In Reply: Dr Fan points out that our definition of corticosteroid insufficiency does not conform to that of the ACCM,1 which states that a corticotropin test should not be used to identify patients with septic shock or acute respiratory distress syndrome who should receive glucocorticoids. In the HYPOLYTE study, patients with adapted corticosteroid function had a short median time of exposure to hydrocortisone (median, 34 hours; interquartile range, 20-49). We reported a decreased hospital-acquired pneumonia rate at day 28 in all patients, but the study was neither designed nor powered to evaluate the safety of treatment in patients with adapted corticosteroid function. Regarding the threshold of cortisolemia, the current definition of critical illness–related corticosteroid insufficiency was not available at the time the study was started. When using this definition,1 the risk of hospital-acquired pneumonia at day 28 remained lower in the hydrocortisone group (hazard ratio, 0.41; 95% CI, 0.20-0.88; P = .02). We thus believe that a corticotropin test should be performed to identify trauma patients who may benefit from prolonged low-dose hydrocortisone. In a systematic review of corticosteroid use in patients with severe sepsis,2 a long course of low-dose corticosteroids did not alter superinfection rate (relative risk, 1.01; 95% CI, 0.82-1.25; P = .92). Also, the rate of polyneuromyopathy, which is low in trauma patients, was not increased in septic patients treated with corticosteroids.3