Examining Factors Associated With Early, Post-Discharge, Acute Healthcare Utilization In Older Adults

Statement of Problem: Inpatient (IP) readmissions within 30-days of discharge from an index hospitalization are prevalent, costly, and propagate risks associated with transitions from various healthcare settings for older adults. The Centers for Medicare and Medicaid Services (CMS) implemented the Hospital Readmissions Reduction Program (HRRP) in 2012 to track excess all-cause readmissions for three target conditions. Identifying readmissions that are potentially preventable may advance efforts to mitigate risk. Post-acute care (PAC) services use has increased in an effort to lower readmission rates but Medicare costs are escalating. PAC referrals are not based on standardized processes; clinical decision support (CDS) may aid clinicians to make appropriate referrals. Methods: Review of the literature of factors associated with IP readmission for target conditions by various readmission types; two secondary data analyses separately examine the impact of hospital discharge disposition, and a PAC referral CDS algorithm on acute health utilization (IP readmission plus emergency department (ED)/observational unit (OBS) returns) within 30 days of hospital discharge. Results: Most literature examining readmission risk for target conditions, analyzed administrative data collected before 2010 for all-cause readmissions therefore it was not possible to examine by HRRP implementation or across various readmission definitions. Significant patient- and organizational-level factors were associated with readmission risk, but PAC services use for target conditions was not widely studied. After secondary data analyses, patients discharged to a PAC facility returned for IP readmission sooner especially within 24 hours of hospital discharge. These patients were younger, had greater functional decline, and heightened fall risk. Return ED/OBS use was highest in home healthcare (HHC) patients particularly five-to-seven days post-hospitalization. Those patients who went home but were flagged by the CDS algorithm as in need of PAC services had a 67.8 percent higher risk for IP readmission and a 73.8 higher risk of return ED/OBS use. Conclusions: Some readmission risk factors e.g., functional decline, cardiac arrest history are not found in administrative data and may prove useful in predicting readmission risk. Improvements in transitions from hospital to PAC settings would be facilitated by CDS tools aimed at assessing discharge readiness and appropriate PAC referral match