MEASURING WHAT MATTERS MOST FOR HEMODIALYSIS OUTCOMES IN INDIA

Maintenance hemodialysis (HD) is the most common renal replacement therapy for end‑stage kidney disease (ESKD) in India. While dialysis treatments have remained out of reach for a significant proportion of the afflicted population, this situation has improved significantly over the past decade. There are ongoing initiatives to expand dialysis services  across the country by the government as well as by private investment. However, in the absence of a fully functional dialysis registry, identification of the parameters that affect patient outcomes in India are currently lacking. This study evaluated the survival outcomes of a large diverse cohort of patients on HD treated at a nationwide dialysis service provider network, to identify factors influencing mortality.

All patients initiating maintenance HD at the 288 centres across 16 Indian states (between August 15, 2008, and April 30, 2025), were included. The outcomes of 24,775 patients who survived more than 90 days were analysed, to provide quality indicators and benchmarks for Indian hemodialysis. Using death and dialysis discontinuation as the event, we estimated Kaplan–Meier survival and compared subgroups by age, gender, centre type (public vs private), diabetes mellitus (DM) status, and mean weekly dialysis frequency. We also used Cox proportional hazards regression to examine the association between risk factors and the event.  

The composite event endpoint occurred in 7,035 patients (deaths: 6,588, dialysis discontinued: 447). The patient survival in the analytic cohort at 1 year was 88% (95% CI, 87–88%), and 58% (95% CI, 57–59%) at 5 years. Survival declined with age: <45 years, 70% (95% CI, 69–72); 45–65 years, 56% (95% CI, 54–57); >65 years, 41% (95% CI, 39–43); (p < 0.001 for inter-group comparisons). In the analytic cohort, diabetes mellitus was present in 39% of patients (n = 6032). The 5‑year age-adjusted survival was lower in patients with DM (50% [95% CI 47–52]) than the rest (61% [95% CI 60–63], p<0.001). A graded association was observed between mean dialysis frequency and 5‑year survival (age and DM adjusted): >2.5–3/week, 69% (95% CI 67–71); >2–2.5/week, 59% (95% CI 56–62); <2/week, 53% (95% CI 52–55) (p<0.001 for intergroup comparisons). There was no difference in survival (age and DM adjusted ) between the genders; males 58.3% [95% CI 57–59] versus females 58% [95% CI 56–60, p=0.1]. However in the multivariate Cox proportionate hazard model, females had a 15% higher risk of mortality as compared to males (p<0.001).There was a significant difference in survival (age and DM adjusted ) between the private centres: 68% [95% CI 66-70] versus public centres: 58% [95% CI 56-60 (p<0.001). However in the fully adjusted Cox proportionate hazard model, which additionally incorporated dialysis frequency and stratified baseline hazard by initiation era, the centre type had minimal impact on the odds of an unfavourable outcome (p=0.6). In the analytic cohort, 6,588 deaths occurred (median age 55.5 years; median time on HD, 551 days); the majority of patients who died (68%) received ≤2.5 sessions/week. The cause of death was known in 5,352 (81%) patients. Cardiac, infectious, and neurologic causes accounted for 90% of known‑cause deaths. 

In this large multi‑centre Indian cohort, 5‑year survival on maintenance HD was 58.2%. Survival was most strongly influenced by age, dialysis frequency, and DM status. As a modifiable risk factor, higher treatment frequency (≥2.5 mean sessions/week) was associated with a significantly better survival, underscoring a call for adherence to a standardised thrice-weekly HD dosing as a key treatment guideline to improve patient outcomes. There was no survival disparity between patients treated in the private centres versus the public centres, reinforcing the role of organisational quality metrics in dialysis delivery.