Our findings characterize the past 6 years of HF care within the VA healthcare system. Medication rates were either sustained or improved throughout the study. The existing VA telehealth model increased in use through the peak of the pandemic. Hospitalizations declined and sustained rates at a lower level than the pre-pandemic period; in contrast, mortality increased and has sustained rates at a higher level. In survival analysis, age, male sex, and numerous comorbidities were associated with increased mortality over time.

Extrapolating from pre-COVID trends, the onset of COVID in March 2020 potentially limited 500-1200 Veterans from receiving GDMT, with newer medications like SGLT2i and ARNI most impacted (Table S8). This amounted to only ∼1% of the eligible HF population, and all medications recovered impressively, with observed prescriptions at study end being between 400 and 19,000 higher than expected based on pre-COVID trends. Again, newer therapies like SGLT2i and ARNI noted the most accelerated adoption during the early and late COVID periods. Most importantly, no medications had observed rates that fell relative to pre-COVID trends. In sum, most Veterans with HFrEF received GDMT throughout the pandemic and afterwards. However, these rates for each class of medication remained well below ideal rates and those reported for HF specialty programs.²

These findings are supported by similar studies from the early COVID period in Veterans with recently diagnosed HFrEF, whose authors suggest that telehealth may have provided continued access through the pandemic.²³ Outside of VA, a study of adults with chronic diseases – including HF – revealed COVID-19-related medication barriers were common, with 51% of respondents reporting a medication-related problem attributed to COVID’s disruption of the healthcare system, and ∼20% noting difficulty obtaining their medications because of offices closed for in-person visits.²⁴ A systematic review noted that COVID led to major interruptions in chronic disease management, particularly related to difficulty in reaching physicians and receiving medications in a timely manner.²⁵ The VA integrated healthcare system model appeared robust, in particular its pharmacy infrastructure, which maintained medication access despite massive healthcare disruptions.

In contrast to settings outside VA, the VA was already delivering a considerable amount of telehealth prior to the pandemic and expanded considerably.¹⁷ Both primary care and cardiology saw a rapid and meaningful growth in telehealth use before partially returning to traditional face-to-face visits in the late COVID period. This pattern aligns with other specialties at VA, other healthcare systems, and across insurance coverage options.^26–29 The rapid transition to telehealth was thus not a unique solution for the VA; a systematic review noted multiple studies reporting telehealth as the primary solution to minimize treatment discontinuation and ensure high adherence.²⁵ Other healthcare systems also noted similar medication rates between HF patients with telehealth visits and those with in-person visits immediately after the pandemic.³⁰ Importantly, in studies both within and outside VA, transition to telehealth was not associated with worsening of clinical outcomes.^5,31,32 Altogether, this suggests some degree of telehealth adoption as the new state of US healthcare delivery.

A marked finding is a reduction in both all-cause and HF hospitalizations but increased mortality for the Veteran HFrEF population that sustained into the late COVID period. While lower volume of admissions for HF patients is documented in the early pandemic,³³ broader national trends suggest improving age-adjusted mortality rates for cardiovascular disease since the peak of the pandemic.³⁴ The elevated mortality in our study, despite maintained medication rates, successful transition to telehealth, and reduced hospitalizations persistent over several years, may reflect a “new normal.” Such findings reveal important limitations in traditional healthcare quality metrics during system-wide disruptions; process measures may be insufficient indicators of care quality in settings of massive healthcare system disruption when patients face barriers beyond the healthcare system’s direct control, such as delayed care-seeking, social isolation, financial hardship, or deferred preventive care. The persistence of the mortality finding four years post-pandemic onset may indicate a long-lasting shift in baseline risk for this highly vulnerable population or may relate to unmeasured confounders; further monitoring is clearly vital and warranted.

Our observed 6-year survival rate of 56.1% for Veterans with HFrEF is largely consistent with other national HF data, and demographic predictors were similarly consistent with the literature.⁴ Female sex was associated with better survival;^35–37 however, female sex was also associated with lower hospitalization risk, which contrasts with prior studies noting higher hospitalization rates among women with HF.^36,38,39 Otherwise traditional prognostic factors including lower hemoglobin and LVEF, as well as comorbidities like smoking and substance use, maintained their expected association with worse outcomes.^4,40–42