Predicting clinical outcomes in peritoneal dialysis patients using small solute modeling.

The power of published models of dialysis adequacy to predict clinical outcomes in renal failure patients maintained on continuous ambulatory peritoneal dialysis (CAPD) is controversial. Inflammation may be an important predictor of morbidity and mortality in CAPD. Baseline data from a 2-yr prospective, longitudinal study of peritoneal dialysis adequacy were analysed. Baseline measures of dialysis adequacy (urea clearance [Kt/V], efficiency number [EN], dialysis index [DI], dialysate-plasma creatinine ratio [D/ Pcreat], creatinine clearance [CrCl weekly PD]) as predictors of outcome were investigated by univariate analysis and by multiple logistic regression modelling. Baseline nutritional and inflammatory markers were also tested as predictors of outcomes. Outcomes were patient survival and technique failure over the succeeding 2 yr. Fifty-three patients consented to the study; 7 patients were unsuitable. Only 6 patients completed the study (13%). Non-survivors (n = 6) had lower protein catabolic rates and lower serum albumin concentrations, and higher C-reactive protein (CRP) levels at baseline than the patients who survived (p <0.05), but there were no differences in any of the measures of dialysis adequacy. The patient group that developed technique failure (n = 9) had significantly higher D/Pcreat (p = 0.037) at baseline. Serum albumin and CRP at study entry were significant negative and positive predictors of death respectively (p <0.05). No baseline variable achieved significance as a predictor of technique failure in the patient cohort. In conclusion, dialysis dose descriptors are poor predictors of clinical outcomes in CAPD patients. Inflammatory and nutritional markers such as CRP and albumin may be more important in predicting patient outcomes than measures of peritoneal small solute clearance.