Respect for patient autonomy via a process of shared decision-making is commonly recommended in guidelines for nephrologists and others working in kidney health care. This is consistent with broader norms in health care provision in many countries that judge paternalism to be ethically unacceptable. Paternalistic practices that were once prevalent have been undermined by research demonstrating that many, if not most, patients expect and desire respect for their autonomy. In particular, withholding information from patients or their families so that the clinician may make a decision on the patient’s behalf is no longer considered ethically appropriate, even if the clinician feels this is necessary to prevent harm. There is evidence that most people prefer to receive information about their health, including details about their prognosis and potential treatment options, even if information may be distressing and treatment may be inaccessible. However, in discussions with clinicians working in resource constrained environments where patients with kidney failure may be unable to access or afford life sustaining treatments such as dialysis or transplantation, several report routinely withholding information from patients and failing to include them in decision-making about treatment. Some justify this as necessary to avoid causing harm, while others suggest that respect for patient autonomy may be irrelevant or unachievable in the setting of severe resource constraints.

In this talk I explore the concept of patient autonomy in the context of kidney care in resource-constrained environments. I examine potential concerns about the relevance and feasibility of respecting autonomy and discuss the potential risks and benefits of withholding information from people with kidney failure in such contexts. I suggest that factors influencing paternalistic practices in resource-constrained environments are likely to reflect similar practices in well-resourced settings, and argue that efforts to promote patient autonomy are especially relevant in resource constrained environments.