Commentary: Ethics in Clinical Trials
By Mirkamal Tolend and Samia Tasmim
Clinical trials aim to elucidate the viability of new treatments or interventions on humans, and are indispensable to the progress of medical science. Although double-blind, randomized, controlled trials are the pinnacle of clinical trials, these are not always practical or ethical to conduct for every clinical question. As such, the interplay of ethics, uncertainty, and the need for high quality scientific evidence makes clinical trials unique.
The importance of ethics in clinical research is seen in documents as old as the Hippocratic Oath. Unfortunately, it took a dark period in research history, involving atrocities such as the use of war prisoners for human experimentation in the Second World War, to cement the emphasis of clinical trial ethics. To ensure such unethical research does not recur, explicit codes of ethical conduct were outlined in the Nuremberg Code, the Declaration of Helsinki, and the Belmont Report.1 Currently in Canada, all federally funded research involving humans or human biologic material must strictly adhere to the Tri-Council Policy Statement.
One of the longest-running controversial clinical trials was the Tuskegee study of syphilis, which ran for 40 years from 1932 to 1972 in Tuskegee, Alabama.2 600 African American men (with or without syphilis) were simply told the purpose of the study was to evaluate “bad blood”. They acquiesced to the study in exchange for free food, burial insurance and medical tests. Notably, they did not receive treatment for syphilis, even after penicillin became the standard of treatment starting in 1947. Astonishingly, this study, which was supposed to run for six months, went on for 40 years. It finally ended when the Associated Press covered the study and an advisory committee ruled that it was ethically unacceptable. Although the participants and their families were later compensated financially and given medical assistance, this trial is an example of the horrific trials that some groups were subjected to.2 It is also one of the reasons why some of these historically mistreated groups are skeptical of medical practitioners and researchers to this day, leading to an additional barrier to them seeking medical care, which in itself is another major issue affecting health.
As the potential benefit of clinical research is usually experienced by future patients rather than by the study participants, ideal participants would be motivated purely by altruism to consent to a study. Policies on informed consent, patient safety monitoring, and careful, peer-reviewed assessment of the study’s scientific and methodological rigor are all important to prevent patient exploitation. The reason for participating in a clinical study may not be purely altruistic when there is the problem of undue inducement, meaning that the process of providing voluntary informed consent is influenced by the financial benefits of participating. For some patients, these benefits may outweigh the risks associated with the study.3 Certain patient groups, such as those who do not respond to current standard-of-care treatments or those from low socioeconomic backgrounds, who would be unable to otherwise obtain either the standard or the experimental treatment may be considered especially vulnerable with regard to participating in a clinical trial.4 To avoid exploitation of these study participants, it is important to be aware of the potential reasons patients may expose themselves to treatments of unproven effectiveness.
Certain biases are important to avoid in clinical trials.5 One prominent example is selection bias, which occurs when the method that is used to sample the cases into the study cohorts leads to a baseline difference between the cohorts. This may differentially affect study results if such a difference confounds the study outcome. Various design-based and analysis-based techniques exist to control for the effect of the confounder variable. These safeguards include matching, restriction, stratification, and multivariable adjustment. However, these techniques can only adjust for known and measured confounders. To protect against bias from known as well as potential unknown confounders, randomization is applied. Hence, a proper randomization is an important design feature in clinical trials to improve internal validity. At times randomization is not possible, due to practical or ethical reasons. To account for this, properly controlled observational studies with large sample sizes have become increasingly rigorous in design, yielding high quality evidence for clinical questions not amenable to randomized controlled trials. The requirement of scientific review board approval before the start of clinical trials is important to ensure that patients are not enrolled in studies with obvious flaws in internal validity, the results of which will be of poor quality and not worth the risk of exposing patients to unnecessary risk.
Ethical issues may affect the interpretation of data. Moving a patient from the treatment group to the control group, or vice versa, can be ethically necessary; if a patient cannot tolerate the side effects of a new treatment, for instance. Intention to treat analysis, as opposed to analysis by treatment received, is an approach to analyzing these cases that “cross over”. The outcome of a patient is analyzed according to the group the patient was initially assigned, i.e., if a treatment patient crosses over to the control cohort, their outcome still belongs to the treatment group data. This method preserves the benefit of the randomization to control for the confounders, which is important because patients who need to cross over may have baseline differences in confounding variables. However, the negative impact this will have on the data is an underestimation of the treatment effect size, i.e., the difference in the outcome between the treatment and control groups. An underestimated effect size may lead to falsely accepting the null hypothesis about the treatment effectiveness, which is an extreme manifestation of the ethical obligation of the clinician overriding the ethical obligation of the scientist. Having a more conservative estimate of the clinical trial result is often favored, as it preserves the control for confounding, and allows participant safety to be better ensured.
Ethical issues also affect the generalizability of results. For instance, during a trial, if an interim analysis shows that treatment of interest shows a significant benefit on early available data, the continued use of placebos or the status quo treatment may be considered unethical.1 However, this assumes that the treatment benefit seen in the interim analysis on early data will persist in the long-term and/or across a greater sample. This assumption may not hold up if adverse effects are difficult to predict for the long term, or if the definition of improvement from the perspective of the patient changes over the course of extended treatment. Furthermore, generalizability of clinical trials can be challenging. Deciding the eligibility criteria in clinical trials can range from two extremes: a pragmatic approach, which is less restrictive and more similar to daily clinical practice, or the explanatory approach, which is highly restrictive to the specific disease presentation under question and occurs in controlled settings. Ethical considerations may favor the explanatory approach, since patients who are less likely to benefit from the treatment are excluded. A problem may arise when results from an explanatory trial are over-generalized to less controlled scenarios and to patients who differ from the clinical trial patients in terms of risk factors, comorbidities, treatment adherence.
Ethics in clinical trials is a broad topic. This brief overview has hopefully sparked an interest in readers to dig deeper and get familiarized with the various ethical considerations in planning, conducting, analyzing, and interpreting clinical trials.
- Nardini The ethics of clinical trials. Ecancermedicalscience; 2014 Jan [cited Sep 5] Available from: doi:10.3332/ecancer.2014.387.
- Centers for Disease Control and Prevention. https://www.cdc.gov[cited Sep 5] Available from: https://www.cdc.gov/tuskegee/timeline.htm
- ArnasonG, Van Niekerk A. Undue fear of inducements in research in developing
Camb Q Healthc Ethics. 2009 Apr [cited Sep 5] Available from: doi:
10.1017/S0963180109090215. PubMed PMID: 19250565.
- Weigmann The ethics of global clinical trials: In developing countries, participation in clinical trials is sometimes the only way to access medical treatment. What should be done to avoid exploitation of disadvantaged populations? EMBO Reports. 2015 Apr [cited Sep 5] Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4428044/.
- JadadAR, Enkin Bias in Randomized Controlled Trials, in Randomized Controlled Trials: Questions, Answers, and Musings. Second Edition, Blackwell Publishing Ltd, Oxford, UK; 2007 [cited Sep 5] Available from: doi: 10.1002/9780470691922.ch3