Relationship between serum phosphate and mortality in critically ill children receiving continuous renal replacement therapy

We aimed to explore the relationship between serum phosphate concentration and 90-day mortality in critically ill children receiving continuous renal replacement therapy (CRRT).Data from the medical records of children aged <13 years who received CRRT at the Pediatric Intensive Care Unit of Hunan Children's Hospital, China from January 2015 to June 2020 were retrospectively collected. Children were grouped into four categories according to the baseline phosphate concentration before CRRT and mean serum phosphate concentration during CRRT: <0.81 mmol/L (hypophosphatemia), 0.81-1.19 mmol/L, 1.2-2.4 mmol/L (normal phosphate concentration), and >2.4 mmol/L (hyperphosphatemia), with the normal phosphate group serving as the comparator group. The correlation of the serum phosphate concentration before and during CRRT with the 90-day mortality after CRRT initiation was analyzed using logistic regression.A total of 177 children were included in our study. The mean serum phosphate concentration before CRRT was 1.46 mmol/L (quartiles: 1.04, 2.20). The 90-day mortality rate was increased in children with a serum phosphate concentration >2.4 mmol/L before CRRT (adjusted odds ratio [aOR] 3.74, 95% confidence interval [CI] 1.42-9.86, P = 0.008). The mean serum phosphate concentration during CRRT was 1.2 mmol/L (quartiles: 0.91, 1.49). The 90-day mortality rate was increased in children with a mean serum phosphate concentration >2.4 mmol/L during CRRT (aOR 7.34, 95% CI 1.59-33.88, P = 0.011).Hyperphosphatemia before and during CRRT predicts a higher 90-day mortality rate.