Off the Beaten Path: Conducting Ethical Pragmatic Trials with Marginalized Populations‪.‬

In October 2009, the Mental Health Commission of Canada (MHCC) began recruiting participants for its At Home/Chez Soi (AHCS) project. The AHCS project is designed as a pragmatic trial intended "to provide policy relevant evidence [i.e., effectiveness data] about what services and system interventions best achieve housing stability and improved health and well-being for those who are homeless and mentally ill." (1) Explanatory trials (e.g., traditional randomized control trials of a drug) test the efficacy or benefits of a treatment or intervention under ideal conditions with narrowly defined participants (i.e., Can the intervention work?), while pragmatic trials test the effectiveness or benefits of a treatment or intervention under routine clinical or health care delivery conditions with a broader pool of subjects (i.e., Does the intervention work?). (2) Given the importance of conducting ethically defensible research and abiding by the standards of mental health research ethics, the authors of this paper wrote an internal report about some research ethics challenges that might occur while conducting the AHCS project, paying particular attention to the sensitivity of working with persons who are homeless and mentally ill. In writing the report, it became readily apparent to us that traditional conceptions and procedures of research ethics in the clinical trials and community-based participatory research contexts were insufficient when conducting pragmatic trials with marginalized populations. Thus, modifying traditional ethical conceptions and procedures was necessary in order to better suit the needs of the AHCS project's participants, researchers, and service providers. In this paper, we begin by outlining the AHCS project and the guiding values and ethical goals of the project. Next, we describe how the existing literature and traditional modes of thinking about research ethics did not meet the project's needs. Finally, we use two examples--one substantive and the other procedural--to show how we modified existing research ethics thinking and practices to better serve project stakeholders.