A Phased Cluster-randomized Trial of Rural Hospitals Testing a Quality Collaborative to Improve Heart Failure Care

Abstract

Use of evidence-based practices for heart failure (HF) patients has the potential to improve outcomes and reduce variations in care delivery. To evaluate the effect of a rural hospital quality collaborative and organizational context (nurse staffing and practice environment) on 4 HF core measures. Phased cluster-randomized trial with delayed intervention control group. The intervention included a HF toolkit, 2 onsite meetings, and a monthly phone call. Twenty-three rural eastern US hospitals, registered nurses who care for HF patients (N=591). Seven quarters of 4 HF core measures, nurse staffing (nursing skill mix, registered nurse hours per patient day, nurse-turnover), and a survey of practice environment. : Using regression models with generalized estimating equation autoregressive methods, no statistically significant changes were found during the intervention period on all 4 core measures for either group. Higher nurse-turnover was related to all 4 core measures: lower compliance with discharge instructions [β=-1.042; 95% confidence interval (CI): -1.777, -0.307], smoking cessation (β=-1.148; 95% CI: -2.180, -0.117), left ventricular ejection fraction (β=-0.893; 95% CI: -1.784, -0.002), and prescribing angiotensin converting enzyme inhibitors on discharge (β=-1.044; 95% CI: -1.820, -0.269). Better practice environment was related to higher left ventricular ejection fraction (β=0.217; 95% CI: 0.054, 0.379). Significant improvements in 4 core measures were realized in stable environments (less nurse-turnover). Assuring appropriate nurse staffing and stability is essential to increase organizational preparation for quality initiatives and adoption of best practices in HF care in rural hospitals.