Herpes zoster impacts the time to first change in diuretic treatment and healthcare resource utilisation in patients with heart failure

Background Herpes zoster (HZ) and its complications are associated with severe pain and reduced quality of life. In individuals with cardiovascular diseases, such as heart failure (HF), the immune-mediated inflammation caused by HZ may contribute to the destabilisation of their underlying condition.1 However, the clinical and healthcare impact of HZ among adults with HF remains poorly understood. Purpose To assess the impact of incident HZ on HF management and associated healthcare resource utilisation (HCRU) among adults aged ≥50 years with HF in Japan. Methods This is a retrospective cohort study based on the Medical Data Vision nationwide claims database of more than 600 acute-care hospitals in Japan (April 2008–December 2023). Patients with HZ were propensity score-matched 1:2 to patients without HZ based on key baseline characteristics, with exact matching on index year. The primary outcome was the time to first change in diuretics (defined as initiation of a diuretic class not used at baseline or dose increase during follow-up), up to 1 year after the HZ episode. The primary outcome was assessed using cumulative incidence probabilities (CIPs) and restricted mean survival time (RMST) differences (HZ−non-HZ), with 95% confidence intervals (CIs). Secondary outcomes evaluated the long-term impact of HZ and included the time to first change in diuretics at 3 and 5 years, as well as annualised all-cause HCRU (hospitalisations, length of stay, and outpatient visits) during follow-up, reported as median (quartile 1, quartile 3 [Q1, Q3]) with differences between cohorts measured with Wilcoxon rank sum test . Results The analyses included 2,378 HZ and 4,753 non-HZ patients with HF (median age [Q1, Q3]: 80.0 [72.0, 86.0] years for both cohorts; female: 60.9% [HZ] vs 59.4% [non-HZ]). Diuretic modification occurred earlier in the HZ cohort versus the non-HZ cohort at all timepoints (Figure 1). For the time to first change in diuretics at 1 year, CIP (95% CI) was 13.3% (11.9, 14.9) vs 9.8% (8.9, 10.8) with RMST difference (95% CI) of −9.6 days (−13.9, −5.2). For the time to first change in diuretics at 3 and 5 years, CIPs (95% CI) were 24.3% (22.3, 26.5) vs 19.3% (18.0, 20.8) and 31.8% (29.2, 34.6) vs 27.7% (25.8, 29.7) with RMST differences (95% CI) of −43.4 days (−61.6, −25.2) and −75.7 days (−111.3, −40.1), respectively. Annualised HCRU analyses showed higher all-cause hospitalisations (median [Q1, Q3]: 0.2 [0.0, 0.8] vs 0.0 [0.0, 0.4]; p<0.001) and outpatient visits (median [Q1, Q3]: 10.9 [6.1, 17.5] vs 7.3 [3.9, 13.0]; p<0.001), and substantially longer length of stay among those hospitalised (median [Q1, Q3]: 51.2 [4.7, 256.3] vs 16.7 [2.8, 153.6] days; p<0.001) in the HZ cohort compared with the non-HZ cohort. Conclusion Among patients with HF, HZ was associated with earlier diuretic modification and increased HCRU, suggesting a short- and long-term impact of HZ on HF management and overall healthcare burden.