SURVIVAL ANALYSIS OF PATIENTS ON LONG-TERM MAINTENANCE HEMODIALYSIS IN INDIA

Introduction:

Maintenance hemodialysis (MHD) forms the largest renal replacement therapy option for CKD patients in India. Several funding initiatives to extend the scope of this therapy to a wider population nationally are underway. However, there is sparse data on the survival outcomes of MHD patients from India. This impacts the measurement and comparisons of the effectiveness of these initiatives.

Methods:

All 36415 patients who were enrolled for maintenance hemodialysis at Apex Kidney Care across the country over the past fifteen years were studied. This data was collected from 220 dialysis centres from Aug 15, 2008 to December 31, 2023. A total of 19974 patients who survived beyond the first 90 days after dialysis initiation were included. Patients were censored for death, recovery from acute kidney injury (AKI), switch to peritoneal dialysis, kidney transplantation, transferred to a different centre, lost to followup and data not available. The impact of patient age, gender, hemodialysis frequency, diabetic status, and type of facility (public or private) on long term patient survival was calculated.

Results:

The median patient survival at 5 years was 61.98%. Age inversely impacted survival, with a 5 year survival of 54.06% for >65 years, 62.55% for 45-65 years and 69.95% for <45 years (p <0.005 for intergroup comparisons). The median 5 year survival of males (62.4%) was higher than females (61.13%) (p=0.0323). Public patients (n=8659) had a median survival of 61.86% as compared to 62.07% for private patients at 5 years (n=11315) (p<0.005). A 5 year median survival of 74.45% was seen in patients on 3-3.5 dialysis/week, 68.30% for 2.5-3 dialysis/week, 61.56% for 2-2.5 dialysis/week and 57.53% in <2 dialysis/week. A small group of patients on >3.5 dialysis/week had a survival of 51.22% (p<0.005). The adjusted prevalence of Hypertension (HTN) was 64.46%. The adjusted prevalence of Diabetes Mellitus (DM) was 42.51%. The 5 year median survival of DM patients was 52.61% (50.39-54.79) and for non DM it was 72.45% (70.84-73.99) (p<0.0001). Deaths were recorded in 4793 patients, whose median age was 58.97 years (IQR 47.64 - 68.17) and they spent 544 days (IQR 244 - 1138) on dialysis. 72.48% of these patients had a frequency <2.5 sessions/week. The cause of death was known in 3487 patients (72.75%). Infections, cardiac and neurological ailments caused more than 90% of these deaths. 963 deaths occurred due to 4 causes potentially linked to dialysis frequency namely, hyperkalemia, pulmonary edema, accelerated HTN and intracranial bleeding. They were responsible for 28.33% of deaths in patients on <2.5 sessions/week, 26.20% for 2.5-3.5/week and 20% for >3.5/week. A total of 686 patients discontinued dialysis (1.88%) after a median duration of 59 days (IQR 12,293). Their median age was 56.96 years (IQR 43.63, 67.29). More patients from public centres (2.41%) discontinued dialysis in comparison to private centres (1.62%) (p<0.0001). Most patients (74.42%) discontinued dialysis out of personal choice, terminal illness, failure to thrive or family apathy. Only 25.58% did so for financial and travel related reasons. Only 658 patients (3.3%) underwent a kidney transplant. Males (3.27%) had a higher transplant rate than females (2.64%) (p<0.014). Recovery from AKI occurred in 1491 patients (4.1%), with a majority (85%) recovering within the first 90 days.

Conclusions:

The long-term survival for hemodialysis patients in this dataset is comparable to Western registry data. Patient age, frequency of dialysis, and diabetic status were the most significant parameters influencing survival. Improvement of patient survival in hemodialysis is an ongoing clinical need, which needs risk factor analysis based on robust existing data.

Declaration: A part of this abstract was also submitted for the ‘Indian Society of Nephrology’ conference, Kolkata, 2023.

I have no potential conflict of interest to disclose.

I did not use generative AI and AI-assisted technologies in the writing process.