Abstract
Objective: To characterize common dosing strategies and to investigate the association between hydrocortisone dosage and in-hospital mortality in infants born extremely premature. Study design: We performed a retrospective review of a cohort of infants born ≤30 weeks' gestational age from 2010 to 2016 from the Pediatrix Clinical Data Warehouse who received hydrocortisone in the first 14 postnatal days. Infants were divided by initial hydrocortisone dosage (high: >2 mg/kg/d vs low: ≤2 mg/kg/d). Baseline characteristics and medication coexposures were compared and mortality was evaluated in a multivariable analysis. Results: A total of 1427 infants were included, 733 with high dosage (51%) and 694 with low dosage (49%). The groups were similar with regard to baseline characteristics. Infants in the high-dosage group had significantly more exposure to any vasopressors (89% vs 84%, P < .001) and greater mortality (50% vs 23%, P < .001) vs the low-dosage group. High dosage of hydrocortisone was associated independently with death (aOR 3.27, 95% CI 2.47-4.34, P < .001) in a multivariable regression analysis including propensity scoring for dosage and other covariates. When the cohort was split into quartiles by dosage, mortality was lower in the lower-dosage quartiles compared with the higher quartiles (mortality range 13%-50%). Conclusions: In this retrospective analysis of a large sample of infants born premature, increased initial hydrocortisone dosage was associated independently with increased mortality. Trials to assess the impact of hydrocortisone dosage in this population are needed. © 2018 Elsevier Inc.