Outpatient Parenteral Antibiotic Therapy (OPAT) Serial Readmissions and High Utilizers: Drilling Down on Readmission Following Discharge on Intravenous Antibiotics

BACKGROUND: OPAT discharges (d/c’s) are at high risk for hospital readmission. We observed that a subgroup of OPAT patients (patients) had multiple 30 day readmissions (30DRs) and wanted to define both global and global minus serial readmitter (SR) or high utilizer (HU) 30DR rates to improve the d/c process and resource utilization. METHODS: Our program tracks all d/c’s on IV antibiotics (excluding Cystic Fibrosis patients) and reviews unsuccessful d/c’s and 30DRs (to DHMC only). OPAT patients include those d/c to home or to facilities. We studied d/c’s from 2014 to 2016 and the 30DRs linked to those index d/c’s (through January 2017). 30DRs linked to non-OPAT d/c’s were not included. Around each index d/c, we quantified: LOS; ID diagnosis; d/c disposition; underlying disease; insurance status; # of 30DRs (emergent, urgent, and elective per current Medicare reporting; excluding deaths and against medical advice d/c’s); readmit cause. SRs were defined as having >2 30DRs following an index d/c; HUs were defined as having >4 OPAT d/c’s over a rolling 12-month period. We compared the entire d/c cohort with the subgroups of SRs and HUs. Statistical analysis was performed using SPC Excel. RESULTS: 1189 unique patients had 1454 OPAT d/c’s and 267 unique 30DRs. The ALOS for all d/c’s was 10.1 days (range 0–113); 65.4% were d/c to home (54% to home infusion). Medicare patients comprised 52.3%. The average monthly 30DR rate was 20.4% (range 3.6% to 40.6%), compared with DHMC all cause 30DR rate of 10.5% (FY2016). 8 SRs had >2 30DRs (range 3–6) around 14 index d/c’s (0.9 % of total). 10 HUs had >4 d/c in 12 consecutive months, accounting for 54 d/c’s (3.7% of total). If SRs were excluded, the average 30DR rate fell to 19.0%; if HUs were excluded, the rate fell to 18.4%. Rates were not combined since 4 patients were both SRs and HUs. Underlying diseases of HUs included: cirrhosis, DM, IBD, cancer, renal transplant. HUs ID diagnoses included recurrent bacteremias from occult TIPS infection, IVC’s for TPN, UTIs and osteomyelitis. CONCLUSION: OPAT patients have high 30DR rates that exceed overall and population specific 30DR rates in our hospital. By excluding SRs and HUs, 30DR rates may more closely typify an average OPAT experience and be a better program process measure. More understanding about drivers for OPAT d/c for SRs and HUs is needed. DISCLOSURES: All authors: No reported disclosures.