
OR WAIT null SECS
A retrospective cohort of 34,398 veterans found similar hospital days in VA and community dialysis settings across pandemic periods.
Veterans receiving hemodialysis in Veterans Affairs (VA) clinics had lower monthly mortality than those dialyzing through VA Community Care (VACC). The gap held through each phase of the COVID-19 pandemic.¹
Virginia Wang, PhD, MSPH, presented the retrospective cohort data at the American Society of Nephrology (ASN) Kidney Week 2026 in Denver, Colorado. The data suggest that in-house VA dialysis may be associated with a survival advantage that did not erode under pandemic strain. The investigators noted the findings could inform VA system planning and decisions about outsourcing dialysis care.
Veterans with end-stage kidney disease (ESKD) are disproportionately affected by social disadvantage and severe illness. Many rely on VA-financed care delivered both by VA facilities and by community providers.¹ Earlier work from Wang’s team found that demand for VA dialysis exceeds VA capacity, and travel distances keep many veterans from VA facilities. As a result, most veterans received dialysis in community settings between 2008 and 2013.²
“My team and I have had longstanding research comparing the outcomes of veterans receiving care in the VA through VA purchased care, what we now call VA community care programs, as well as those veterans who bypass their VA benefit and receive dialysis treatments through their Medicare benefit,” said Wang, research health scientist with Health Services Research and Development at the Durham VA Health Care System and professor in the Department of Population Health Sciences at Duke University School of Medicine, in an interview with HCPLive. “We had found differences then. Now, with the global pandemic, we were very curious the extent to which those differences may have changed over the course of the pandemic.”
The investigators noted that heavy reliance on care across VA and non-VA systems may increase care fragmentation and adverse outcomes, which likely worsened during the pandemic.¹ Whether outcomes differed between VA and VACC dialysis during this period had not been examined.
The team analyzed a cohort of veterans receiving VA-financed hemodialysis. Each veteran was assigned to their most common setting, either VA or VACC. Outcomes of interest were hospital days and mortality.¹
The investigators compared outcomes across 3 periods:
They also examined differences in monthly mortality by race (White, Black, and Hispanic/Other). Changes were estimated with marginal structural mean models, fit using inverse probability of treatment weighted generalized estimating equations to address confounding.¹
Of 34,398 patients, 29% received dialysis in VA and 71% in VACC. Patients in VA dialysis had greater comorbidity.¹
After weighted adjustment, estimated monthly hospital days did not differ between VA and VACC dialysis (2.3 vs 2.0 days). Hospital days also did not change across COVID-19 periods (P = .36).¹
Mortality was lower among veterans in VA dialysis in every period. Odds ratios for mortality were:
Estimated monthly mortality for VA vs VACC was:
These differences did not vary across COVID-19 periods (P = .43), including by race (P = .62).¹
The investigators wrote that outcomes for veterans receiving VA-financed dialysis were not affected by COVID-19. They noted this contrasts sharply with previously reported experience of patients receiving dialysis in US private-sector community settings. The mechanisms behind the VA’s pandemic experience and the lower mortality in VA clinics warrant further investigation, they added.¹
“What are the underlying factors associated with the better outcomes we’re observing in the VA clinics versus the non-VA clinics?” Wang said. “There’s something really special, we think, about the structure of the VA health care system, not only for dialysis-specific treatments, but the entire constellation of services and care that veterans receive in-house in the VA system.”
Editor’s Note: Wang reports no relevant disclosures.
Related Content: