Is parenteral phosphate replacement in the intensive care unit safe?

Hypophosphatemia is well recognized in the intensive care setting, associated with refeeding and continuous forms of renal replacement therapy (CCRT). However, it is unclear as to when and how to administer intravenous phosphate supplementation in the general intensive care setting. There have been recent concerns regarding phosphate administration and development of acute kidney injury. We therefore audited our practice of parenteral phosphate administration. We prospectively audited parenteral phosphate administration (20mmol) in 58 adult patients in a general intensive care unit in a University tertiary referral center. Fifty-eight patients were audited; mean age 57.2 +/- 2.0 years, 70.7% male. The median duration of the infusion was 310min (228-417), and 50% of the patients were on CRRT. 63.8% of patients were hypophosphatemic (<0.87mmol/L) prior to the phosphate infusion, and serum phosphate increased from 0.79 +/- 0.02 to 1.07 +/- 0.03mmol/L, P<0.001. Two patients became hyperphosphatemic (>1.45mmol/L). There was no correlation between the change in serum phosphate and the pre-infusion phosphate. Although there were no significant changes in serum urea, creatinine or other electrolytes, arterial ionized calcium fell from 1.15 +/- 0.01 to 1.13 +/- 0.01mmol/L, P<0.01. Although infusion of 20mmol phosphate did not appear to adversely affect renal function and corrected hypophosphatemia in 67.7% of cases, we found that around 33% of patients who were given parenteral phosphate were not hypophosphatemic, and that the fall in ionized calcium raises the possibility of the formation of calcium-phosphate complexes and potential for soft tissue calcium deposition.