Abstract 14940: Determinants of Rehospitalization After Cardiac Surgery in Prolonged ICU Length of Stay Survivors

Introduction: Determining predictors of rehospitalization of prolonged ICU length of stay (prICULOS) post cardiac surgery (CS) survivors would help inform processes of care in regards to optimization of patients pre-discharge and having additional community supports for this vulnerable population. Methods: Data of CS patients with prICULOS (ICULOS of ≥5 days post CS) from Jan. 1, 2000 to Dec. 31, 2011 was extracted from linked administrative and clinical databases to describe rehospitalization parameters. Regression analysis was used to determine predictors of time to first rehospitalization. Results: Of 9209 CS patients discharged home and alive at 30 days (d), 596 (6.5%) experienced prICULOS. The 30d rehospitalization rate was 109/596 (18.3%) for prICULOS patients vs. 996/8613 (11.6%) for non-prICULOS patients (p<0.01). For the prICULOS cohort, the cumulative rehospitalization rate at 1 year was 44.3%, and the median hospital LOS for readmissions was 7d. The most common reason for rehospitalization was heart failure. Out of 54 prICULOS patients that died within 1 year of discharge post CS, 32 (59%) died in hospital during rehospitalization. Predictors of time to first rehospitalization are in the table. Median time to first hospitalization of patients with the risk factors in the table was 54d compared to a median time of 70d in patients without these factors. Any rehospitalization within 1 year was independently associated with increased risk of death [OR 2.91, p<0.01]. Conclusions: Survivors of prICULOS after CS have a high rate of rehospitalisation and nearly a 3 fold increase risk of death if needing rehospitalisation within 1 year of initial discharge. We have identified 6 independent risk factors associated with rehospitalisation that can be used to optimize patients pre-discharge as well as be used to identify patients that may need closer follow-up in the community to prevent rehospitalization.