Does hydrocortisone reduce death and BPD in preterm infants who remain intubated beyond a week of life?

February 21, 2020


Onland, W et al. Effect of Hydrocortisone Therapy Initiated 7 to 14 Days After Birth on Mortality or Bronchopulmonary Dysplasia Among Very Preterm Infants Receiving Mechanical Ventilation A Randomized Clinical Trial. Jama 2019; 321(4):354-363. PMID: 30694322


Brittany Butler, MD
Section of Neonatology, Department of Pediatrics
Children’s Hospital Colorado and University of Colorado School of Medicine

Clyde J. Wright, MD
Associate Professor
Section of Neonatology, Department of Pediatrics
Children’s Hospital Colorado and University of Colorado School of Medicine




(P) In infants born at gestational age less than 30 weeks and birthweight less than 1250 g who remain ventilator dependent at 7 to 14 days of life, (I) does hydrocortisone (C) compared to placebo (O) reduce the outcome of bronchopulmonary dysplasia or death at (T) 36 weeks postmenstrual age?


  • Design: Double-blind, placebo-controlled randomized clinical trial
  • Allocation: Patients were randomized to hydrocortisone or placebo stratified by study center and gestational age with a computer-generated randomization sequence
  • Blinding: Double-blind
  • Follow-Up: Until hospital discharge or death
  • Setting: 19 NICUs in the Netherlands and Belgium
  • Patients: 372 infants randomly assigned to a treatment group
    • Included:
      • GA <30 weeks
      • BW <1250 g
      • Remained mechanically ventilated with a high risk of BPD determined by a respiratory index, defined as mean airway pressure × fraction of inspired oxygen, >2.5 for 12 hr/day during a period of 48 hrs
    • Excluded:
      • chromosomal defects or major
      • congenital malformations
      • received corticosteroids for improving lung function in the first week of life
  • Intervention:
    • Infants were treated with hydrocortisone 5mg/kg per day divided in 4 doses for 7 days, followed by 3.75 mg/kg per day divided in 3 doses for 5 days, then 2.5 mg/kg divided in 2 doses for 5 days, and finally 1.25 mg/kg given once a day for 5 days. This resulted in 22 days of treatment. The total cumulative dose was 72.5 mg/kg.
    • The use of hydrocortisone outside of the study protocol was strongly discouraged but could be considered when the respiratory index was >10 for 6 hrs or when patients had received 10 days of study treatment but did not show improvement. When open-label hydrocortisone was started, the study treatment was stopped.
  • Outcomes:
    • The primary outcome was death or BPD at 36 weeks postmenstrual age, defined as requiring positive pressure, greater than 30% FiO2, or requiring 22-29% FiO2 and failing an oxygen reduction test.
    • Secondary outcomes included
      • Individual components of the primary outcome: mortality at 28 days and at hospital discharge, BPD at 28 days
      • Hospital length of stay
      • Use of open-label hydrocortisone
      • Respiratory outcomes including extubation failure at 3, 7, 14, and 21 days after starting treatment, total duration of mechanical ventilation and supplemental oxygen
      • Morbidities of prematurity including necrotizing enterocolitis, gastrointestinal bleeding, spontaneous intestinal perforation, intraventricular hemorrhage, periventricular leukomalacia, retinopathy of prematurity, hypertension, hyperglycemia, nosocomial infection (ie, sepsis, pneumonia, meningitis), patent ductus arteriosus
      • Growth as assessed by body weight and head circumference at 36 weeks’ postmenstrual age
  • Analysis and Sample Size:
    • Investigators aimed to recruit 200 patients in each group to give the study 80% power to detect an absolute risk reduction of 15% and a two-tailed type 1 error of 0.05 with an anticipated dropout rate of 10%
    • Main results were analyzed according to the intention-to-treat principle despite any protocol deviations.
    • Preplanned per-protocol and as-treated analysis were also performed.
    • Subgroup analysis according to gestational age, chorioamnionitis, respiratory index at randomization, sex, multiple birth, and study center steroid preference were performed using logistic regression models.
  • Patient Follow-Up:
    • 7065 infants assessed for eligibility.
    • 6693 excluded
      • 6519 did not meet inclusion criteria
        • 5911 did not require mechanical ventilation with a respiratory index of at least 2.5 at 7-14 days of life
        • 608 died in the first week of life
        • 162 parents declined
        • 12 excluded for other reasons
      • 372 infants randomly assigned to a treatment group
        • 182 to hydrocortisone group (1 died before 1st dose)
        • 190 to placebo group (2 did not receive placebo, 1 died before 1st dose, 1 physician chose not to treat)
      • 371 infants were included in the analysis (1 excluded because consent was withdrawn)


Randomized infants had similar baseline characteristics overall, as shown in table 1. However, there were more SGA infants and lower birthweights in the placebo group. Age, duration of invasive ventilation, and FiO2 requirement at randomization were similar in both groups.

Table 1
Select infant characteristics Hydrocortisone (n=182) Placebo (n=190)
Gestational Age, median (IQR), wk 25.4 (24.9-26.4) 25.6 (24.7-26.4)
Birthweight, median (IQR), g 775 (644-865) 710 (629-810)
SGA, NO (%) 26 (14.3) 38 (20.0)
Duration of invasive respiratory support before randomization, median (IQR), days 9 (7-11) 9 (7-11)
Age at randomization, median (IQR), days 10 (8-13) 0.35 11 (9-13)
Average FiO2 at randomization, median (IQR) 0.35 (0.30-0.45) 0.34 (0.29-0.40)

There was no difference in the primary composite outcome of death or BPD in the hydrocortisone or placebo groups, as shown in table 2. This remained true after adjustment for stratification risk factors and in prespecified sensitivity and subgroup analysis.

Table 2 demonstrates that multiple secondary outcomes favored the hydrocortisone group. The only statistically significant adverse outcome in the hydrocortisone group was the use of insulin for hyperglycemia.

Table 2
Select Outcomes Hydrocortisone (n=182) NO/Total (%) Placebo (n=190) NO/Total (%) P value
Death or BPD at 36 wks GA 128/181 (70.7) 140/190 (73.7) 0.54
Death at 36 wks GA 28/181 (15.5) 45/190 (23.7) 0.048
Death before hospital discharge 36/181 (19.9) 54/190 (28.4) 0.06
Failure to extubate after starting study medication
3 days 146/173 (84.4) 170/183 (92.9) .01
7 days 92/169 (54.4) 139/178 (78.1) <.001
14 days 56/166 (33.7) 86/168 (51.2) .001
Use of open label medication 51/181 (28.2) 108/190 (56.8) <.001
Pneumonia 45/181 (24.9) 64/190 (33.7) .048
Weight at 36 wks GA, mean (SD), g 2235 (395) 2125 (468)
Hyperglycemia requiring insulin 33/181 (18.2) 15/190 (7.9) .004


The authors conclude that among mechanically ventilated very preterm infants, administration of hydrocortisone between 7 and 14 days after birth, compared with placebo, did not improve the composite outcome of death or BPD at 36 weeks’ postmenstrual age. These findings do not support the use of hydrocortisone for this indication.


Bronchopulmonary dysplasia affects a majority of extremely low birthweight infants, leading to poor pulmonary and neurologic outcomes. Dexamethasone, a potent anti-inflammatory steroid, facilitates extubation and reduces the risk of BPD or death (1). However, it can cause hyperglycemia, hypertension, and may lead to abnormal neurologic development (1). Low dose prophylactic hydrocortisone also improves these outcomes (2), but this approach exposes some babies at low risk of death or BPD to potential side effects. Seeking a safer alternative, this is the first controlled trial to evaluate hydrocortisone in preterm babies that remain intubated beyond a week of life.

While this study failed to demonstrate a significant decrease in the composite primary outcome of death or BPD, there was a trend toward reduced death (16% vs 24%, p value 0.048, at 36 weeks; 20% vs 28%, p value 0.06, at hospital discharge) in infants treated with hydrocortisone. When using a composite outcome, an outcome that exhibits no effect can negate the effect of treatment on another outcome (3). Here, BPD occurred in 65% of all survivors regardless of treatment (100/(181-28) hydrocortisone; 95/(190-45) placebo). Importantly, the rate of severe BPD was not different between groups, suggesting that hydrocortisone did not lead to more survivors with severe morbidity. In contrast, treated patients were more likely to be extubated after starting treatment, less likely to be diagnosed with pneumonia, and gained more weight.

Currently, use of steroids in preterm babies requiring mechanical ventilation beyond the first week of life varies. In an American survey, only 13% of responding neonatologists would start steroids in a 21 day old requiring moderate ventilatory support (4). In contrast, 90% of responding UK providers would treat intubated babies with steroids after the first week (5). Similarly, participants in this study started hydrocortisone around 10 days of life and required a median 0.35 FiO2. While the inclusion of infants requiring moderate FiO2 relatively early in their course could attenuate hydrocortisone’s effect, the high rate of the primary outcome in all study participants argues against this possibility. Additionally, high rates of off label steroid use in the placebo could decrease hydrocortisone’s effect in the treatment group, but preplanned per-protocol and as-treated analyses also revealed no difference in the primary outcome.

Many neonatologists treat preterm babies who remain intubated after the first week of life with steroids (4,5). Although this study did not show a reduction in the combined outcome of death or BPD, it suggests that hydrocortisone may aid in successful extubation and reduce the risk of death. Given the morbidity associated with prolonged mechanical ventilation including air leak, ventilator associated pneumonia, and worse neurodevelopment (6), early extubation may improve longer term outcomes not yet reported for this study. Furthermore, those with severe BPD represent a diverse group with heterogenous outcomes (7); whether hydrocortisone has a meaningful treatment effect in any of these subgroups is unknown. Finally, we await the results of the NICHD Neonatal Research Network study: Hydrocortisone for BPD, which will also assess survival without BPD and neurodevelopmental outcomes at 18-22 months.


  1. Doyle LW, Cheong JL, Ehrenkranz RA, Halliday HL. Late (<7 days) systemic postnatal corticosteroids for prevention of bronchopulmonary dysplasia in preterm infants (Review). Cochrane Database Syst Rev.2017; 10. Art. No.: CD001145
  2. Baud O, Maury L, Lebail F, Ramful D, El Moussawi F, Nicaise C, et al. Effect of early low-dose hydrocortisone on survival without bronchopulmonary dysplasia in extremely preterm infants (PREMILOC): a double-blind, placebo-controlled, multicentre, randomised trial. Lancet. 2016; 387(10030):1827-36
  3. Freemantle N, Calvert M, Wood J, Eastaugh J, Griffin C. Compositeoutcomes in randomized trials: greater precision but with greater uncertainty?  2003; 289(19): 2554-2559.
  4. Niwas R, Kamat M, Kling P, Yeh TF, Pyati S. Postnatal corticosteroids to treat or prevent chronic lung disease: a survey of neonatologists’ practices. Arch Dis Child Fetal Neonatal Ed. 2009; 94(1): F77.
  5. Job S, Clarke P. Current UK practices in steroid treatment of chronic lung disease. Arch Dis Child Fetal Neonatal Ed. 2015; 100(4): F371.
  6. Vliegenthart RJ, Kaam AH, Aarnoudse-Moens CS, Van Wassenaer AG, Onland W. Duration of mechanical ventilation and neurodevelopment in preterm infants. Arch Dis Child Fetal Neonatal Ed. 2019; 104: F631–F635.
  7. Akangire G, Manimtim W, Nyp M, Noel-MacDonnell J, Kays A, Truog W, Taylor J. Clinical Outcomes among Diagnostic Subgroups of Infants with Severe Bronchopulmonary Dysplasia through 2 Years of Age. Am J Perinatol. 2018; 35(14):1376-1387