[Hypophosphatemia in parenteral nutrition: prevention and associated risks factors]

To determine the incidence of hypophosphatemia in parenterally fed patients, the phosphate amount necessary to prevent this complication and associated risks factors.

Observational study, not controlled, in a third level hospital.

In-patients with parenteral nutrition with at least a complete laboratory work-up.

For a complete year, days on parenteral nutrition, administered phosphate and plasmatic ionised calcium levels, y-glutamiltranspeptidase, glucose, phosphate, pre-albumin, urea, and leukocytes were recorded. A multiple stepwise regression analysis and logistic regression are used for data analysis.

Eight hundred and twenty seven determinations, corresponding to 401 patients, were included. Significant variables (p < 0.05) were: administered phosphate and ionised calcium serum levels, glucose, pre-albumin, and urea; regression coefficients were 0.004 (95%CI: 0.002 to 0.006), -0.156 (95%CI: -0.270 to 0.037), -0.014 (95%IC: -0.022 to 0.009), 0.005 (95%CI: 0.002 to 0.009) and 0.019 (95%CI: 0.016 to 0.022), respectively; the constant was 1.0735 (95%CI: 0.939 to 1.2079). The risk for developing hypophosphatemia decreased from 0.65 (95%CI: 0.33 to 1.26) to 0.16 (95%CI: 0.078 to 0.35) when administered phosphate varied from the span 7.5-17.5 mmol to values higher than 27.5 mmol.

It is necessary to routinely supplement nutrition with phosphate since its content in commercially available lipidic emulsions is not sufficient to prevent hypophosphatemia in the majority of patients with parenteral nutrition. Phosphate intake must be sufficient to restore the intracellular phosphate deficit and to compensate for the plasmatic phosphate fall, with special attention to poorly nourished, hyperglycaemic or with renal failure patients. Phosphate intakes around 27-37 mmol dramatically decrease the incidence of hypophosphatemia in studied patients, with no recorded cases of severe hypophosphatemia.