rule, people don't get feedback except anecdotally," according to Partridge.
Sharing the results of studies with participants may improve patient-physician communication and even lead to greater participation in trials, but the process is not as simple as it may seem, Partridge and Winer point out in the report.
In the interview, Partridge noted that some patients may not want to find out the results, particularly if they received a treatment that was less effective than hoped. And in the case of some trials that take years to complete, some patients may have died, she said.
To find out more about the barriers to sharing information, as well as the benefits of doing so, Partridge and colleagues are now surveying a group of doctors and nurses who run cancer trials. The researchers also plan to conduct a study of sharing results with individuals who have participated in cancer clinical trials.
SOURCE: The Journal of the American Medical Association 2002;288:358-365.
Copyright 2002 Reuters