Non-Randomized Study

Because non-randomized (NR) studies do not involve randomization, participants can select the group they join or some other factors leads to their group selection. For example, a new program is implemented in one region and the traditional program used in another region, and the outcomes in both regions compared. A benefit in non-randomized studies is that they have fewer barriers for recruitment and informed consent in populations where scientific literacy is limited. They are often proposed as an ethical alternative to a randomized controlled study in one of two humanitarian-related scenarios:

1. Research evaluating mental health interventions or psychosocial programs. Here, it may be impossible practically or ethically to randomize a population to an intervention that is considered novel or untested. For example, a new nutritional program could be introduced into some refugee camps and the health outcomes compared with other camps which maintained the traditional nutritional programs. Randomizing the individuals or camps to the programs may not be feasible. Another example could be where non-randomized research was used to evaluate reintegration programs for former child soldiers as it would have been unethical to put them on a waitlist.

2. Biomedical research during an acute health crisis with a high fatality rate as a way to rapidly determine a safe and effective treatment. Because all participants receive the study treatment, it quickly allows the affected community equal access to a potentially life-saving treatment if no other standard treatment is available.

Despite the ethical justifications that non-randomized studies can reduce risk to affected populations, there is significant debate regarding the ethics of conducting non-randomized clinical trials, particularly during an infectious disease outbreak. While NR trials do create more flexible research designs, require fewer participants, and provide equal access to treatment, others argue that randomization is required to reduce the risk of bias and essential to establishing the cause of any changes. Without randomization, outcome differences may arise from selection bias or other differences between the participants, leading to ongoing uncertainty about the intervention’s efficacy. Conducting non-randomized clinical trials should take care to preserve scientific integrity and be resource-conscious, considering the long-term effects of the scientific knowledge that is produced. The 2014-15 Ebola outbreak in Western Africa is a useful example for understanding the ethical pros and cons of utilizing non-randomized clinical trials during acute healthcare emergencies.