Simulation Testing for Selection of Critical Care Medicine Trainees. A Pilot Feasibility Study

Selection of physicians into anesthesiology, intensive care, and emergency medicine training has traditionally relied on evaluation of curriculum vitae, letters of recommendation, and interviews, despite these methods being poor predictors of subsequent workplace performance.

In this study, we evaluated the feasibility and face validity of incorporating assessment of nontechnical skills in simulation and personality traits into an existing junior doctor selection framework.

Candidates short-listed for a critical care residency position were invited to participate in the study. On the interview day, consenting candidates participated in a simulation scenario and debriefing and completed a personality test (16 Personality Factor Questionnaire) and a survey. Timing of participants' progression through the stations and faculty staff numbers were evaluated. Nontechnical skills were evaluated and candidates ranked using the Ottawa Crisis Resource Management Global Rating Scale (Ottawa GRS). Nontechnical skills ranking and traditional selection method ranking were compared using the concordance correlation coefficient. Interrater reliability was assessed using the concordance correlation coefficient.

Thirteen of 20 eligible participants consented to study inclusion. All participants completed the necessary stations without significant time delays. Eighteen staff members were required to conduct interviews, simulation, debriefing, and personality testing. Participants rated the simulation station to be acceptable, fair, and relevant and as providing an opportunity to demonstrate abilities. Personality testing was rated less fair, less relevant, and less acceptable, and as giving less opportunity to demonstrate abilities. Participants reported that simulation was equally as stressful as the interview, whereas personality testing was rated less stressful. Assessors rated both personality testing and simulation as acceptable and able to provide additional information about candidates. The Ottawa GRS showed moderate interrater concordance. There was moderate concordance between rankings based on traditional selection methods and Ottawa GRS rankings (ρ = 0.52; 95% confidence interval, -0.02 to 0.82; P = 0.06).

A multistation selection process involving interviews, simulation, and personality testing is feasible and has face validity. A potential barrier to adoption is the high number of faculty required to conduct the process.