Comparison of Charlson's weighted index of comorbidities with the chronic health score for the prediction of mortality in septic patients.

BACKGROUND: Comorbidity is one of the most important determinants of short-term and long-term outcomes in septic patients. Charlson's weighted index of comorbidities (WIC) and the chronic health score (CHS), which is a component of the acute physiology and chronic health evaluation (APACHE) II, are two frequently-used measures of comorbidity. In this study, we assess the performance of WIC and CHS in predicting the hospital mortality of intensive care unit (ICU) patients with sepsis.

METHODS: A total of 338 adult patients with sepsis were admitted to a multisystem ICU between October 2010 and August 2012. Clinical data were collected, including age, gender, underlying diseases, key predisposing causes, severity-of-sepsis, and hospital mortality. The APACHE II, CHS, acute physiology score (APS), sequential organ failure assessment (SOFA) and WIC scores were assessed within the first 24 hours of admission. Univariate and multiple Logistic regression analyses were used to compare the performance of WIC and CHS. The area under the receiver operating characteristic curve (AUC) was used to predict hospital mortality over classes of risk.

RESULTS: Of all the enrolled patients, 224 patients survived and 114 patients died. The surviving patients had significantly lower WIC, CHS, APACHE II, and SOFA scores than the non-surviving patients (P < 0.05). Combining WIC or CHS with other administrative data showed that the hospital mortality was significantly associated with age, severe sepsis, key predisposing causes such as pneumonia, a history of underlying diseases such as hypertension and congestive cardiac failure, and WIC, CHS and APS scores (P < 0.05). The AUC for the hospital mortality were 0.564 (95% confidence interval (CI) 0.496-0.631) of CHS, 0.663 (95% CI 0.599-0.727) of WIC, 0.770 (95% CI 0.718-0.822) of APACHE II, 0.856 (95% CI 0.815-0.897) of the CHS combined with other administrative data, and 0.857 (95% CI 0.817-0.897) of the WIC combined with other administrative data. The diagnostic value of WIC was better than that of CHS (P = 0.0015).

CONCLUSIONS: The WIC and CHS scores might be independent determinants for hospital mortality among ICU patients with sepsis. WIC might be an even better predictor of the mortality of septic patients with comorbidities than CHS.