Baseline and longitudinal determinants of survival in end-stage renal disease: A joint longitudinal-survival analysis.

Abstract

Introduction: Evidence remains limited regarding how repeated biomarker measurements influence survival in individuals with end-stage renal disease (ESRD). This study examined the prognostic role of both baseline biomarker levels and their longitudinal patterns in ESRD patients.

Materials and methods: This retrospective cohort included maintenance hemodialysis patients from Shahid Ayatollah Ashrafi Esfahani Hospital, Tehran, Iran, with at least 3 months of dialysis history. Cox regression assessed baseline predictors, while a joint longitudinal-survival model evaluated repeated albumin, hemoglobin, Kt/V, and urea reduction ratio (URR) measurements in relation to survival.

Results: Overall, 330 ESRD patients were analyzed, with a mean age of 55.63 ± 15.57 years. In the joint model, each one-unit increase in the estimated current value of albumin, hemoglobin, Kt/V, and URR was linked to a lower risk of mortality over follow-up, with hazard ratios of 0.996 ( P = 0.009), 0.999 ( P = 0.015), 0.994 ( P = 0.039), and 0.988 ( P = 0.043), respectively. In Cox regression, higher white blood cell count increased mortality risk (Hazard ratio [HR]: 1.09, P = 0.025), whereas higher baseline log-ferritin, hemoglobin, and URR reduced risk, with HRs of 0.83 ( P = 0.042), 0.85 ( P < 0.001), and 0.13 ( P = 0.035), respectively.

Conclusions: The current study showed that longitudinal changes in markers such as albumin, hemoglobin, Kt/V, and URR may be important predictors of survival in patients with ESRD. Evaluation of baseline marker levels, along with longitudinal changes, nutritional status, age, and body mass index, may be important for determining survival among patients with ESRD.