Limitations of the peritoneal equilibration test in prescribing and monitoring dialysis therapy

Based on correlation analysis with 24-h dialysis collections the peritoneal equilibration test (PET) has been promoted as an aid to prescribe and monitor dialysis dose. However correlation is an incorrect statistical technique to demonstrate the similarity of one measure to another. The closeness (or limits) of agreement should be measured using the technique of Bland and Altman. One hundred and nineteen patients underwent a 24-h dialysate collection and a PET. D/P ratios for urea and creatinine, dialysate volume, urea and creatinine clearance and KT/V (urea) were calculated using both methods and compared using correlation analysis. In addition the limits of agreement, reflecting the potential margins of difference between the two methods, were determined. When used to calculate a daily dialysate volume required to achieve adequacy targets, the PET was found to result in a prescription error range of -0.6 to +1.51/day for creatinine clearance and -0.9 to +0.61/day for urea clearance. The tendency of the PET to exaggerate clearance resulted in 14% of patients incorrectly achieving a target creatinine clearance of 50 1/week and 17% incorrectly reaching a target KT/V of 1.7. The PET cannot be used in place of 24-h dialysis collections to prescribe or monitor dialysis therapy.