Critical care research on patients with advance directives or do-not-resuscitate status: Ethical challenges for clinician-investigators

Abstract

Clinician-investigators face challenges in conducting research on critically ill patients when they have do-not-resuscitate orders, advance directives, or are in need of end-of-life care. Potential conflicts of interest for clinician-investigators include either financial stakes or academic and reputational stakes. The dual roles for intensive care unit physician or nurse clinician-investigators as healthcare professionals and scientists also present conflicts of interest, as does the dual purpose for the physical plant of the intensive care unit, which simultaneously serves as a site for patient care and a site for clinical research. Intensive care unit patients who become human research subjects also have dual roles that present conflict or confusion that can result in the therapeutic misconception. There are no scientifically or ethically sound reasons to exclude patients from participation in critical care research based on the presence of an advance directive or do-not-resuscitate order, as it would create a biased study sample that does not reflect the critically ill patient population, and it would treat a large group of potential research subjects differently from others without justification. There are four values in tension for critical care clinician-investigators in relation to patients/human research subjects: curative intent, palliative intent, research, and fiduciary obligations. A patient's decision to participate in research does not relieve clinician-investigators of their obligation to serve patient/human research subject's interests, even when doing so involves decisions to limit or withdraw life-sustaining interventions or withdraw the patient/human research subject from research. Critical care research involving patients with advance directives or do-not-resuscitate status is both possible and desirable because it is just, respects patient autonomy, and results in study populations that better reflect the clinical population in all respects.