INTRODUCTIONDespite well-established guidelines, multiple recent studies have demonstrated variability in the conduct of brain death certification. This is undesirable given the gravity of the diagnosis. We sought therefore to survey local clinicians involved in brain death certification to identify specific areas of variability, if any, and to elicit information on how the testing process can be improved.

MATERIALS AND METHODSAn anonymous questionnaire was sent to all clinicians on the brain death certification roster in a tertiary neurosciences referral centre. This survey covered clinician demographics, evaluation of current and proposed resources to assist clinicians in certification, knowledge of the legislation governing brain death and organ procurement, technical performance of the brain death tests, and their views on the appropriate limits of physiological and biochemical preconditions for brain death testing.

RESULTSWe found significant variability in the conduct of brain death testing, especially in performing the caloric and apnoea tests. Of the existing resources to assist clinicians, written aide-memoires were the most popular. Respondents felt that bedside availability of a more detailed written description of the brainstem tests, and a formal accreditation course would be useful. There was wide variation in the limits of serum sodium and glucose, and the minimum core temperature and systolic blood pressures that respondents felt would preclude testing but we were able to identify thresholds at which the majority would be happy to proceed. We addressed the issues identified in our study by improving our written hospital brain death protocol, and designing an instructional course for clinicians involved in brain death certification.

CONCLUSIONSOur findings confirm that variability in the performance of brain death testing is indeed a universal phenomenon. Formal training appears desirable, but more importantly, clear and detailed protocols for testing should be made available at the bedside to assist clinicians. These protocols should be tailored to provide step-by-step instructions so as to avoid the inconsistencies in testing identified by this and other similar studies.