By Karl E. Peace
Georgia Southern University
March 17, 2008 | The article New England IRB on Ethical Issues in Recruiting Patients, in the served as a reminder to those involved with the execution of clinical trials of the importance of exercising due diligence to ensure that all aspects of clinical trial conduct are ethical.
I would like to bring to the attention of readers of eCliniqua some of my thoughts about making the informed consent process more ethical (see refs. 1, 2, and 3). Many are aware, particularly in Phase III confirmatory proof of efficacy trials, that the determination of sample size carries with it an efficacy imperative. That is, in order to confirm a question regarding efficacy of a compound, it is imperative that the sample size of the trial be large enough to provide a priori a high probability (large power) that the question will be answered.
Sample size determination carries with it an ethical imperative as well. Before I would agree to participate in a clinical trial, I would want to know first: that the trial was needed to answer a medically important question (and what that question was), and second: what the likelihood was -- for the number of planned participants -- that the question would be answered. Including this second issue in the informed consent document in the protocol, would, I believe, make the informed consent process more ethical.
It is easy to argue that for small trials, we know with relatively large confidence before such trials are conducted, that all we’ll be able to say after trial completion is that “we failed to confirm or refute the question.” This may be observed from the table, which summarizes 90 percent confidence intervals (CIs) on the difference between -- and the ratio of -- two treatments, and the length of the confidence intervals for various sample sizes ranging from 10 to 200 per group.
We observe that all confidence intervals on the difference between treatments, cover 0. Therefore, we cannot conclude that the treatments are different (at the one-sided 5 percent level of significance). However, with the possible exception of the interval corresponding to 200 patients per group, all confidence intervals are so wide that we cannot claim that the treatments are similar. So if at the outset of a trial, one knows that it is highly likely that the outcome will be summarized as: “We can’t say that the test treatment is different from the standard; neither can we say that it is similar,” why do the trial? Unless of course, the primary reason for conducting the trial is pilot in nature: to gain experience with the attendant methods, to gather data so that a confirmatory follow-up trial may be conducted, etc. In that case, patients should be fully informed prior to their decision regarding trial participation.
Having argued that it should be ethically imperative to include a statement in the informed consent document regarding the likelihood that the trial will answer the question or objective, I acknowledge that writing such a statement geared to -- at most -- an 8th grade reading level, is most challenging, since even many highly educated people find the concept of power difficult to comprehend.
References:
1. Peace, KE: "Some Thoughts on the Biopharmaceutical Section and Statistics"; ASA Joint Sesquicentennial Meetings, Publication of the American Statistical Association; 98-105; Arlington, Va. August, 1989.
2. Peace, K.E.; "An Ethical Issue Concerning Informed Consent in Clinical Trials." Biopharmaceutical Report. 1 (2); 1992.
3. Peace, K. E. (2005): Making Informed Consent More Ethical. Drug Information Forum, 5(4), 26-27.
Karl E. Peace, Ph.D., is Georgia Cancer Coalition Distinguished Cancer Scholar and Senior Research Scientist and Professor, Jiann-Ping Hsu College of Public Health, Georgia Southern University.
------------------------
This story first appeared in eCliniqua,one of Bio-IT World’s free e-newsletters. Subscribe here.