Lessons learned from predicting hospital admission criteria comparing two COVID-19 waves: An Emergency Department's perspective.

Introduction Emergency Departments (ED) play a central role in providing healthcare during crises such as pandemics. Evaluation of admission criteria under the circumstances of wave-dynamics is the bottleneck question for both clinical practice and planning on a structural level.Methods This study compares patient characteristics, clinical biomarkers and outcomes during wave 1 (W1) and wave 2 (W2) of the COVID-19 pandemic at two tertiary care EDs in Berlin. We validated parameters predicting hospital admission by a CART-analysis of W1 in W2 and compared the findings to a CART-model derived from W2 by a decision curve analysis.Results A total of 1,204 patients with suspected COVID-19 presented in W1 and 1,440 in W2; 106 in W1 and 722 in W2 tested positive. Patients in W2 were older (>65 years, W1: 22.2%; W2: 34.6%) and multimorbid. Mortality of positive tested (W1: 1.9%; W2:8.6%) rose in W2. The admission-predicting parameters O2 saturation <95%, dyspnea, and cardiovascular comorbidities of W1 were applicable with a sensitivity of 98.5% (CI-95%: 96.3% - 99.4%) and specificity of 68.4% (61.0%-74.9%) in W2.Conclusions Our results show transferability of admission-criteria despite changing pandemic conditions and serve as lessons learned to support future pandemic preparedness and capacity planning.