Remote care, unequal entry: the equity gap in digital heart failure program

Digital therapeutics programs (DTx) for heart failure (HF) such as a remote patient monitoring program reduce all-cause mortality, cardiovascular mortality and readmissions1 while accelerating guideline-directed medical therapy optimisation2. Structured programs with digitally-enabled care support further decrease length of stay and mortality3. However, digital health access remains inequitable: older adults, minorities, and rural populations face barriers including limited broadband, device unavailability, digital illiteracy, and cost constraints 4. Previous work showed equitable DTx participation across marginalisation levels within an Ontario, Canada outpatient heart function clinic5. However, whether referral to such programs occurs equitably remains unknown. This study examines economic, ethno-racial, age-based, and social marginalisation between HF patients referred and not referred to DTx in the clinic. We conducted a retrospective observational cohort study of adult HF patients attending an outpatient heart function clinic at a tertiary centre in Ontario, Canada, between April 2022 and May 2025. Patients were stratified by referral status to an DTx program (referred vs not referred) by 10 HF specialists. Equity-related variables included age, sex, language, patient portal registration status, distance from clinic, and area-level marginalisation. Area-level marginalisation was assessed using the Ontario Marginalization Index (ON-MARG 2021), across four domains: Households & Dwellings, Material Resources, Age & Labour Force, and Racialized & Newcomers. Postal codes were linked to dissemination areas and categorized into quintiles (Q1=least to Q5=most marginalised) using the Statistics Canada Postal Code Conversion File and R version 4.3.1. Among 1088 patients (622 referred; 466 not referred) (Table 1), mean age was similar (63.9 vs 61.9 years; p=0.053), as was sex distribution (66.4% vs 64.6% male; p=0.578). More referred patients resided >100 kms (17.1% vs 10.1%; p=0.018). Language preference was similar (9% vs. 8% non-English among n=808 with data, p=0.730). Patient portal activation was higher in referred patients (88.2% vs 83.0%, p=0.009). Marginalisation patterns varied (Figure 1.) Material resources distribution differed (χ²=9.96, p=0.04), with referred patients more concentrated in the most marginalised quintile (Q5: 22.5% vs. 17.4%, p-trend=0.006). Racialized & Newcomers distribution differed (χ²=16.42, p=0.003), with referred patients overrepresented in the least marginalised quintile (Q1:13.2% vs. 6.7%) but similar at Q5 (31.2% vs. 29.2%). Age & Labour Force showed significant trend (p-trend=0.025), with non-referred more concentrated in Q5 (22.3% vs 15.8%). Households & Dwellings did not differ (p=0.58). Referral patterns showed geographic equity and appropriate targeting by material deprivation but underrepresentation of racialised and employment-dependent groups.