More hospitalizations: Secret to improving HF mortality?

In what researchers acknowledge is a counterintuitive finding, a study of real-world outcomes in multidisciplinary heart failure (HF) clinics found a higher rate of hospitalizations in patients seen at clinics compared with those who were not. The results were published online Dec. 10 in Circulation: Heart Failure.

Harindra C. Wijeysundera, MD, PhD, of the Schulich Heart Centre at Sunnybrook Health Sciences Centre in Ontario, and colleagues wanted to assess HF clinics in a real-world population. They wrote that while trials have shown clinics to improve patient outcomes, the results don’t reflect the heterogeneity of clinics or interventions. “Moreover, there remains uncertainty about which components of specialized HF clinics are most important,” they wrote. “For example, are beneficial effects mediated through more aggressive medication titration, or through enhanced surveillance?”

Using administrative data in Ontario, they designed a study to compare the effectiveness of treatment at specialized HF clinics with usual care and pinpoint clinic characteristics that were associated with improved patient outcomes. They identified 14,468 adult patients who between 2006 and 2007 were discharged alive after an HF hospitalization, 1,288 of whom received treatment at specialized HF clinics. The primary effectiveness outcome was all-cause mortality, with secondary outcomes of all-cause readmission and hospitalization for HF through March 31, 2010.

They used propensity scores to match clinic and usual-care patients; cox-proportional hazards models to evaluate clinic-level characteristics that were associated with improved outcomes; and a validated HF Disease Management Scoring Instrument to derive intensity scores for the 21 HF clinics included in the study.  

All-cause mortality at the four-year follow-up was 52.1 percent in the clinic group compared with 54.7 percent in the control group. The clinic group had higher rates of all-cause readmission and HF hospitalization, at 87.4 vs. 86.6 percent and 58.7 vs. 47.3 percent, respectively.

The breakdown for high-, medium- and low-intensity clinics was eight, eight and five; they found no relationship between intensity classification and mortality or readmission. Increased intensity of medication management was associated with reduced all-cause and HF readmission, while more involvement of caregivers and more comprehensive self-care education programs were associated with increased hospitalization.

“Treatment at HF clinics was associated with a small but statistically significant reduction in mortality, but increased all-cause and HF readmissions,” Wijeysundera and colleagues noted, adding that they observed a more modest improvement in mortality than other studies. They cautioned that there was a potential for survivorship bias in their study.

They described the results on hospitalizations as “robust,” and wrote that their large sample size facilitated the evaluation of clinic-level characteristics and outcomes. The findings on caregiver involvement and self-care education programs may indicate that screening leads to earlier intervention and increased hospitalization.

“These counterintuitive findings … suggest that the mortality benefit afforded by HF clinics may be mediated in part by earlier hospitalization and intervention, and thus avoidance of critical deterioration,” Wijeysundera and colleagues wrote. “In this setting, one can argue that these hospitalizations are not avoidable, but may be an important mediator of improved survival.”

As they have done with previous HF studies, the researchers planned to next conduct a cost-effectiveness analysis.