Determining the optimal disease management strategies for patients with cardiovascular (CV) disease is an ongoing area of research. Several studies comparing a nurse-led home-based intervention (HBI) with standard care (SC) in patients with chronic heart disease have recently been completed. Simon Stewart, MD, PhD, Australian Catholic University, Melbourne, Australia, presented the results of a composite analysis of 3 randomized trials: the WHICH? trial [Stewart S et al. Int J Cardiol. 2014], the SAFETY trial [Stewart S et al. Lancet. 2015], and the NIL-CHF study [Stewart S et al. Eur J Heart Fail. 2015]. The hypothesis was that HBI would be superior to high levels of SC in the prevention of repeated hospitalizations and premature mortality and that the effectiveness of HBI would increase as the complexity of the clinical cases increased.

The 3 trials enrolled patients across the spectrum of cardiac disease. The WHICH? trial enrolled patients with heart failure (HF) with reduced ejection fraction and preserved ejection fraction, the SAFETY trial enrolled patients with chronic atrial fibrillation without HF, and the NIL-CHF study enrolled cardiac patients, most with acute coronary syndrome, without HF. In all of the studies, patients were recruited during acute hospitalization before returning home. All 3 trials were compliant with the Consolidated Standards of Reporting Trials, had independent data management and statistical analysis, and had blinded end point acquisition and adjudication. Follow-up ranged from 2 years (SAFETY) to 3 years (WHICH? and NIL-CHF).

A total of 1226 patients were analyzed, of which 612 received HBI and 614 received SC. The demographics of the study cohort were well matched across interventions. Patients were older (approximately 70 years), had multiple comorbidities and high clinical complexity, and had received appropriate levels of treatment. Thirty percent of patients were women.

Several aspects of recurrent hospital stays significantly favored HBI, including the median length of stay per patient in unplanned admission days (P = .011), CV admission days (P = .039), and all admissions (days in the hospital; P = .017). Further, all-cause mortality was lower with HBI vs SC (15.4% vs 20.2%; adjusted HR, 0.56; 95% CI, 0.41 to 0.78; P = .001). Patients in the HBI group also achieved a mean of 1210 ± 463 days alive and out of the hospital (90.1%; 95% CI, 88.2 to 92.0) compared with 1184 ± 494 days event free in the SC group (87.2%; 95% CI, 85.1 to 89.3; P = .02).

HBI was associated with worse event-free survival in lower clinical complexity cases, and Dr Stewart noted that HBI worked best when the clinical complexity was increased. Accordingly, to reduce the chance for harm, HBI should be reserved for cases that are more clinically complex.

Limitations of this study include that it was a post hoc analysis of studies in which the participants were not blinded. The mechanisms through which HBI increases events at low clinical complexity and benefits cases of high clinical complexity need to be further explored.