Predictive value of admission SOFA scores for ICU mortality: A single-centre study from a South African tertiary hospital

Abstract

Background. The Sequential Organ Failure Assessment (SOFA) score was initially developed as a tool to quantify the degree of organ dysfunction in patients with sepsis. It is now a commonly used severity-of-illness score across a variety of settings. Although the SOFA score is versatile and simple, it has not been well evaluated in South African (SA) critical care populations. Objective. To determine whether an association exists between admission SOFA score and patient outcomes in a heterogeneous population of critically ill patients in a tertiary intensive care unit (ICU) in SA. Methods. This study was a single-centre, retrospective, observational study conducted at a level 3 ICU in Durban, SA. Data required to calculate SOFA scores were obtained from patient charts. The primary outcome of interest was ICU mortality. Results. Complete data for SOFA score calculation were collected from 279 patient charts. Of the total cohort, 90 patients (32.3%) died, and 189 (67.7%) survived. The median SOFA scores of survivors and non-survivors were 7 and 10, respectively (p<0.001). Receiver operating characteristic curve analysis demonstrated an area under the curve of 0.739 (95% confidence interval (CI) 0.677 - 0.800; p<0.001) for SOFA score prediction and ICU mortality. Conclusion. The SOFA score is a simple severity-of-illness tool derived from clinical variables routinely available on ICU admission. In this study, admission SOFA score was independently associated with ICU mortality in a multidisciplinary SA ICU and demonstrated fair discriminative ability for predicting ICU mortality.