Association of Hospital or Ambulatory-Based HF Transitional Care Services and 30 Day Readmissions

Introduction Hospital (hosp)/ambulatory (amb) transitional care services (TCS) aim to reduce 30-day readmissions by promoting optimized care and improving provider-patient communication. Study aims were to explore hosp and amb TCS availability and association of specific TC services with 30-day all-cause unadjusted readmissions. Methods Using a cross-sectional descriptive design and survey methods, members of ACC, HFSA and AAHFN who treat adults for decompensated HF in hosp or hosp-based amb centers completed an on-line questionnaire about TCS and provided 30-day readmission rates for 3 months. Multivariable logistic regression models assessed relationships between hosp and amb bundles and readmissions below the median rate after adjusting for presence of HF educator and hospitalists, and hosp type (hosp services only). Results Of 76 hosp and 60 amb responders who provided TCS and readmission data, many sites were teaching centers (80.2%) and part of multicenter systems (84.0%). Of sites, 90.8% used hospitalists, 42% used hosp-employed MDs and 52.6% had less than 500 beds. Hosp-owned services inc. home care (50%), palliative/hospice care (63.2%) and SNFs (32.9%). Median (25 th , 75 th IQR) 30-day readmission rate was 19.33% (15.30, 22.70). Of 27 hosp- and 30 amb-based TCS assessed, 9 hosp- and 2 amb-based TCS were in place by over 90% of sites; however, these prominent TCS were not associated with lower 30-day readmissions using lowest quartile or below median rates. Of all 57 hosp and amb TCS assessed, only 9 were associated with or trended toward lower readmissions ( Table ) and 2 amb TCS were associated with higher readmissions based on lowest quartile analyses: home care for all and HF specialty clinic. TCS were grouped by themes and assessed by high vs. low use. In multivariable models of hosp TCS bundled into 6 themes, readmission was more likely below the median when TCS included 2-3 quality-monitoring initiatives (vs. 0-1 services), p = 0.028. When amb TCS were bundled into 7 themes, readmission was more likely below the median when TCS included 6-7 communication-education services (vs. 0-5 services), p = 0.039. Conclusion Despite efforts by hospitals to improve 30-day HF readmissions, rates remain high. Hosp-based quality monitoring and amb-based patient communication-education HF TCS themes were associated with lower 30-day readmission and should be enhanced.