Lower mean phosphate independently predicts mortality in critically ill patients: Results from a prospective cohort study.

PURPOSE: To evaluate lower mean phosphate as a prognostic tool in critically ill patients. METHODS: This is a prospective single-center cohort study including adult patients (> 18 years) with a length of intensive care unit (ICU) stay of at least 24 h. Phosphatemia was evaluated within 1 h of ICU admission and once daily. Mean phosphate, calculated by the simple arithmetic mean of daily phosphate measurements, was proposed and tested. Standard severity scores were applied. Multivariate and survival analyses were performed. RESULTS: A total of 317 patients were included, of whom 111 (35%) presented hypophosphatemia. Hypophosphatemia associated with surgical conditions, nutritional therapy, hypovitaminosis D, hyperparathyroidism, mechanical ventilation (need and duration), and ICU and hospital length of stay were evaluated. Admission APACHE II and SOFA (ICU days 1, 3, and 7) scores and ICU and in-hospital mortality were greater in the hypophosphatemia group than control group. Higher APACHE II (RR: 1.1; 95%CI: 1.01-1.2; p = 0.045) and lower mean phosphate (RR: 0.02; 95%CI: 0.001-0.09; p = 0.044) independently predicted ICU and in-hospital mortality. CONCLUSIONS: Hypophosphatemia is frequent in the ICU, and was associated with unfavorable outcomes. This study introduces the importance of longitudinal monitoring of phosphate levels, since lower mean phosphate is an independent predictor of mortality in critically ill patients.