Survival and Quality of Life on Maintenance Hemodialysis: Results from a Multicenter Study

Introduction

CKD is a growing global health concern, now the 7^th leading risk factor for mortality. The burden is particularly high in lower-middle-income countries like India, where 17% of the population is at risk, and 0.7% suffer from end-stage kidney disease (ESKD).^1,2 India has a CKD incidence of 229 cases per million, with a significant shortage of kidney donors, making maintenance hemodialysis (MHD) essential for many patients.^3,4 To address this, the Indian government, along with private and non-profit organizations, provides free or subsidized MHD. As of 30 September 2025, according to the PMNDP website, there are 1,743 dialysis centers across 751 districts in 36 states, equipped with 12,402 hemodialysis machines. These have delivered 375.88 lakh dialysis sessions to 29.46 lakh patients with ESKD.^3,5-7 However, data on patient survival and quality of life (QoL) in India remain limited.

The aim of the present multicenter study was to assess the survival and QoL of incident patients on MHD.^6,7

Materials and Methods

A total of 1039 patients were enrolled from 30 MHD centers from West, Central, and Southern Indian states [Figure 1]. The centers were from the Government, Not-For-Profit charitable, and the private sectors. There were 4 (13.33%), 7 (23.33%), and 19 (63.33%) public, private, and public-private partnership (PPP) model centers, respectively. The center-specific data have been published previously in this journal

Close

Figure 1:

Study flow. CAPD: Continuous ambulatory peritoneal dialysis.

Export to PPT

Incident patients on MHD defined as dialysis vintage >3 months and <1 year, >18 years, and of both sexes were included.

Patients on MHD for >1 year, those planning for a transplant, or migrating to PD within 1 year, and those with advanced malignancies, liver, lung, or heart failure, extreme frailty, or any other medical reasons with an anticipated survival of <1 year were excluded.

The study was initiated by a steering committee comprising Principal Investigators from all participating centers, which finalized key components, including the study protocol, data entry forms, and center registration forms. [Supplementary Forms].

Each enrolled patient was assigned a unique computer-generated ID, ensuring centers could only access their own data. Following registration with a username and unique password, data was collected electronically via online forms across four monthly follow-up visits. Offline data entry was also allowed, with the option to upload it to the study website. Saved data on the server could not be changed without the study administration’s approval. Any corrections during data validation required agreement between the study center and administration. Centers could download their data in Excel format at any time. Regular reports of cumulative data were shared with centers, maintaining patient privacy. Data was securely stored on a server, with weekly backups to a separate server. The baseline characteristics of the enrolled patients have been detailed in Table 1.

Causes of mortality definitions were: (i) Cardiac death- unexpected death due to cardiac cause, occurring in a very short time (< 1 hour) after onset of symptoms or Coronary event (documented acute myocardial infarction). (ii) Infection- clinically determined acute infection with biochemical or microbiological, or radiological evidence, including COVID infections, and (iii) Unknown cause- at-home mortality or undetermined demise.

Cardiac and non-cardiac death were based on documentary evidence/physician assessment [Supplementary Figure 1].

QoL was studied using a detailed questionnaire based on the EuroQOL-5 Dimensional 3 Level (EQ-5D 3L) questionnaire, which has been pretested and validated in the CKD population in India and elsewhere.^8-10 It was utilized locally in vernacular Indian languages based on the center location. EQ-5D 3L score ranges from 5 (best QoL) to 15 (poor QoL). The scores (individual and comprehensive) were noted every scheduled follow-up [Supplementary Table 1, 1a and Supplementary Figure 1a]. If the last follow-up QoL comprehensive score ≤ baseline QoL score, then the patient was deemed to have improved QoL. If the comprehensive QoL score > baseline QoL, then patients were deemed to have worsened QoL.

Results

The study enrollment started on 1^st April 2019 and was planned to continue till June 2027, but the study was stopped in June 2023 due to a lack of funding.

The mean age was 49±14.9 years (males: 48.92±15.20 years, females: 49.37±14.36 years) [Table 1]. The survival of MHD patients in the present cohort was 94.1%, 86.6%, 77.9%, 58.4%, and 47.1% at 4 months, 8 months, 1 year, 2 years, and 3 years, respectively. The survival of patients was 28.10±0.73 months [Figure 2].

Close

Figure 2:

Overall Kaplan Meier survival analysis.

Export to PPT

Of 1039 patients enrolled the study, 169 exited the study (118 patients moved to other non-participating centers, 8 converted to PD, 30 underwent unplanned renal transplant); 13 were lost to follow-up. [Figure 1]. The subjects were followed-up for 13.06 ± 10.06 months, with a total of 13,574 patient months. A total of 289 deaths were recorded; the risk of one death occurred every 46.96 patient months [Supplementary Table 2, Supplementary Figure 2].

Discussion

Despite improved availability of MHD in India over the last decade, patient outcomes are generally not known.. The mean age was 49 years, and 70.1% were male; incident dialysis patients in their productive part of life were affected by ESKD status. Median-survival time on dialysis for males and females showed no disparity. Along with socioeconomic factors, both qualitative and quantitative factors defining dialysis delivery may be responsible for this dismal outcome.

Uneducated individuals had significantly lower median survival compared to educated individuals, demonstrating that educational status is crucially influential in survival outcomes. Survival rates increased with higher levels of education, ranging from 23 months for primary education to 34 months for graduates. A lack of disease and treatment insight and timely dialysis access, and transplantation adversely affects survival.

Survival rates varied significantly based on the funding source for dialysis. Patients who were self-funded or insurance-funded had better survival compared to those with employer and government-funded dialysis. Government-funded dialysis patients experienced a QoL 1.44 times inferior to those whose dialysis was supported by employer or self-funding. Notably, employer-funded dialysis patients demonstrated the highest QoL. This could potentially be attributed to the imperative for these patients to maintain employment, coupled with the supportive environment provided by employers to foster productive employees.

A higher mortality risk in dialysis patients with both diabetes and hypertension was observed. Interestingly, patients with hypertension demonstrated better survival rates compared to normotensive patients. This paradoxical finding might be attributed to underlying cardiac failure in normotensive patients, potentially leading to earlier cardiac-related deaths. Furthermore, no significant differences in survival rates among patients with coronary artery disease (CAD) were noted when compared to the group with no CAD.

Patients with stroke had the worst survival rates, with the risk of stroke being 5-30 times higher in dialysis patients and an associated 90% case fatality rate. Factors such as vascular disease, inflammation, uremic toxins, and dialysis techniques contribute to increased stroke risk, particularly during the initial dialysis phase. In a study from South India, patients with CKD stages III-V experienced worse survival and diminished functional outcomes following a stroke.¹⁰

Patients with 2/week dialysis had numerically better survival than those on 3/week dialysis (p=0.051), with no survival difference between those who underwent emergent or planned dialysis initiation. A retrospective observational study of 9-month follow-up conducted in a stand-alone dialysis center in India also revealed an insignificant difference in mortality rate between patients on twice-weekly and thrice-weekly hemodialysis.¹¹ In patients with kidney failure, CVD remains the primary mortality driver and has become the focus of efforts to reduce mortality risk.¹² In a study from South India, 463 MHD patients of a mean age of 36 years were studied over 1 year, with a mean dialysis vintage of 100 days. Additionally, 92% were on 3/week-MHD and 8% were on 2/week-MHD; all-cause mortality was 9% with 60% deaths in the first 28 days of initiation of HD and 75% in the first 2 months.¹³ The inference that 2/week and 3/week have the same outcomes may be related to the short duration of study.

Several Indian studies provide valuable comparative median survival (or equivalently, mean/estimated survival times) data among patients undergoing MHD. For example, a retrospective cohort of 229 patients in rural Puducherry reported a median survival of 819 days (∼27 months) on dialysis.¹⁴ Another single-center Bengaluru study of 96 patients found an estimated mean survival time of 570 days (∼19.2 months) after dialysis initiation.¹⁵ A large nationwide network-based study of 23,601 patients from 193 centers reported that 71 % survived past 180 days and, after adjustment, the predicted 180-day survival ranged from 83 % to 97 % (median center ∼90 %).¹⁶ Thus, when interpreting our own cohort’s survival data, it appears consistent with, but somewhat modest compared to these Indian benchmarks, suggesting significant early-phase mortality and highlighting potential opportunities for improvement in dialysis delivery and support services.