COVID-19 vaccination status impact on mortality in end-stage kidney disease

To the Editor: Patients with end-stage kidney disease (ESKD) are particularly vulnerable to adverse clinical outcomes associated with COVID-19, with an estimated 20%–30% mortality risk.1–4 There is limited information on clinical outcomes, including mortality, following vaccination among patients on dialysis, with many relevant phase 3 trials excluding patients with “serious kidney disease” and chronic conditions. To characterize the impact of COVID-19 vaccination on all-cause mortality, we conducted a retrospective analysis using data from all US adults (i.e., aged ≥18 years) with ESKD receiving dialysis through Fresenius Medical Care (FMC) between March 1 and September 18, 2021. Data for a total of 239,660 patients were available with clinical and demographic, including vaccination status, being updated throughout the study period. As of March 1, 2021, 16,140 patients were classified as fully vaccinated (defined as the time period beginning 2 weeks after either an Ad26.COV2.S vaccination or a second mRNA vaccination), 39,938 were partially vaccinated (defined as the time period up to 2 weeks after vaccination with Ad26.COV2.S or the period from initial vaccination to 2 weeks after a second mRNA vaccination), and 114,403 were unvaccinated (defined as having no COVID-19 vaccination history). Among patients at the start of the study period (N = 170,481), 13.8% were on peritoneal dialysis, 3.8% were on home hemodialysis, and 82.4% were on in-center hemodialysis, 42.6% were female, 63.9% had a history of diabetes, and 51.6% were younger than age 65. Patient demographics remained relatively consistent throughout the study period. Overall, 102,717 patient-years of follow-up were available for analysis: 30,689 for unvaccinated patients, 14,478 for partially vaccinated patients, and 57,550 for fully vaccinated patients. During the analysis period, 19,356 deaths occurred, equating to an overall mortality rate of 18.8 deaths per 100 patient-years. The unadjusted rate of death (per 100 patient-years) was 29.6, 13.8, and 14.4 among unvaccinated, partially vaccinated, and fully vaccinated patients, respectively. When adjusted for sex, age, race/ ethnicity, diabetes history, and US geographic region, the risk of death was 159% higher among unvaccinated patients than that observed during times when patients were vaccinated (hazard ratio [HR]: 2.6). Unvaccinated patients were at higher risk regardless of modality (Figure 1). In FMC electronic medical records, COVID-19 was listed as the primary or secondary cause of death in 1207 of the total 19,356 deaths (1.18 deaths/100 years) and as the primary cause of death in 832 patients (0.81 deaths/100 years). It was estimated unvaccinated patients were six times more likely than fully vaccinated patients to have a COVID-19-related death using either definition (Figure 2). Furthermore, cardiovascular causes were listed as the primary or secondary cause of death for a total of 9280 patients; estimated unvaccinated patients were two times more likely than fully vaccinated patients to experience cardiovascular death. The estimated HRs (95% confidence interval [CI]) for the unadjusted and adjusted analyses comparing unvaccinated versus fully vaccinated states were 2.12 (2.06–2.19) and 2.59 (2.51–2.68), respectively. Additional sensitivity analyses (i.e., with two-way interaction terms, with age as a continuous variable, examining follow-up time relative to admission, and removing patients who initiated the study as fully vaccinated to assess possible immortal time bias) had HRs ranging from 2.57 to 2.65. An adjusted analysis comparing unvaccinated and partially vaccinated states demonstrated a survival benefit with partial vaccination (HR [95% CI]: 2.31 [2.20–2.43]). Finally, an analysis simultaneously assessing vaccination status and clinically reported COVID-19 history was conducted. When mortality risk among unvaccinated patients with COVID-19 history was compared to that of unvaccinated patients with no COVID-19 history, fully vaccinated patients with no COVID-19 history, and fully vaccinated patients with COVID-19 history, the hazard ratios were 1.2, 3.1, and 2.7, respectively. Our results are consistent with those from a smaller cohort of 35,206 hemodialysis patients vaccinated against COVID-19 and matched unvaccinated patients. In that analysis, the cumulative risk of all-cause mortality at day 120 was 0.031 among BNT162b2-vaccinated patients (vs. 0.071, p < 0.001) and 0.028 among Received: 7 June 2022 Revised: 6 November 2022 Accepted: 19 January 2023