A risk scoring model based on vital signs and laboratory data predicting transfer to the intensive care unit of patients admitted to gastroenterology wards

Abstract Purpose To compare the ability of a score based on vital signs and laboratory data with that of the modified early warning score (MEWS) to predict ICU transfer of patients with gastrointestinal disorders. Materials and methods Consecutive events triggering medical emergency team activation in adult patients admitted to the gastroenterology wards of the Asan Medical Center were reviewed. Binary logistic regression was used to identify factors predicting transfer to the ICU. Gastrointestinal early warning score (EWS-GI) was calculated as the sum of simplified regression weights (SRW). Results Of the 1219 included patients, 468 (38%) were transferred to the ICU. Multivariate analysis identified heart rate ≥ 105 bpm (SRW 1), respiratory rate ≥ 26 bpm (SRW 2), ACDU (Alert, Confused, Drowsy, Unresponsive) score ≥ 1 (SRW 2), SpO2 /FiO2 ratio < 240 (SRW 2), creatinine ≥ 2.0 mg/dL (SRW 2), total bilirubin ≥ 9.0 mg/dL (SRW 2), prothrombin time/international normalized ratio (INR) ≥ 1.5 (SRW 2), and lactate ≥ 3.0 mmol/L (SRW 2) for inclusion in EWS-GI. The area under the receiver operating characteristic curve of the EWS-GI was larger than that of MEWS (0.76 vs. 0.64; P < 0.001). Conclusions EWS-GI may predict ICU transfer among patients admitted to gastroenterology wards. The EWS-GI should be prospectively validated.