Hospital Admissions, Mortality and Comorbidities Among New York State Sickle Cell Patients, 2005-2013

Analyses of administrative and large data sources in Sickle Cell Disease (SCD) can answer questions not suitable for prospective study but have been hampered by lack of validated methods to adjust for individual comorbidities and lack of baseline utilization data over time. We sought to develop a database to characterize inpatient SCD care across New York State and generate a re-weighted sickle-cell specific Charlson Comorbidity index (S-CCI) for use in future large data SCD research. We identified 18,541 individual SCD patients admitted to New York State hospitals between 2005 and 2013 from the SPARCS database. We present data from both a randomly selected derivation cohort, used to develop the S-CCI and a validation cohort, The S-CCI resulted in small improvements in model fit and discrimination while using fewer covariates, allowing a more parsimonious model. Despite being the most common comorbidity, chronic pulmonary disease was not predictive of mortality. Mortality per hospitalization was 0.61%. Many patients (32%) were admitted only once during the nine year period. However, the majority was admitted more frequently with over 15% of patients being admitted more than once per year.