High prevalence of hypophosphataemia amongst patients with infectious diseases. A retrospective study.

Abstract

To describe the prevalence of hypophosphataemia amongst all patients treated during 1993 for infectious diseases. The associations between serum C-reactive protein, serum albumin and white blood cells in blood were studied amongst hypophosphataemic patients. Comparisons were made according to the severity of infection. A retrospective study of the prevalence of hypophosphataemia, and a case-control study amongst a subgroup of patients. The Department of Infectious Diseases, University Hospital, Umeå, Sweden. For the prevalence study all 967 patients (449 women and 518 men) treated during 1993 were included. In the case-control study, 108 cases, with serum phosphate < or =0.64 mmol L(-1), 216 age-and sex-matched controls with serum phosphate >0.82 mmol L(-1) (men) and >0.86 mmol L(-1) (women), respectively, were included. In 402 of 967 patients (42%) at least one serum phosphate value was below the reference level. Hypophosphataemia was found in 573 of 1966 blood samples (29%). Severe hypophosphataemia (<0.30 mmol L(-1)) was seen in 1.2% of the patients (0.9% of blood samples). The prevalence of hypophosphataemia was higher amongst women than amongst men. In the case-control study, the serum C-reactive protein and the number of white blood cells was higher amongst the cases compared with the controls (124 vs. 94 mg L(-1) and 11.4 vs. 9.3 cells/L x10(9), respectively). The multiple logistic regression showed a 4-fold higher risk of having low serum phosphate in patients with severe infection, compared with mild infections. None of the other variables (albumin, days of hospital stay or white blood cells in blood) increased the risk for hypophosphataemia. The high prevalence of low serum phosphate levels and the increased risk of having low levels in severe infections shown in this study needs further attention.