Interrater Agreement and Reliability of Burn Size Estimations Between Emergency Physicians and Burn Unit.

Abstract

The initial approach to burn injuries has remained relatively unchanged over the past several decades and revolves around trauma assessment and fluid resuscitation, frequently initiated in the emergency department (ED). While previous research suggests that emergency physicians (EP) are poor estimators at total body surface area (TBSA) affected, we believe that estimation differences are improving, specifically at academic centers with co-located burn units that emphasize burn injury education. This study investigated the interrater agreement and reliability of burn size estimations at an academic ED and its co-located burn unit. This single-center, retrospective study was conducted at a large academic ED with a co-located burn unit. The study included adult patients admitted to the burn unit after receiving paired burn size estimations from EPs and the burn unit. The primary endpoint was the interrater agreement, measured by the kappa coefficient, κ, of 10% TBSA estimation intervals. The secondary endpoint was the intraclass correlation coefficient (ICC), evaluating the reliability of exact, nonranged, and TBSA estimations. A chart review was performed for patients evaluated from November 1, 2016 to July 31, 2019. One thousand one hundred and eighty-four patients were admitted to the burn unit, 1176 of which met inclusion criteria for the primary endpoint. The κ of TBSA between EPs and the burn unit was 0.586, while the weighted κ was 0.775. These values correlate with moderate and substantial agreements, respectively. Additionally, 971 patients had exact TBSA estimations from paired EPs and the burn unit which were used for the secondary endpoint. The ICC between EPs and the burn unit was 0.966, demonstrating an excellent reliability. Further sub-analysis was performed, revealing mean over- and underestimation differences of exact TBSA estimations of 3.93 and 2.93, respectively. EPs at academic institutions with co-located burn units are accurate estimators of TBSA in the assessment of burn injuries. We believe that burn education, to include core rotations within the burn unit, plays a major role in improved burn size estimations.