Utility of an emergency department clinical protocol for early identification of coronavirus infection

Abstract

Introduction: We assessed the utility of an emergency department (ED) protocol using clinical parameters to rapidly distinguish likelihood of novel coronavirus 2019 (COVID-19) infection; the applicability aimed to stratify infectious-risk pre-polymerase chain reaction (PCR) test results and accurately guide early patient cohorting decisions. Methods: We performed this prospective study over a two-month period during the initial surge of the 2020 COVID-19 pandemic in a busy urban ED of patients presenting with respiratory symptoms who were admitted for in-patient care. Per protocol, each patient received assessment consisting of five clinical parameters: Presence of fever; hypoxia; cough; shortness of breath/dyspnea; and performance of a chest radiograph to assess for bilateral pulmonary infiltrates. All patients received nasopharyngeal COVID-19 PCR testing. Results: Of 283 patients studied, 221 (78%) were PCR+ and 62 (22%) PCR-. Chest radiograph revealed bilateral pulmonary infiltrates in 85%, which was significantly more common in PCR+ (94%) vs PCR- (52%) patients (P < 0.0001). The rate of manifesting all five positive clinical parameters was significantly greater in PCR+ (63%) vs PCR- (6.5%) patients (P < 0.0001). For PCR+ outcome, the presence of all five positive clinical parameters had a specificity of 94%, positive predictive value of 98%, and positive likelihood ratio of 10. Conclusions: Using an ED protocol to rapidly assess five clinical parameters accurately distinguishes likelihood of COVID-19 infection prior to PCR test results, and can be used to augment early patient cohorting decisions.