To investigate the morbidity of continuous renal replacement therapy(CRRT)-related hypophosphatemia in critically ill patients, risk factors for CRRT-related hypophosphatemia and clinical benefits of phosphorus supplementation.

A total of 451 critically ill patients who received CRRT in the First Affiliated Hospital of Xi 'an Jiaotong University from January 1, 2017 to December 31, 2018 were recruited. According to whether serum phosphorus level less than 0.81 mmol/L, the patients were divided into hypophosphatemia group (n=297) and non-hypophosphatemia group (n=154) . Basic clinical information of patients were collected, including age, gender, cormorbidities, major diagnosis, clinical scoring, indication for CRRT, electrolytes before CRRT, length of hospital stay, prognosis, and possible risk factors for hypophosphatemia were analyzed. Patients with hypophosphatemia were further subdivided into subgroups according to whether or not receiving phosphorus supplementation, and were analyzed on length of hospital stay and prognosis.

The incidence of CRRT-related hypophosphatemia in critically ill patients was 65.85%. Female (OR=2.484, 95%CI: 1.413-4.367), patients with lower blood phosphorus levels before CRRT [1.31 (0.93, 1.82) mmol/L vs 1.68 (1.35, 2.20) mmol/L, OR=0.469, 95%CI: 0.326-0.673] and patients with longer duration on CRRT [66.3 (39.3, 131.8) h vs 24.0 (16.0, 34.0) h, OR=1.054, 95%CI: 1.038-1.070] may be more likely to develop hypophosphatemia. Female patients (P=0.014), complicated with urologic diseases (P=0.012), primary diagnosis as urologic disease (P=0.013), lower blood phosphorus level before CRRT [mild 1.50 (1.10, 2.04) mmol/L vs severe 1.25 (0.83, 1.73) mmol/L, P=0.019], longer duration of CRRT [mild 45.38 (29.50, 71.10) days vs severe 116.50 (57.70, 194.80) days, P<0.001], and length of ICU stay[mild 0 (0, 5) days vs severe 11 (5, 21) days, P<0.001] may be associated with more severe hypophosphatemia. In terms of benefits from phosphorus supplementation, although there was no statistically difference in hospital mortality, mortality at 30 days after discharge, and mortality at 90 days after discharge between these two groups, blood phosphorus level at the end of CRRT in phosphorus supplementation group did increase [(0.11±0.50) mmol/L] compared with average blood phosphorus level during CRRT.

The incidence of CRRT -related hypophosphatemia is high in critically ill patients. Gender and total duration of CRRT can be characteristics combined to predict the occurrence of CRRT-related hypophosphatemia. For patients developing hypophosphatemia during CRRT, phosphorus can be supplemented appropriately.