Abstract
BACKGROUND: Heart failure (HF) causes substantial morbidity and mortality, in part because evidence-based pharmacotherapies remain underused. Remote, asynchronous digital platforms may accelerate the initiation and titration of pharmacotherapies outside of traditional clinic visits; however, their efficacy and safety have not been adequately tested in randomized trials. METHODS: VITAL-HF (Virtual Care to Improve Heart Failure Outcomes) was a multicenter, parallel-arm, unblinded, prospective, randomized controlled trial. The primary objective was to assess the effect of a remote, digital intervention targeting the use and titration of evidence-based therapies for HF. Adults ≥18 years old with stable HF with reduced ejection fraction (HFrEF), a left ventricular EF ≤ 40%, not yet optimized on HF therapy, and with smartphone access were eligible. Participants were randomized 1:1 using a randomization scheme generated by computer software and uploaded into a secure, web-based application to the digital intervention or usual care. Each participant in both treatment groups received a cellular-connected blood pressure cuff and scale. Participants were asked to measure vital signs and report symptoms daily. Data were transmitted to treating clinicians along with personalized treatment recommendations. Individualized care plans were developed by local clinicians and implemented with assistance from the digital platform and associated health coaches. Participants were then followed with study visits at 3 and 6 months after enrollment The primary outcome was the change in a HF medical therapy score (range 0-12) developed by the HF Collaboratory. Key secondary outcomes included other metrics of medication titration and clinical outcomes, such as those related to HF medication use (e.g., hyperkalemia). FINDINGS: A total of 178 participants were randomized (90 intervention, 88 usual care) across 7 US sites. The median age was 66.4 (interquartile range [IQR], 56.1-73.1) years, 39% were women, 17% were Black, and the median left ventricular EF was 26% (IQR, 20-35). Baseline medical therapy included 84.8% beta-blockers, 48.9% angiotensin receptor neprilysin inhibitors, 48.9% mineralocorticoid receptor antagonists, and 47.2% sodium-glucose cotransporter 2 inhibitors. At six months, the mean HF Collaboratory score increased from 5.2 (±1.9) to 7.2 (±2.0) in the intervention group, a mean change of 2.0 (±2.4); and in the usual care group, the mean score increased from 4.8 (±2.5) to 6.3 (±2.1), a mean change of 1.4 (±2.2). The least squares mean difference between intervention and usual care scores was 0.78 (95% confidence interval, 0.21-1.34; p = 0.007). There were 4 possible safety events in the digital group and 7 in the usual care group. INTERPRETATION: Among participants with HFrEF who were not yet optimized on medical therapy, use of a remote digital intervention outside routine clinical encounters was safe and improved the HF medical therapy score at six months compared to usual care. FUNDING: The VITAL-HF study was funded by Story Health.