Effect of nurse-implemented transitional care for chinese individuals with chronic heart failure in Hong Kong : A randomized controlled trial

Objectives: To determine the effect of nurse-implemented transitional care (TC) on readmission and mortality rates in Chinese individuals with chronic heart failure (CHF) in Hong Kong. Design: Single-center randomized controlled trial of TC versus usual care (UC). Setting: University-affiliated hospital in Hong Kong. Participants: Hospitalized Chinese individuals with CHF (N = 178; aged 78.6 ± 6.9, 45% male). Measurements: The TC group received a predischarge visit, two home visits, and then regular telephone calls over 9 months to provide self-care education and support, optimized health surveillance, and facilitation in use of community services. Primary endpoints were event-free survival, all-cause hospital readmission, and mortality during the 9-month follow-up. Secondary endpoints were length of hospital stay, self-care, and health-related quality of life (HRQL). Data were analyzed using survival analysis and generalized estimating equations, following an intention-to-treat principle. Results: Survival analysis indicated no significant differences in event-free survival, hospital readmission, or mortality between the TC and UC groups, although the TC group had a lower hospital readmission rate at 6 weeks (8.1% vs 16.3%, P = .048) and lower mortality at 9 months (4.1% vs 13.8%, P = .03). The TC group also had a shorter hospital stay (P = .006) and significantly better self-care and HRQL. Because of attrition, sensitivity analyses were conducted to examine whether the intention-to-treat assumption affected the results. Per-protocol population analyses (hazard ratio (HR) = 0.40, 95% confidence interval (CI) = 0.17–0.93) and worst-case-scenario analysis (HR = 0.44, 95% CI = 0.25–0.77) suggested a lower mortality risk in the TC group. Conclusion: The translation of individual-centered nurse-implemented TC to the Chinese culture and healthcare context of Hong Kong appears beneficial.