Of the 3,757 titles and abstracts screened, 20 studies were included: 7 RCTs and 13 observational studies (
PRISMA diagram of the literature search.
Characteristics of the randomized controlled trials included in the meta-analysis.
| Author, Year, Country | Intervention (number of participants) | Comparator (number of participants) | GA (weeks) |
BW (g) |
Time of initiation of hydrocortisone | Dose and duration of hydrocortisone | Other comments |
|---|---|---|---|---|---|---|---|
| Hochwald, 2013, Canada | 11 | 11 | I: 26.1 ± 1.5 |
I: 870 |
Early (HC started along with the administration of the first inotrope) | Initial dose of 2 mg/kg, 6 h after 1 mg/kg q 6 h for three doses, followed by 0.5 mg/kg q 6 h for four doses |
≤30-week-old infants with ≤ 1,250 g BW within 48 h of postnatal age were enrolled Definition of hypotension: MAP < GA IBP monitoring was done Both groups received two doses of fluid boluses (10 ml/kg) prior to randomization The initial dose of dopamine used was 5 μg/kg/min and then titrated. If needed, an epinephrine infusion at 0.05 μg/kg/min was added and titrated |
| Kovacs, 2019, USA | 16 | 19 | I: 39 (38–40) |
I: 3,625 |
Early (HC started along with the administration of the first inotrope) | 0.5 mg/kg/d q 6 h until the re-warming period in neonates undergoing therapeutic hypothermia |
Infants were enrolled at 36 weeks who were undergoing therapeutic hypothermia for HIE Definition of hypotension: MAP < GA IBP monitoring was done Both groups received two doses of fluid boluses (10 ml/kg over 15 min) prior to randomization The dose of dopamine used was 6–20 μg/kg/min in both groups. Dobutamine was added if there was myocardial dysfunction. Inotropes were tapered at the clinicians’ discretion Baseline cortisol levels were low in both groups and the values decreased in the placebo group The HC group had a statistically significant decreased risk of electrographic seizures, but the requirement for anti-epileptic drug therapy and abnormal MRI were similar in both groups The cumulative and maximum dose of inotrope therapy was significantly lower in the HC group The regression modeling predicted HC to be associated with a 2.2 mm Hg increase in pulse pressure |
| Ng, 2006, Hong Kong | 24 | 24 | I: 27.2 |
I: 918 |
Early (HC was initiated if the dopamine requirement was ≥10 μg/kg/min) | 1 mg/kg q 8 h for 5 days |
<32 weeks and <1,500 g infants within 7 days of life were enrolled Definition of hypotension: MAP < GA IBP monitoring was done Both groups received fluid boluses (30 ml/kg) prior to randomization HC was initiated if the dopamine requirement was ≥10 μg/kg/min. When 20 μg/kg/min dopamine and dobutamine failed to maintain MAP, an epinephrine infusion at a rate of 0.2 μg/kg/min was started. The dose was increased at a rate of 0.2 μg/kg/min q 30 min until the BP normalized Serum cortisol level prior to randomization was comparable between the groups Glycosuria was significantly higher in the HC group |
| Salas, 2014, Argentina | 25 | 25 | I: 38 ± 1.4 |
I: 2,936 ± 680 |
Late (HC was started at dopamine ≥14 μg/kg/min and/or after starting epinephrine) | 2.5 mg/kg q 12 h for 48 h |
≥37-week-old infants of any postnatal age were included Definition of hypotension: MAP ≤ 5th centile Both IBP and NIBP monitoring was done Both groups received 10–20 ml/kg of volume expansion Dopamine was started at 5–10 μg/kg/min and gradually increased. Dobutamine was used in neonates with poor cardiac contractility. If hypotension was persistent, epinephrine at 0.1 μg/kg/min was started This was an open trial and six out of the 25 patients in placebo group received HC Hyperglycemia was significantly higher in those infants who received epinephrine and HC (50%) Though not statistically significant, two neonates in the HC group developed hypertension compared to none in the placebo group |
GA, gestational age; BW, birth weight; I, intervention group; C, control group; SD, standard deviation; q, every; HC, hydrocortisone; IVH, intraventricular hemorrhage; BPD, bronchopulmonary dysplasia; BP, blood pressure; MAP, mean arterial pressure; IBP, invasive BP; NIBP, non-invasive BP; VLBW, very low birth weight. HIE, hypoxic-ischemic encephalopathy; NI, no information.
Characteristics of the observational studies included in the meta-analysis
| Author, Year, Country, Study design | Intervention (number of participants) | Comparator (number of participants) | GA (w) Mean ± |
BW (g) |
Time of initiation of hydrocortisone | Dose and duration of hydrocortisone | Other comments |
|---|---|---|---|---|---|---|---|
| Altit, 2018, Canada, retrospective study | 39 | 23 | I: 26 ± 1.9 |
I: 828 ± 339 |
Late (mean VAI score in HC group was 15) | There was no uniform protocol for HC initiation. The cumulative dose of HC received was 10.5 mg/kg |
Preterm neonates of ≤34 weeks with culture-proven sepsis and/or NEC were included Definition of hypotension was not reported Both IBP and NIBP were used Both groups received volume expansion prior to HC initiation Dose of inotropes and the titration were clinician-dependent The HC group was sicker compared to the no HC group VAI decreased and BP, urine output, and oxygen requirement improved significantly following HC. Despite the HC group being sicker at baseline, those who received HC had similar survival from septic episodes. However, after adjustment for baseline sickness, a multivariate analysis revealed that receipt of HC and GA were associated with a higher risk of death and/or BPD at 36 weeks PMA After adjustment for baseline sickness, the HC group had a higher risk of mortality at 1 year PMA (HR: 6.08, |
| Mizobuchi, 2011, Japan, matched case–control | 12 | NA | I: 24.6 ± 1.3 |
I: 554 ± 100 |
Early (HC initiated along with the first inotrope) | A single HC dose of 2 mg/kg was used. If hypotension persisted, the dose was repeated q 12 h. The duration of HC treatment was not mentioned |
ELGANs within the first 7 days of life were enrolled Definition of hypotension: MAP <30 mm Hg for infants with GA between 25 and 27 weeks, and <25 mm Hg for infants with GA between 23 and 24 weeks IBP was used Volume expansion with 10 ml/kg was given before inotrope initiation. The type of fluid used was not mentioned The initial and maximum doses of dopamine were not mentioned. HC was given if hypotension persisted despite a dopamine dose of 5 μg/kg/min Infants <25 weeks GA were given prophylactic indomethacin (0.1 mg/kg of continuous infusion for 12 h) for prevention of both PDA and IVH and were also treated with prophylactic glucose-insulin therapy for prevention of hyperkalemia on the 1st day of life The MAP in the HC group increased significantly at 2 h after HC treatment and was comparable to the control group at 5 h and remained at normal levels until 12 h after HC treatment |
| Noori, 2006, USA, prospective (pre-post) | 15 preterm infants and 5 term infants | NA | Preterm: |
Preterm: |
Late (HC was started at ≥ 15 μg/kg/min of dopamine and/or dobutamine) | HC dose was 2 mg/kg followed by 1 mg/kg q 12 h up to four additional doses. A full course of HC was given only when the patient remained hypotensive or maintained only minimum acceptable BP with a dopamine dose of 5–8 μg/kg/min |
Both term and preterm infants of less than 72 h of age were included Definition of hypotension: BP values <10th percentile Both IBP and NIBP were used. BP recordings were reported at different time points after HC initiation from 1 to 48 h Hypotensive infants received one dose of 10–20 ml/kg of normal saline prior to inotrope initiation Dopamine dosage used was 5–20 μg/kg/min Echocardiography was used to monitor the effect of HC on cardiac function In both the preterm and term neonates, HC improved BP without compromising cardiac function, systemic perfusion, or cerebral and renal blood flow Hyperglycemia (defined as two consecutive blood sugar values of >200 mg/dl) was significantly higher in the HC group. The requirement for insulin for the treatment of hyperglycemia was comparable between the two groups The authors also evaluated the outcomes of low vs. high-dose HC therapy with the cut-off being 4 mg/kg/day. There were no significant differences in the outcomes except for cortisol levels and the proportion of infants with hypertension (BP> 90th centile) which was higher in the high-dose group Infants treated with HC who had pre-treatment cortisol levels >15 μg/dl had significantly higher rates of hyperglycemia, insulin therapy. and death than those with levels <15 μg/dl The authors concluded that the lowest effective HC dose should be used and that HC therapy should be avoided in infants who are not cortisol deficient |
| Peeples, 2017, USA, retrospective cohort | 70 | 36 | I: 25 ± 1 |
I: 687 ± 161 |
Late (HC was initiated in neonates requiring a mean vasopressor dose of 12.4 μg/kg/min) | HC dose was 1 mg/kg followed by 0.5–1 mg/kg q 8–12 h. There was variation in the cumulative dosage of HC due to clinician discretion from 1 to 4 mg/kg/d. Duration of HC therapy not mentioned |
Preterm infants <28 weeks who were of postnatal age up to 28 days were enrolled. Except for birth weight, all other major baseline characteristics were similar between the two groups Definition of hypotension: There was no specific BP cut-off for defining hypotension. It was based on the clinician's discretion based on abnormal perfusion suggested by decreased urine output, acidosis, or clinical evidence of organ dysfunction No information regarding how BP was monitored No information as to whether any saline boluses were given preceding inotrope initiation The HC group had a higher incidence of hypertension ( The increased incidence of death in the high cortisol group was independent of the HC dose used Vasopressor requirement decreased significantly in the cortisol-deficient group who were treated with HC over a period of 14 h |
| Seri, 2001, USA retrospective (pre-post) | 21 | NA | 26.9 ± |
952 ± |
Late [HC was initiated at 20 μg/kg/min of dopamine or when dopamine (10–15 μg/kg/min) with dobutamine and/or epinephrine failed] | HC dose and duration varied based on the etiology of hypotension. HC dose used was 1 mg/kg/dose q 12 h. Preterm infants with severe capillary leak syndrome and/or previous steroid treatment received 3–6 mg/kg/d divided twice daily or four times daily for 2–3 days for a total of six treatment courses |
Preterm infants of ≤14 weeks postnatal age were included Definition of hypotension not mentioned IBP was used. BP recordings were monitored until 24 h Hypotensive infants received volume resuscitation (dose not mentioned) Maximum dosage of dopamine used was 20 μg/kg/min or 10–15 μg/kg/min and dobutamine/epinephrine was concurrently used Response to treatment was evaluated at 24 h |
GA, gestational age; BW, birth weight; I, intervention group; C, control group; SD, standard deviation; CDH, congenital diaphragmatic hernia; HC, hydrocortisone; IVH, intraventricular hemorrhage; NEC, necrotizing enterocolitis; PDA, patent ductus arteriosus; BPD, bronchopulmonary dysplasia; BP, blood pressure; MAP, mean arterial pressure; IBP, invasive BP; NIBP, non-invasive BP; ELBW, extremely low birth weight; ELGANs, extremely low gestational age neonates; VAI, vasoactive-inotropic; HR, hazard ratio; PMA, post-menstrual age; NA, not applicable.
Narrative review of the included studies.
| Author, Year, Country, Study design | Intervention (number of participants) | Comparator (number of participants) | GA (weeks) |
BW (g) |
Time of initiation of hydrocortisone | Dose and duration of hydrocortisone | Other comments |
|---|---|---|---|---|---|---|---|
| Baker, 2008, USA, retrospective study | 117 | NI | 35 (26–39) | 2,620 (880–3,400) | Late |
Stress dose |
All infants with refractory hypotension were included The maintenance dose of HC was given only when the baseline cortisol level (12 h after stress dose) was <5 μg/dl and/or hemodynamically unstable Definition of hypotension: MAP < GA IBP and/or NIBP were used Treatment with HC increased the MAP at 2, 6, 12, and 24 h after initiation; decreased the total inotrope dose at 6, 12, and 24 h; and was associated with the resolution of oliguria No difference was observed for the outcomes of IVH > grade 2, PVL, bacterial or fungal sepsis, and spontaneous intestinal perforation after HC treatment |
| Batton, 2012, USA, RCT | Pilot study enrolling 10 infants |
NA | 230/7–266/7 | NA | Early [enrolled infants were administered study infusion (dopamine or placebo) and study syringe medication (HC or placebo)] | Initial dose of 1 mg/kg followed by six doses of 0.5 mg/kg q 12 h |
Only preterm infants <27 weeks within the first 24 h were included Definition of hypotension: protocol-driven low BP from various centers Studied infants received two doses of crystalloid at Enrolled infants were administered study infusion (dopamine or placebo) and study syringe medication (hydrocortisone or placebo) 2 × 2 factorial randomization: (1) dopamine/placebo; (2) dopamine/hydrocortisone; (3) placebo/placebo; (4) placebo/hydrocortisone 6 μg/kg/min of dopamine (or placebo volume equivalent) was initiated and increased every 20 min to a maximum rate of 15 μg/kg/min Only 17% of eligible infants were enrolled and enrollment of the study subjects was poor None of the infants received open-label anti-hypotensive therapy in the first postnatal week Two infant deaths occurred in the placebo/placebo group, the time of death was distant from enrollment (days 20 and 62) |
| Bourchier, 1997, New Zealand, RCT | 21 | 19 | I: 26.6 ± 2.1 |
I: 923 ± 188 |
Early (HC was started before inotrope initiation) | An initial dose of HC (2.5 mg/kg q 4 h) was continued for 48 h, followed by 1.25 mg/kg for 48 h, then 0.625 mg/kg for 48 h |
VLBW infants of postnatal age less than 7 days were enrolled Definition of hypotension: MAP <25 mm Hg (birthweight 500–749 g), <30 mm Hg (BW 750–999 g), 35 mm Hg (BW 1,000–1,499 g) on two occasions 30 min apart The method of BP measurement (IBP or NIBP) was not mentioned Infants in both groups received a dose of colloid at 10 ml/kg prior to randomization In the I group (HC alone), persistent hypotension after the first dose of HC was managed by a dose of colloid. If hypotension continued or developed after the second dose of HC, this was judged as treatment failure and the infant was started on dopamine In the C group (inotrope alone), dopamine was started at 5 μg/kg/min, increasing stepwise to a maximum dose of 20 μg/kg/min if necessary. If hypotension persisted, noradrenaline infusion (0.05–0.5 μg/kg/min) was added. Once normotension was achieved for 24 h, the inotropes were reduced over a 24–48 h period There was no pre- or post-correlation baseline cortisol (after ACTH stimulation) or maximum dopamine in either of the two groups Slower weaning of HC (over 144 h) resulted in hypertension in four neonates There were four treatment failures in HC group who were successfully treated with dopamine There was no difference in the outcomes of survival, BPD, ROP (stage 2–4), IVH (grades 2–4), NEC, symptomatic PDA, insulin requirement, and sepsis |
| Heckman, 1999, Germany, retrospective (pre-post) | 30 | NA | 26.0 ± 1.7 | NA | Early (HC started prior to the initiation of the first inotrope) | HC dose used was 2–3 mg/kg/dose. Duration and frequency of dosing were not mentioned |
Only preterm infants <28 w within the first 72 h were included Definition of hypotension: MAP <25 mm Hg (birthweight 500–749 g), <30 mm Hg (BW 750–999 g), 35 mm Hg (BW 1,000–1,499 g) on two occasions 30 min apart IBP/NIBP were not mentioned Studied infants received two doses of crystalloid at 15 ml/kg before HC and/or dopamine initiation Dopamine was initiated at 5 μg/kg/min and the maximum dose was not mentioned Of the 30 hypotensive infants, only 16 required dopamine, and four out of the 16 infants were weaned off dopamine within the first 12 h of HC initiation |
| Helbock, 1993, USA, prospective study (pre-post) | 6 | NI | NI | 510–750 | Late (HC after the highest dosage of vasopressor therapy was required) | 24–60 mg/m2/day was used. |
Six ELBW neonates were evaluated Definition of hypotension: MAP < 32 mm Hg IBP was used HC was added once the MAP was 21 mm Hg despite the addition of 10–50 ml/kg of volume expansion (saline, 5% albumin, or packed blood red blood cells) and 24 to 60 μg/kg/min of cumulative inotrope therapy Four of the enrolled neonates had cortisol concentrations <5 μg/dl Increase in BP occurred as early as 30 min following the administration of HC, and within 12–30 h all the newborns were normotensive without inotrope support |
| Krediet, 1998, Holland, RCT | 13 | 13 | NI | NI | Late (HC was initiated if the neonate was hypotensive despite receiving dopamine and/or dobutamine at 20 μg/kg/min) | 5 mg/kg/d in four divided doses |
Only the abstract was available Definition of hypotension: MAP < GA No difference in additional requirement of inotropes or volume expanders was found In the HC group, the increase in MAP to normal values occurred more rapidly (within 6 h) No differences were found between groups for IVH |
| Rios, 2014, USA, retrospective cohort | 1,501 | 4,033 | I: 32 ± 6.2 C:33 ± 5.4 | I: 2,047 ± 1,232 |
NI | Dose of HC not mentioned |
Both preterm and term infants within the first 28 days of life were evaluated Definition of hypotension: International Classification of Disease, Ninth Revision diagnosis code for hypotension (458.0, 458.1, 458.2, 458.8, or 458.9) NI regarding how BP was monitored NI as to whether any saline boluses were given preceding inotrope initiation The trend of use of inotropes between 2001 and 2012 was reported. The most used inotrope was dopamine (65.3%), followed by dobutamine (19.9%), HC (18.8%), epinephrine (16.6%), vasopressin (0.8%), norepinephrine (0.6%), and phenylephrine (0.4%) Whilst dopamine and dobutamine usage declined, the use of HC and vasopressin increased For the sub-group of ELBW infants, the most commonly used inotrope was dopamine (83.4%) followed by HC (33%), epinephrine (32.8%), dobutamine (28.4%), phenylephrine (0.9%), vasopressin (0.7%), and norepinephrine (0.3%). The use of dobutamine declined, and the use of HC substantially increased over the study period There was no significant improvement in survival of hypotensive neonates during the 10-year study period Those neonates who required ≥4 inotropes had a >60% mortality rate |
| Robertson, 2017, USA, retrospective study | 53 | 102 | I: 38.2 ± 1.8 |
I: 3,021.8 ± 565 |
NI | 1–1.5 mg/kg every 6 h for 2 days followed by 0.5 mg/kg every 6 h until the clinical situation necessitates |
Neonates ≥34 weeks of gestation with CDH were included Definition of hypotension not reported IBP/NIBP not mentioned Use of the volume expander was not mentioned. Dose of inotropes, and the titration and initiation of steroid not mentioned Patients treated with HC had refractory hypotension (definition not mentioned) and included a significantly sicker group of infants (low APGAR score at 1 and 5 min and type C and D CDH defect) HC stress therapy was initiated within 48 h of delivery in 58.5% of the patients Total duration of HC averaged 17.8 ± Mortality was significantly higher for those treated with HC as expected No significant differences were found in survival between the high and low cortisol groups, but mortality trended higher in patients with high cortisol levels who received HC After multivariate analysis, the HC stress dose administration duration was associated with an increased risk of mortality, and the total duration of HC treatment was associated with an increased risk of sepsis |
| Ramanathan, 1996, USA, retrospective study | 19 | 15 | I: 25 ± 2 |
I: 705 ± 172 |
Late (corticosteroids initiated when highest dosages of dopamine and dobutamine were reached) | NI |
ELBW infants were included Only the abstract was available I: HC with inotropes, C: Dexamethasone with inotropes Definition of hypotension: NI In all infants, inotropes could be weaned off within 12–48 h after corticosteroid initiation Serum cortisol levels were low and the levels did not correlate with the degree of hypotension No difference in death and incidence of bacterial sepsis was noted between the groups |
| Vishveshwara, 1996, USA, prospective (pre-post study) | 41 | NI | NI | 2,297 ± 893 | Late (HC was initiated when the requirement of dopamine was >17 μg/kg/min) | 10 mg/kg followed by 5 mg/kg q 8 h for 2 days |
Only the abstract was available HC was significantly efficacious in reducing the dose of dopamine and in maintaining the BP within the target limits |
| Verma, 2017, USA, retrospective study | 69 |
74 (non-refractory hypotension) | I: 25.1 ± 1.3 |
I: 676.1 ± 121.4 |
Late (HC was started when the maximum dose of dopamine, dobutamine, and/or epinephrine was required) | 2–4 mg/kg/d |
ELBW infants with refractory hypotension were included. Refractory hypotension was defined as hypotension persisting despite treatment with maximum doses of dopamine, dobutamine (20 μg/kg/min), and/or epinephrine (0.5–1 μg/kg/min) in addition to volume expansion Definition of hypotension: MAP < GA IBP was used Hypotensive infants received 10–20 ml/kg of normal saline or 5% albumin prior to inotrope initiation (the number of boluses was not defined and was based on the decision of the clinical team) Infants in the HC group had lower BW, GA, and higher mean airway pressure and oxygen requirements The receipt of ANS and incidence of maternal GDM were lower in the HC group The HC group had a higher risk for IVH which declined in the multivariate analysis Hypotension in ELBW infants who had a GA of ≤25 and were unexposed to ANS and GDM were refractory to inotrope therapy. The authors concluded that such neonates may benefit from an initial therapy with, or earlier institution of hydrocortisone No difference in BPD, PDA, NEC, and SIP was found between the two groups |
GA, gestational age; BW, birth weight; I, intervention group; C, control group; SD, standard deviation; CDH, congenital diaphragmatic hernia; HC, hydrocortisone; ANS, antenatal corticosteroids; GDM, gestational diabetes mellitus; IVH, intraventricular hemorrhage; BPD, bronchopulmonary dysplasia; NEC, necrotizing enterocolitis; PDA, patent ductus arteriosus; SIP, spontaneous intestinal perforation; BP, blood pressure; MAP, mean arterial pressure; IBP, invasive BP; NIBP, non-invasive BP; ELBW, extremely low birth weight; NA, not applicable; NI, no information.
Except for two RCTs (
Whilst only two non-RCTs had a moderate risk of overall bias (
The meta-analyses of RCTs indicated that clinical benefit or harm could not be ruled out for the outcome of mortality with the use of early or late hydrocortisone along with inotropes when compared to inotropes alone as the effect estimates were statistically non-significant and the certainty of evidence was very low [early initiation risk ratio (RR): 0.46, 95% confidence interval (CI): 0.03–7.92; late initiation RR: 0.43, 95% CI: 0.12–1.47)]. No observational study reported the outcome of mortality for the use of early hydrocortisone (
Forest plots depicting the outcomes of mortality and treatment response from randomized controlled trials (RCTs) and non-RCTs.
The very low certainty of evidence suggested that early and late hydrocortisone administration possibly improved the response to inotrope therapy (early initiation RR: 1.85, 95% CI: 1.26–2.71; late initiation RR: 2.50, 95% CI: 1.16–5.39) (
Clinical benefit or harm could not be ruled out for any of the clinical outcomes from the meta-analyses of RCTs for either early or late hydrocortisone due to statistically non-significant effect estimates and very low to low evidence certainty (
The very low certainty of evidence from an observational study indicated that the late hydrocortisone therapy possibly increased the risk of NEC ≥ stage 2 (RR: 2.46, 95% CI: 1.19–5.08) (
Clinical benefit or harm could not be ruled out for this outcome for early hydrocortisone therapy as reported from a single RCT (
Forest plots depicting the outcomes of additional inotrope requirement, change in blood pressure (BP,) and inotrope duration from randomized controlled trials (RCTs) and non-RCTs.
The low certainty of evidence suggested that early hydrocortisone therapy possibly decreased the risk of the requirement of additional inotropes (RR: 0.18, 95% CI: 0.05–0.69) (
Whilst early hydrocortisone initiation possibly decreased the duration of inotrope therapy (MD: −39.8, 95% CI: −30.29 to −49.31; very low certainty), late hydrocortisone initiation possibly increased the duration (MD: 61.75 h, 95% CI: 43.98–79.52; low certainty) (
Clinical benefit or harm could not be ruled out for the outcome of duration of ventilation for early and late hydrocortisone administration (
The study by Ng et al. found a higher incidence of glycosuria in the hydrocortisone-treated group (
Bourchier and Weston reported that there was no correlation between the levels of cortisol (after ACTH stimulation) either at baseline or during therapy and the maximum dopamine dosage (
Amongst the eight studies included in the narrative review, all except one had evaluated late hydrocortisone therapy. (
The EtD framework and the recommendations are provided in