RESUMO
BACKGROUND: The Surgical Apgar Score (SAS) describes a feasible and objective tool for predicting surgical outcomes. However, the accuracy of the score and its correlation with the complication severity has not been well established in many grounds of low resource settings. OBJECTIVE: To determine the accuracy of Surgical Apgar Score in predicting the severity of post-operative complications among patients undergoing emergency laparotomy at Muhimbili National Hospital. METHODS: A prospective cohort study was conducted for a period of 12 months; patients were followed for 30 days, the risk of complication was classified using the Surgical Apgar Score (SAS), severity of complication was estimated using the Clavien Dindo Classification (CDC) grading scheme and Comprehensive Complication Index (CCI). Spearman correlation and simple linear regression statistic models were applied to establish the relationship between Surgical Apgar Score (SAS) and Comprehensive Complication Index (CCI). The Accuracy of SAS was evaluated by determining its discriminatory capacity on Receiver Operating Characteristics (ROC) curve, data normality was tested by Shapiro-Wilk statistic 0.929 (p < 0.001).Analysis was done using International Business Machine Statistical Product and Service Solution (IBM SPSS) version 27. RESULTS: Out of the 111 patients who underwent emergency laparotomy, 71 (64%) were Male and the median age (IQR) was 49 (36, 59).The mean SAS was 4.86 (± 1.29) and the median CCI (IQR) was 36.20 (26.2, 42.40). Patients in the high-risk SAS group (0-4) were more likely to experience severe and life-threatening complications, with a mean CCI of 53.3 (95% CI: 47.2-63.4), compared to the low-risk SAS group (7-10) with a mean CCI of 21.0 (95% CI: 5.3-36.2). A negative correlation was observed between SAS and CCI, with a Spearman r of -0.575 (p < 0.001) and a regression coefficient b of -11.5 (p < 0.001). The SAS demonstrated good accuracy in predicting post-operative complications, with an area under the curve of 0.712 (95% CI: 0.523-0.902, p < 0.001) on the ROC. CONCLUSION: This study has demonstrated that SAS can accurately predict the occurrence of complications following emergency laparotomy at Muhimbili National Hospital.
