"Sick" or "not-sick": Accuracy of System 1 diagnostic reasoning for the prediction of disposition and acuity in patients presenting to an academic ED

Objective System 1 decision-making is fast, resource economic, and intuitive (eg, "your gut feeling") and System 2 is slow, resource intensive, and analytic (eg, "hypothetico-deductive"). We evaluated the performance of disposition and acuity prediction by emergency physicians (EPs) using a System 1 decision-making process. Methods We conducted a prospective observational study of attending EPs and emergency medicine residents. Physicians were provided patient demographics, chief complaint, and vital sign data and made two assessments on initial presentation: (1) likely disposition (discharge vs admission) and (2) "sick" vs "not-sick". A patient was adjudicated as sick if he/she had a disease process that was potentially life or limb threatening based on pre-defined operational, financial, or educationally derived criteria. Results We obtained 266 observations in 178 different patients. Physicians predicted patient disposition with the following performance: sensitivity 87.7% (95% CI 81.4-92.1), specificity 65.0% (95% CI 56.1-72.9), LR+ 2.51 (95% CI 1.95-3.22), LR - 0.19 (95% CI 0.12-0.30). For the sick vs not-sick assessment, providers had the following performance: sensitivity 66.2% (95% CI 55.1-75.8), specificity 88.4% (95% CI 83.0-92.2), LR + 5.69 (95% CI 3.72-8.69), LR - 0.38 (95% CI 0.28-0.53). Conclusion EPs are able to accurately predict the disposition of ED patients using system 1 diagnostic reasoning based on minimal available information. However, the prognostic accuracy of acuity prediction was limited.