Clinicians' Approaches to Prognostication After Traumatic Brain Injury and Cardiac Arrest: A Multi-Hospital, Qualitative Study.

BACKGROUND

Understanding how clinicians prognosticate is important for creating interventions that improve current practice and clinical trials. We compared clinicians' reported approaches to prognostication with traumatic brain injury (TBI) and comatose cardiac arrest (CA) for patients enrolled in a multicenter clinical trial.

METHODS

We conducted semi-structured interviews with clinicians who treated patients with severe traumatic brain injury (TBI) enrolled in the Brain Oxygen Optimization in Severe TBI Phase-3 (BOOST-3) trial (NCT03754114). We compared these reports with ones in a previous study of patients who were comatose after cardiac arrest (CA) and were enrolled in the Influence of Cooling Duration on Efficacy in Cardiac Arrest Patients (ICECAP) trial (NCT04217551). We performed deductive coding using our codebook from CA interviews and then used inductive coding to add new topics raised in the TBI interviews. We looked specifically for reported reliance on initial "clinical gestalt" as observed in the CA interviews.

RESULTS

We interviewed 18 clinicians at 13 hospitals. Predicting poor outcomes was less common with TBI cases than in the CA study, consistent with records showing that final prognostication was determined later in the TBI cases (7 [interquartile range (IQR) 2-18.5] vs. 3 [IQR 2-7] days). Similar percentages of clinicians reported high confidence in their initial prognostic assessments in the two settings (TBI, 33%; CA, 40%). Fewer clinicians reported relying on initial clinical gestalt predictions with patients with TBI (22%) than with patients who had experienced CA (70%). With patients with TBI, more clinicians reported having used later subjective assessments to revise their initial uncertain prognostication.

CONCLUSIONS

In interviews with clinicians practicing at multiple institutes, we found that clinicians were less likely to report relying on initial gestalt impressions with patients with TBI than with patients who had experienced CA and were more likely to report relying on later subjective assessments to refine uncertain initial prognostic judgments. Fewer clinicians reported high confidence in initial assessments of patients with TBI.