Predicting in-hospital mortality for dementia patients after hip fracture surgery - A comparison between the Charlson Comorbidity Index (CCI) and the Elixhauser Comorbidity Index.

ABSTRACT

BACKGROUND: Considerable in-hospital mortality was reported in geriatric patients with dementia sustaining femoral neck or inter-trochanteric fracture. We intended to establish a predictive model of in-hospital mortality for dementia patients after hip fracture surgery. METHODS: We collected 8080 registrants ≧ 65 years old from the subset (LHID2000) of the National Health Insurance Research Database (NHIRD) that met the following inclusion criteria1. Admitted with the ICD of hip fracture; 2. Underwent operation of hip fracture during the same hospitalization; 3. Co-existing diagnosis of dementia (ICD-9-CM codes 290). The co-morbidity was recorded according to validated Charlson Comorbidity Index (CCI) and Elixhauser Comorbidity Index (ECI) from the diagnoses of hospitalization. The main outcome measure was in-hospital mortality that was defined as death being reported during hospitalization. The comparison of predictability was conducted by Receiver operating characteristic (ROC) curves and the area under the ROC curve (AUC) among different predictive models. RESULTS: The Charlson Comorbidity Index (CCI) score and Elixhauser Comorbidity Index (ECI) score showed similar ability in predicting in-hospital mortality (AUC = 0.653, 95% CI = 0.611-0.695 for CCI; AUC = 0.624, 95% CI = 0.582-0.665 for ECI, p = 0.0717). By adding age grouping (≥80 yrs = 1, 65-80 yrs = 0) and gender difference (Male = 1, Female = 0), these two models were shifted to models CCI_new1 and ECI_new1. Consequently, the AUC greatly increased in the CCI_new1 (AUC = 0.682, 95% CI = 0.643-0.722). It therefore provided better prediction of in-hospital mortality than ECI_new1 (AUC = 0.651, 95% CI = 0.611-0.691) (p = 0.0444). CONCLUSIONS: Utilizing the CCI with addition of grouping for age and gender provides a better prediction for in-hospital mortality than the ECI among elderly patients with concomitant dementia and hip fracture who underwent surgical intervention.