BRAIN NATRIURETIC PEPTIDE VERSUS THE SIMPLIFIED ACUTE PHYSIOLOGY SCORE 3 FOR PREDICTING MORTALITY IN A SOUTH AFRICAN GENERAL INTENSIVE CARE UNIT.

Abstract

Background: The biomarker brain natriuretic peptide (BNP) may be more useful for risk stratification in general intensive care unit (ICU) patients when compared with more complex scores, such as the Simplified Acute Physiology Score 3 (SAPS 3). This study sought to investigate the prognostic relevance of BNP levels and SAPS 3 for ICU mortality at a South African quaternary hospital. Methods: This was a retrospective chart review of adult patients admitted to the general ICU of the Inkosi Albert Luthuli Central Hospital in Durban, South Africa in 2018 and 2019. Patient characteristics, BNP levels, and SAPS 3 totals were collected for each patient. The study outcome was ICU mortality following admission to the ICU. The prognostic relevance of BNP levels and SAPS 3 was investigated using receiver-operator characteristic curve statistics and area under the curve (AUC). Results: The sample size was 376 patients (medical: 133; surgical: 243). ICU inpatient mortality was 29%. SAPS 3 demonstrated superior prognostic accuracy when compared with BNP (AUC = 0.815 vs. 0.657). However, when BNP levels were compared between patients stratified as high- and low-risk for mortality based on SAPS 3, the BNP levels were significantly higher in low-risk patients who died vs. low-risk patients who survived (median: 306.0 ng/L vs. 92.0 ng/L, p<0.001). Conclusion: Overall, SAPS 3 was superior to BNP at predicting ICU inpatient mortality. However, BNP might potentially have some prognostic utility in patients stratified as low risk according to SAPS 3. Recommendation: Future research, specifically larger multicentre studies, should be conducted in the African continent to validate the results of the present study.