Objective: To define risk factors for ultrafiltration failure (UFF) during early stages of peritoneal dialysis (PD).
Design: Retrospective analysis of a group of patients whose peritoneal function was prospectively followed.
Setting: A tertiary-care public university hospital.
Patients: Nineteen of 90 long-term PD patients required a peritoneal resting period to recover UF capacity: 8 had this requirement before the third year on PD (early, EUFF group) and 11 had a late requirement (LUFF group). The remaining 71 patients, those with stable peritoneal function over time, constituted the control group.
Main outcome measures: Peritoneal UF capacity under standard conditions (monthly) and small solute peritoneal transport (yearly).
Results: None of the conditions appearing at the start of PD or during the observation period could be definitely identified as the cause of UFF. There were no differences in characteristics between the EUFF group and the other two groups, except for the higher prevalence of diabetes in the EUFF group. Residual renal function (RRF) declined in all three groups during the first 2 years, with rapid loss during the third year in the EUFF group. This rapid loss in RRF was coincident with UFF. Peritoneal solute and water transport at baseline was similar in the three groups. After 2 years on PD, individuals in the EUFF group showed a significantly lower UF and higher creatinine mass transfer coefficient values than those in the LUFF group. Diabetic patients in the control group showed remarkable stability in UF capacity over time. During the second year on PD, requirement for increases in dialysate glucose concentration was 3.4 +/- 0.5% in the LUFF group, but as high as 25.5 +/- 24.2% in the EUFF group. The accumulated days of active peritonitis (APID, days with cloudy effluent) were similar for the three groups after 1, 2, and 3 years on PD. Interestingly, diabetic patients in the control group showed an APID index significantly lower than the overall EUFF group. Diabetics in the control group also had significantly lower APID versus nondiabetics in the control group (p = 0.016).
Conclusions: Our findings suggest that certain patients develop early UFF type I. Diabetic state and a higher glucose requirement to obtain adequate UF suggest that glucose on both sides of the peritoneal membrane could be responsible. The mechanisms for this higher requirement remain to be elucidated. The identification of a larger cohort of these early UFF patients should lead to a better exploration of the primary pathogenic mechanisms.