Measuring total body water in peritoneal dialysis patients using an ethanol dilution technique

Background. The accuracy with which total body water (TBW) is estimated is a direct determinant of the reliability of Kt/V urea measurements in perito neal dialysis (PD) patients. Ethanol dilution has been previously shown to be a reliable measure of TBW. Advances in breath alcohol technology make th is a feasible clinical tool. Methods. We gave 19 fasting chronic PD patients 0.3 g/kg of ethanol (EtOH) orally on two separate occasions. Breath alcohol concentrations (BrACs), de termined by dual-beam infrared-analysis, were recorded at baseline and peri odically thereafter until BrACs were less than 0.01%. The TBW was then dete rmined by standard pharmacokinetic techniques. Results. TBW measurements were reproducible, with a mean between-run differ ence of -0.004 liter/kg (95% limits of agreement-0.040 to 0.032 by Bland-Al tman). The Watson equations tended to underestimate TBW, with a mean differ ence (EtOH - Watson) of +3.0 liters (SD 4.0 liters, P = 0.004) and a mean a bsolute difference of 4.1 liters (SD 2.7 liters, range -4.4 to 9.5 liters). Kt/V was calculated from dialysate and urine collection, using V as determ ined from TBW estimates from EtOH and Watson. The mean Kt/V(EtOH) was 2.31 (SD 0.50) compared with 2.46 (SD 0.52) using Watson. The mean absolute diff erence between the two Kt/V estimates was 0.26 (SD 0.20, range. -0.87 to 0. 57), with Kt/V overestimated by Watson in 14 patients. EtOH was well tolera ted, and the procedure was completed in about four hours. Conclusions. Measuring V by the BrAC technique does not require blood sampl ing, is reliable, and is reproducible. It is a potentially useful method fo r a periodic determination of volume that may allow for more accurate Kt/V measurement in PD patients.