Evolution Over Time of Volume Status and PD-Related Practice Patterns in an Incident Peritoneal Dialysis Cohort

Abstract

Background and objectives: Volume overload is frequent in prevalent patients on kidney replacement therapies and is associated with outcome. This study was devised to follow-up volume status of an incident population on peritoneal dialysis (PD) and to relate this to patient-relevant outcomes.

Design, setting, participants, & measurements: This prospective cohort study was implemented in 135 study centers from 28 countries. Incident participants on PD were enrolled just before the actual PD treatment was started. Volume status was measured using bioimpedance spectroscopy before start of PD and thereafter in 3-month intervals, together with clinical and laboratory parameters, and PD prescription. The association of volume overload with time to death was tested using a competing risk Cox model.

Results: In this population of 1054 participants incident on PD, volume overload before start of PD amounted to 1.9±2.3 L, and decreased to 1.2±1.8 L during the first year. At all time points, men and participants with diabetes were at higher risk to be volume overloaded. Dropout from PD during 3 years of observation by transfer to hemodialysis or transplantation (23% and 22%) was more prevalent than death (13%). Relative volume overload >17.3% was independently associated with higher risk of death (adjusted hazard ratio, 1.59; 95% confidence interval, 1.08 to 2.33) compared with relative volume overload ≤17.3%. Different practice patterns were observed between regions with respect to proportion of patients on PD versus hemodialysis, selection of PD modality, and prescription of hypertonic solutions.

Conclusions: In this large cohort of incident participants on PD, with different treatment practices across centers and regions, we found substantial volume overload already at start of dialysis. Volume overload improved over time, and was associated with survival.

Keywords: Confidence Intervals; Follow-Up Studies; Hypertonic Solutions; Proportional Hazards Models; Prospective Studies; Renal Replacement Therapy; Spectrum; Treatment Outcome; Water-Electrolyte Imbalance; bioimpedance; diabetes mellitus; fluid overload; fluid status; peritoneal dialysis; renal dialysis.

Graphical abstract

Figure 1.

Patterns of prescription practice by region, with (A) use of APD versus CAPD and (B) use of polyglucose and hypertonic solution.

Figure 2.

Relative volume overload at baseline (left panel) and mean change during the study (right panel) by (A) sex, (B) diabetes status, and (C) region. The Δ of relative volume overload/depletion in percent points was calculated for each patient and visit (baseline, month 1, month 3, month 6, and so on, until month 36) as the absolute difference between relative volume overload or depletion at the respective visit and at baseline. The plots show the interquartile range (IQR; box), mean (cross), median (line), minimum and maximum value within the IQR±1.5 IQR (whiskers), and outliers (circles). AP, Asia Pacific; LA. Latin America; EEME, Eastern Europe & Middle East; WE, Western Europe.

Figure 3.

Relative volume overload at baseline (left panel) and mean change of relative volume overload (right panel) of participants in the following categories: volume depleted (<−7%), euvolemic (−7% ≤ relative volume status ≤7%), and volume overloaded (>7%). The plots show the interquartile range (IQR; box), mean (cross), median (line), minimum and maximum value within IQR±1.5 IQR (whiskers), and outliers (circles).

Figure 4.

Cumulative incidence of death by volume status 1 month after initiating PD, adjusted for competing risks of transfer to HD and transplantation in participants with relative volume overload >17.3% versus ≤17.3% for a participant with median age, no diabetes, and no cardiovascular disease.