Using the age-adjusted Charlson comorbidity index to predict outcomes in emergency general surgery

E. St-Louis,S. Iqbal,L. Feldman,M. Sudarshan,D. Deckelbaum,T. Razek,K. Khwaja
DOI: https://doi.org/10.1097/TA.0000000000000457
2015-02-01
Journal of Trauma and Acute Care Surgery
Abstract:BACKGROUND We evaluated the role of the Charlson age-comorbidity index (CACI), a weighted comorbidity index that reflects cumulative increased likelihood of 1-year mortality, in predicting perioperative outcomes in an emergency general surgery population at a large Canadian teaching hospital. METHODS A retrospective chart review of emergency general surgery admissions in 2010 was conducted. Patients who had surgery were identified. Mode of surgery and CACI were recorded, as well as measures of outcome, including 30-day mortality and intensive care unit (ICU) admission. A multivariate stepwise logistic regression model was created to assess the effect of age-adjusted Charlson comorbidity index on postoperative outcomes while controlling for the effect of possible confounders. The prediction ability of CACI for mortality was assessed using receiver operating characteristic analyses considering the area under the curve and its 95% confidence intervals (CIs). RESULTS Of the 529 admissions to general surgery from the emergency department, 257 patients underwent a surgical intervention. The CACI scores ranged from 0 to 16. We described a total of 11 deaths (4.3%) and 30 ICU admissions (11.7%). CACI was associated with an increased risk of 30-day mortality (adjusted odds ratio,1.39; 95% CI, 1.11–1.73; p = 0.0034). Receiver operating characteristic analysis was consistent with high accuracy of CACI for mortality prediction alone, resulting in area under the curve or c statistic of 0.90 (95% CI, 0.84–0.95). CACI was similar in predicting mortality to a multivariate model. CACI was also found to be associated with ICU admission (adjusted odds ratio, 1.17; 95% CI, 1.01–1.37; p < 0.0382). CACI is not as good a predictor for ICU admission when compared with the multivariate model. CONCLUSION We have shown that the CACI is a valid tool for 30-day mortality prediction in the context of emergency general surgery. LEVEL OF EVIDENCE Prognostic study, level III.
What problem does this paper attempt to address?