Hyponatraemia is one complication of peritoneal dialysis1–4 which has been referenced as associated with the retention of osmolytes (icodextrin), hyperglycaemia, catabolic states, intracellular potassium or phosphorus loss5 and ultrafiltration failure.6 A positive correlation has been demonstrated between the loss of sodium and the ultrafiltrate volume.; thus adjusting salt intake to said volume is therefore recommended.3 This aspect of peritoneal dialysis can complicate clinical recovery from hyponatraemia.

We present the case of a 28-year-old woman with ESRD on peritoneal dialysis, (CAPD [continuous ambulatory peritoneal dialysis]) and rheumatoid arthritis, receiving treatment with prednisone. She arrived for consultation due to vomiting and diarrhoea that had persisted for a week. Her regular peritoneal dialysis regimen included two 1.36% glucose exchanges (4-h dwell time), one exchange with 2.27% glucose (6-h dwell time) and a fourth exchange with icodextrin (7.5%). With a CAPD she achieved a daily negative balance of around 1900ml. Residual diuresis was counted at around 200ml per day. The analytical control demonstrated evidence for hyponatraemia (Na: 119mEq/l), hypochloraemia (Cl: 78mEq/l) and hypokalaemia (K: 3.2mEq/l). Hydroelectrolytic reposition was begun with physiological saline supplemented with potassium chloride and the CAPD regimen was modified, eliminating the exchange with icodextrin and substituting it with a 1.36% glucose exchange. The three exchanges at 1.36% were maintained with a dwell time of 5 hours. Dwell time for the 2.27% glucose exchange was 9h. Microbiological studies were negative. Symptomatology was self-limited to the few hours after admission, during which volume reposition was completed and potassium levels were normalised, although there was no evidence that the hyponatraemia was resolved. The sodium sieving level was 15%. During the study, it was observed that, at the end of the dwell time for a 1.36% glucose exchange, the ratio between the sodium concentration of the peritoneal dialysate effluent and that of the plasma was 100%, which meant that, based on the plasma sodium concentration at 122mEq/l, the final sodium concentration in the dialysate effluent was 122mEq/l. The final exchange volume was 2300ml. The initial sodium concentration from the exchange was 134mEq/l, which would have corresponded to a theoretical final concentration of 116.5mEq/l. This discrepancy between the real value (122mEq/l) and the theoretical value (116.5mEq/l) pointed to a negative sodium balance and the exchange dwell time as the perpetuating causes of hyponatraemia. The next day, starting with a plasma sodium concentration of 120mEq/l, the exchange dwell time was limited to 3h and the plasma sodium increased to 124mEq/l.

This case emphasises the importance of the influence of the peritoneum in sodium homeostasis in patients on peritoneal dialysis, going beyond the concept of sodium sieving in order to evaluate ultrafiltration failure. Free water transport through aquaporins (AQP1) induced by the osmolality of the peritoneal fluid is assessed by sodium sieving, which is considered to be reduced if a decrease of less than 5mEq/l of sodium in the peritoneal fluid occurs (glucose concentration at 3.86%) after 60min of infusion.7 In our patient, who maintained normal sodium sieving, the comparison between the theoretical sodium concentration in the dialysate effluent and the actual measurement at the end of the dwell time pointed towards the aetiology of persistent hyponatraemia. Although some concepts have still not been elucidated with respect to peritoneum physiology, adjusting peritoneal dialysis regimens to the serum sodium concentration may prevent the perpetuation of hydroelectrolytic imbalance. The theoretical/real sodium ratio with respect to dialysate volume upon completion of the exchange dwell time could be used to quickly evaluate the tendency of the sodium balance in peritoneal dialysis.