Online training improves medical students' ability to recognise when a person is dying: the ORaClES randomised controlled trial

Background: Recognising dying is a key clinical skill for doctors, yet there is little training. Aim: To assess the effectiveness of an online training resource designed to enhance medical students’ ability to recognise dying. Design: Online multicentre double-blind randomised controlled trial (NCT03360812). The training resource for the intervention group was developed from a group of expert palliative care doctors’ weightings of various signs/symptoms to recognise dying. The control group received no training. Setting/participants: Participants were senior UK medical students. They reviewed 92 patient summaries and provided a probability of death within 72 hours (0% certain survival – 100% certain death) pre, post, and 2 weeks after the training. Primary outcome: (1) Mean Absolute Difference (MAD) score between participants’ and the experts’ scores, immediately post intervention. Secondary outcomes: (2) weight attributed to each factor, (3) learning effect and (4) level of expertise (Cochran–Weiss–Shanteau (CWS)). Results: Out of 168 participants, 135 completed the trial (80%); 66 received the intervention (49%). After using the training resource, the intervention group had better agreement with the experts in their survival estimates (δ_MAD = −3.43, 95% CI −0.11 to −0.34, p = <0.001) and weighting of clinical factors. There was no learning effect of the MAD scores at the 2-week time point (δ_MAD = 1.50, 95% CI −0.87 to 3.86, p = 0.21). At the 2-week time point, the intervention group was statistically more expert in their decision-making versus controls (intervention CWS = 146.04 (SD 140.21), control CWS = 110.75 (SD 104.05); p = 0.01). Conclusion: The online training resource proved effective in altering the decision-making of medical students to agree more with expert decision-making.