Post-hospitalization dialysis facility processes of care and hospital readmissions among hemodialysis patients: a retrospective cohort study

BACKGROUND: Both dialysis facilities and hospitals are accountable for 30-day hospital readmissions among U.S. hemodialysis patients. We examined the association of post-hospitalization processes of care at hemodialysis facilities with pulmonary edema-related and other readmissions. METHODS: In a retrospective cohort comprised of electronic medical record (EMR) data linked with national registry data, we identified unique patient index admissions (n = 1056; 2/1/10–7/31/15) that were followed by ≥3 in-center hemodialysis sessions within 10 days, among patients treated at 19 Southeastern dialysis facilities. Indicators of processes of care were defined as present vs. absent in the dialysis facility EMR. Readmissions were defined as admissions within 30 days of the index discharge; pulmonary edema-related vs. other readmissions defined by discharge codes for pulmonary edema, fluid overload, and/or congestive heart failure. Multinomial logistic regression to estimate odds ratios (ORs) for pulmonary edema-related and other vs. no readmissions. RESULTS: Overall, 17.7% of patients were readmitted, and 8.0% had pulmonary edema-related readmissions (44.9% of all readmissions). Documentation of the index admission (OR = 2.03, 95% CI 1.07–3.85), congestive heart failure (OR = 1.87, 95% CI 1.07–3.27), and home medications stopped (OR = 1.81, 95% CI 1.08–3.05) or changed (OR = 1.69, 95% CI 1.06–2.70) in the EMR post-hospitalization were all associated with higher risk of pulmonary edema-related vs. no readmission; lower post-dialysis weight (by ≥0.5 kg) after vs. before hospitalization was associated with 40% lower risk (OR = 0.60, 95% CI 0.37–0.96). CONCLUSIONS: Our results suggest that some interventions performed at the dialysis facility in the post-hospitalization period may be associated with reduced readmission risk, while others may provide a potential existing means of identifying patients at higher risk for readmissions, to whom such interventions could be efficiently targeted. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s12882-018-0983-5) contains supplementary material, which is available to authorized users.