Hypophosphatemia in the intensive care unit: Incidence, predictors and management

Abstract

Introduction: Our objectives are to identify the incidence of hypophosphatemia and the associated risk factors. We also want to establish intravenous replacement therapy that is effective for ICU patients. Methods: A prospective observational study assessing adults admitted to ICU in between March and May 2009. All patients without baseline phosphate level and renal failure were excluded. They were evaluated for the occurrence of common risk factors. Association with independent variables that includes age, gender and BMI were verified. Evaluation of IV replacement therapy was done in the treated patients. Results: From 50 patients that were reviewed, nine were excluded. There were 66% male and 34% female with mean age 46.88±17.89. The mean ICU stay was 8.00±6.41 days. The incidence of hypophosphatemia was 29% (n=12/41). Gender and creatinine clearance was found to be significantly different between normophosphatemia and hypophosphatemia patients. There was no significant association for each potential risk factor and the number of risk factors (≥3) with the incidence of hypophosphatemia. Multi-linear regression analysis showed that lactate, creatinine clearance and pH were significant predictors to the serum levels. A significant difference of mean serum phosphate was seen after repletion by total dose of 10, 20 and 40 mmols in the treatment subgroups. Conclusions: The incidence of hypophosphatemia in our ICU was high and comparable to previous studies. None of the commonly reported risk factors is associated with hypophosphatemia in this studied population. Among all significant correlated variables, only pH was found to be a significant predictor for serum phosphate. Baseline phosphate level may guide the initial replacement dose to prevent delay in normalization of serum levels.