Background
Multi-disciplinary heart failure (HF) clinics improve outcomes for HF patients in randomized clinical trials. It is unclear if this efficacy translates to real world effectiveness. Accordingly, our objectives were to 1) compare real world outcomes of HF patient treated in HF clinics vs that in standard care and 2) identify HF clinic features associated with improved outcomes.
Methods
The service components at all 34 HF clinics in Ontario, Canada were evaluated and scored using a validated instrument. Based these scores, the clinics were categorized by an expert panel into high/medium or low intensity strata. Our cohort consisted of all patients discharged alive after a HF hospitalization in 2006-07. Patients were classified as either HF clinic or standard care patients and followed until March 31st, 2010, to evaluate mortality, all-cause hospitalization, and HF hospitalization. Propensity score matching was used to compare outcomes between comparable groups of patients in the two groups, using Kaplan-Meier survival curves. We explored the clinic level characteristics associated with improved outcomes by developing marginal Cox-proportional hazard models, restricted to the overall sample of HF clinic patients, so as to account for clustering by HF clinic.
Results
We identified 14,468 HF patients, of whom 1,288 were seen in HF clinics. In a matched sample of 1,288 pairs, systematic differences between groups were substantially reduced. Over 3 years of follow-up, 52.1% of HF clinic patients died, compared to 54.7% of standard care patients (p-value 0.02). HF clinic patients had a significant increase in hospitalization (87.4% vs 86.6% for all-cause [p-value 0.009]; 58.7% vs 47.3% for HF-related [p-value <0.001]). Clinics in the high intensity strata were associated with lower mortality (hazard ratio [HR] 0.68 (95% confidence interval [CI] 0.48-0.98; p-value 0.04) but higher rates of all-cause hospitalization (HR 1.48; 95% CI 1.01-2.18; p-value 0.04) and HF hospitalization (HR 1.98; 95% CI 1.42-2.77; p-value <0.001), compared to low intensity clinics. HF clinics that targeted both the patient and caregiver were associated with improved survival compared to those that only focused on the patient, as were clinics with an emphasis on peer support. Clinics with frequent contacts between providers and patients had a significant reduction in mortality (HR 0.15; 95% CI 0.09-0.25; p-value <0.0001). A more intensive medication management program was associated with reduced all cause and HF hospitalization (HR 0.35 and HR 0.27 respectively).
Conclusions
Multi-disciplinary HF clinics are associated with a decrease in mortality but increase in re-hospitalizations compared to standard care. A gradient was observed between clinic intensity and outcomes whereby greater intensity of clinic services was associated with mortality reductions but increased hospitalization.
Implementation and Methodology of a Multidisciplinary Disease-State-Management Program for Comprehensive Diabetes Care
