Serum estradiol level predicts acute kidney injury in medical intensive care unit patients.

Abstract

Previous studies have shown that serum estradiol (E2) levels can predict mortality in intensive care unit patients. Our study investigated the predictive role of admission estradiol level on patient mortality and development of acute kidney injury in medical intensive care unit patients with a wide range of diagnoses. We conducted a prospective cohort study using serum samples from hospitalized patients in medical, cardiac, and pulmonary intensive care units at the Ege University Hospital within 6 months. Serum estradiol levels from 118 adult patients were collected within 48h of hospitalization. Receiver operating curves and multiple logistic regression analyses were performed to investigate its relationship with acute kidney injury development and mortality. Serum estradiol levels were significantly higher in non-survivor patients than in survivor patients [85 (19-560) pg/mL vs. 32 (3-262) pg/mL, p < 0.001]. Admission estradiol levels were significantly higher in patients with AKI on admission than in patients with chronic kidney disease (p = 0.002) and normal renal function (p = 0.017). Serum E2 levels were higher in patients with renal deterioration during follow-up than patients with stable renal functions [62 (11-560) pg/mL vs. 38 (3-456) pg/mL, p = 0.004]. An admission estradiol level of 52.5pg/mL predicted follow-up renal deterioration with 63% sensitivity and 74% specificity. A combined (APACHE II-E) score using APACHE II and serum estradiol level predicted overall mortality with 66% sensitivity and 82% specificity. Admission estradiol level is a good marker to predict the development of acute kidney injury and mortality in medical intensive care unit patients.