Severe Hypophosphatemia: Its Prevalence and Predictors Associated with In-Hospital Mortality

Background: It has been unclear about prevalence and in-hospital mortality of severe hypophosphatemia among hospitalized medical patients. Predictors for in-hospital mortality have not yet been determined in terms those may include serum nadir phosphate level, hemolytic complication from hypophosphatemia, or baseline disease as a cause of hypophosphatemia .

Methods: A retrospective cohort study was conducted for all adult patients with severe hypophosphatemia (<1.5 mg/dL) at a Japanese community teaching hospital for 6-year period from January 2009 to December 2014. Prevalence of severe hypophosphatemia was determined and in-hospital mortality was compared between patients with severe hypophosphatemia and all hospitalized adult patients during the same study period. Predictors for in-hospital mortality were analyzed among patients with severe hypophosphatemia based on multivariable logistic regression model adjusted for demographics, causes of hypophosphatemia, nadir phosphate level, and presence of hemolysis.

Results: From a total of 482,812 admitted adult patients, 97 patients (0.02%) had severe hypophosphatemia. The common causes of severe hypophosphatemia were malnutrition, infection, or malignancy. Hemolysis was identified in patients with severe hypophosphatemia in 5 patients (5.15% of those with severe hypophosphatemia) and no one with hemolysis died during the hospitalization. In-hospital mortality among those with severe hypophosphatemia was 21.6% and it was significantly greater than that (0.26%) of all hospitalized adult patients (P<0.001). Concomitant malignancy was associated with increased in-hospital mortality with odds ratio of 10.98 (95% CI, 1.56-77.1). Mortality was not related to lower nadir level of serum phosphate or presence of hemolysis.

Conclusions: Although severe hypophosphatemia was uncommon in hospitalized patients, in-hospital mortality among patients with severe hypophosphatemia was high (about 20%). However, this in-hospital mortality was associated with baseline malignant disease but not with the magnitude of serum nadir phosphate level or associated hemolysis, which was noted only rarely in those with severe hypophosphatemia.