ESICM LIVES 2016: part two

Abstract

Introduction: Despite initial recovery, many critically ill patients discharged from the intensive care unit (ICU) may experience deteri-oration1. Early identification of patients at risk might facilitate improvements in quality of care. Objective: To determine risk factors associated with short-term hospital outcome (unplanned ICU readmission and unexpected death on the ward). Methods: Prospective cohort study from August 2014 to May 2015 performed at Hospital Sirio Libanes (Sao Paulo, Brazil). During this period, we analyzed 527 patients who were admitted in our ICU and discharged alive. We used univariate and multivariate analysis to identify risk factors associated with latter ICU readmission or unexpected ward death in the same hospitalization period. Results: Forty seven patients (8.9 %) were readmitted after ICU discharge and further forty seven died in the ward. Patients who had unexpected outcomes were older compared to those with successful outcomes (age 72 +/- 17 vs 66 +/- 18 respectively, p = 0.003), were sicker (SAPS 3 of 53 [IQR 43-60] vs 38 [29-49] respectively, p < 0.001), male sex (62.8 % vs 47.1 % respectively, p = 0.006), had a non-surgical reason for hospitalization (75.5 % vs 45 % respectively, p < 0.001), more frequently came from the ward (21.3 % vs 7.6 % respectively, p < 0.001), were less frequently independent for daily activities (54.3 % vs 76.7 % respectively, p < 0.001), had sepsis (39.4 % vs 11.8 % respectively, p < 0.001), required mechanical ventilation (44.7 % vs 14.1 % respectively, p < 0.001), vasoactive drugs (54.3 % vs 24.2 % respectively, p < 0.001) and dialysis (16 % vs 1.2 % respectively, p < 0.001), had higher SOFA score both at ICU admission (4 [2-7] vs 1 [0-3] respectively, p < 0.001) and discharge (2 [1-4] vs 1 [0-2] respectively, p < 0.001), had higher C-reactive protein values at admission (5.8 mg/dL [1.9-15.6] vs 3.6 [0.8-8.6] respectively, p < 0.001) and lactate values (1.9 mmoll/L [1.2-3.7] vs 1.7 [1.1-2.5] respectively, p = 0.03), and higher Sabadell score at discharge (1 [1-2] vs 0 [0-1] respectively, p < 0.001). On multivariate analysis, only sepsis and Sabadell score were independent risk factors for worse outcomes (AUC 0.73; p < 0.001). Conclusions: Sabadell score at ICU discharge might be a promising subjective tool for hospital outcomes.