Peritoneal Equilibration Test: Time on Treatment and Reference Values

Dr. G. Amici, Nephrology and Dialysis Division, Regional Hospital, I-31100 Treviso (Italy) Table 1. Mean PET measurements and differences between the two groups of patients with different times on treatment Dear Sir, We are using the peritoneal equilibration test (PET) with a 3.86% glucose solution [1] in peritoneal dialysis to evaluate and monitor peritoneal function and to optimize dia-lytic prescription. We are applying Twar-dowski’s statistical method [2, 3] of classifying patients into four categories according to their PET results for solute peritoneal transport – D/P(4 h) – and ultrafiltration capacity, using the mean and standard deviation of the studied population. In our opinion this method is at present the only practical way to link test results to peritoneal dialysis prescription. There were no doubts about interpreting results of a first cross-sectional study of our patients’ peritoneal transport characteristics. However, after 1 year, patients from the first round repeated the test while many new patients did it for the first time, and so we found ourselves faced with an important methodological problem. It would seem that a criterion has not yet been established for choosing patients to obtain the population from which reference values can be calculated and consequently the intervals of each of the four categories [2, 3]. Do we consider the changes of the mean and standard deviation of our PET population by adding new first-PET patients or is the first year’s classification the only valid one? How many patients are enough for this purpose? Do we consider time on treatment? It is not acceptable to periodically recalculate category intervals in a sort of ‘dynamic’ classification every time new PETs are added to the list. But it is not useful for every center to consider only the first-year results for classification because of the small number of patients. Of course, it is an error considering repeated PETs for this classification. We also feel that it is not completely correct in a cross-sectional study to include patients with just any length of time on treatment. In our opinion, usable data must refer to a population that can be considered ‘normal’ or ‘basic’, i.e. patients with a functionally intact peritoneum as far as peritoneal dialysis is concerned and consequently a reasonably short time on treatment. With this aim in mind we carried out a one-way factorial Anova for solute transport rates [D/P(4 h) for creatinine, BUN, potassium, phosphorus and D/Do(4 h) for glucose] and ultrafiltration taken from 57 PETs performed between November 5, 1991 and April 7, 1993 in 37 patients (31 PETs from the first cross-sectional study, 20 PETs repeated after 1 year and 6 new first PETs) among six groups with different times on treatment (0-3, 3-6, 6-12, 12-24, 24-36, >