Measured-to-predicted creatinine generation ratio increases with time and decline in residual renal function in continuous ambulatory peritoneal dialysis

The expression of measured-to-predicted creatinine generation ratio (M/P) has been proposed as an index of compliance in continuous ambulatory peritoneal dialysis (CAPD) patients. Although M/P may not be sensitive enough for cross-sectional study, serial monitoring has been suggested to identify noncompliance. We attempted to evaluate serial changes of M/P from a nonselected group of CAPD patients. Sixty-three patients, all followed up for 2 years, were reviewed retrospectively. Their M/P ratios at years 0 and 2 were computed and compared. Baseline M/P had a normal distribution with a mean of 0.96 +/- 0.26. There was significant correlation between baseline M/P and residual glomerular filtration rate (GFR; r = -0.81; P < 0.0001). There were weak correlations between M/P and duration of dialysis (r = 0.52; P < 0.0001), body weight (r = -0.52; P < 0.0001), Kt/V (r = 0.31; P < 0.02), weekly creatinine clearance normalized to body surface area (r = 0.53; P < 0.0001), and serum albumin level (r = -0. 28; P < 0.05). After 2 years, M/P increased in 56 of 63 patients (88. 9%). Average M/P increased from 0.96 +/- 0.26 to 1.31 +/- 0.27 (P < 0.0001). Multivariant analysis showed M/P at year 0, which was largely determined by residual GFR, was the only independent factor affecting increase in M/P from year 0 to year 2. The general trend of increasing M/P was still present when only anuric patients were analyzed, although that was not statistically significant (1.21 +/- 0.14 to 1.32 +/- 0.24; P = 0.12). The finding of increasing M/P with time in CAPD patients, particularly those with significant residual renal function, suggests M/P may not be a reliable indicator of noncompliance, even for serial follow-up of the same patient. Better methods for assessment of compliance in CAPD patients are required.