Scolaris Content Display Scolaris Content Display

Systemic corticosteroid regimens for prevention of bronchopulmonary dysplasia in preterm infants

Collapse all Expand all

Background

Systematic reviews showed that systemic postnatal corticosteroids reduce the risk of bronchopulmonary dysplasia (BPD) in preterm infants. However, corticosteroids have also been associated with an increased risk of neurodevelopmental impairment. It is unknown whether these beneficial and adverse effects are modulated by differences in corticosteroid treatment regimens related to type of steroid, timing of treatment initiation, duration, pulse versus continuous delivery, and cumulative dose.

Objectives

To assess the effects of different corticosteroid treatment regimens on mortality, pulmonary morbidity, and neurodevelopmental outcome in very low birth weight infants.

Search methods

We conducted searches in September 2022 of MEDLINE, the Cochrane Library, Embase, and two trial registries, without date, language or publication‐ type limits. Other search methods included checking the reference lists of included studies for randomized controlled trials (RCTs) and quasi‐randomized trials.

Selection criteria

We included RCTs comparing two or more different treatment regimens of systemic postnatal corticosteroids in preterm infants at risk for BPD, as defined by the original trialists. The following comparisons of intervention were eligible: alternative corticosteroid (e.g. hydrocortisone) versus another corticosteroid (e.g. dexamethasone); lower (experimental arm) versus higher dosage (control arm); later (experimental arm) versus earlier (control arm) initiation of therapy; a pulse‐dosage (experimental arm) versus continuous‐dosage regimen (control arm); and individually‐tailored regimens (experimental arm) based on the pulmonary response versus a standardized (predetermined administered to every infant) regimen (control arm). We excluded placebo‐controlled and inhalation corticosteroid studies.

Data collection and analysis

Two authors independently assessed eligibility and risk of bias of trials, and extracted data on study design, participant characteristics and the relevant outcomes. We asked the original investigators to verify if data extraction was correct and, if possible, to provide any missing data. We assessed the following primary outcome: the composite outcome mortality or BPD at 36 weeks' postmenstrual age (PMA). Secondary outcomes were: the components of the composite outcome; in‐hospital morbidities and pulmonary outcomes, and long‐term neurodevelopmental sequelae. We analyzed data using Review Manager 5 and used the GRADE approach to assess the certainty of the evidence.

Main results

We included 16 studies in this review; of these, 15 were included in the quantitative synthesis. Two trials investigated multiple regimens, and were therefore included in more than one comparison. Only RCTs investigating dexamethasone were identified.

Eight studies enrolling a total of 306 participants investigated the cumulative dosage administered; these trials were categorized according to the cumulative dosage investigated, 'low' being < 2 mg/kg, 'moderate' being between 2 and 4 mg/kg, and 'high' > 4 mg/kg; three studies contrasted a high versus a moderate cumulative dose, and five studies a moderate versus a low cumulative dexamethasone dose. We graded the certainty of the evidence low to very low because of the small number of events, and the risk of selection, attrition and reporting bias. Overall analysis of the studies investigating a higher dose versus a lower dosage regimen showed no differences in the outcomes BPD, the composite outcome death or BPD at 36 weeks' PMA, or abnormal neurodevelopmental outcome in survivors assessed. Although there was no evidence of a subgroup difference for the higher versus lower dosage regimens comparisons (Chi2 = 2.91, df = 1 (P = 0.09), I2 = 65.7%), a larger effect was seen in the subgroup analysis of moderate‐dosage regimens versus high‐dosage regimens for the outcome cerebral palsy in survivors. In this subgroup analysis, there was an increased risk of cerebral palsy (RR 6.85, 95% CI 1.29 to 36.36; RD 0.23, 95% CI 0.08 to 0.37; P = 0.02; I² = 0%; NNTH 5, 95% CI 2.6 to 12.7; 2 studies, 74 infants). There was evidence of subgroup differences for higher versus lower dosage regimens comparisons for the combined outcomes death or cerebral palsy, and death and abnormal neurodevelopmental outcomes (Chi2 = 4.25, df = 1 (P = 0.04), I2 = 76.5%; and Chi2 = 7.11, df = 1 (P = 0.008), I2 = 85.9%, respectively). In the subgroup analysis comparing a high dosage regimen of dexamethasone versus a moderate cumulative‐dosage regimen, there was an increased risk of death or cerebral palsy (RR 3.20, 95% CI 1.35 to 7.58; RD 0.25, 95% CI 0.09 to 0.41; P = 0.002; I² = 0%; NNTH 5, 95% CI 2.4 to 13.6; 2 studies, 84 infants; moderate‐certainty evidence), and death or abnormal neurodevelopmental outcome (RR 3.41, 95% CI 1.44 to 8.07; RD 0.28, 95% CI 0.11 to 0.44; P = 0.0009; I² = 0%; NNTH 4, 95% CI 2.2 to 10.4; 2 studies, 84 infants; moderate‐certainty evidence). There were no differences in outcomes between a moderate‐ and a low‐dosage regimen.

Five studies enrolling 797 infants investigated early initiation of dexamethasone therapy versus a moderately early or delayed initiation, and showed no significant differences in the overall analyses for the primary outcomes. The two RCTs investigating a continuous versus a pulse dexamethasone regimen showed an increased risk of the combined outcome death or BPD when using the pulse therapy. Finally, three trials investigating a standard regimen versus a participant‐individualized course of dexamethasone showed no difference in the primary outcome and long‐term neurodevelopmental outcomes.

We assessed the GRADE certainty of evidence for all comparisons discussed above as moderate to very low, because the validity of all comparisons is hampered by unclear or high risk of bias, small samples of randomized infants, heterogeneity in study population and design, non‐protocolized use of ‘rescue’ corticosteroids and lack of long‐term neurodevelopmental data in most studies.

Authors' conclusions

The evidence is very uncertain about the effects of different corticosteroid regimens on the outcomes mortality, pulmonary morbidity, and long term neurodevelopmental impairment. Despite the fact that the studies investigating higher versus lower dosage regimens showed that higher‐dosage regimens may reduce the incidence of death or neurodevelopmental impairment, we cannot conclude what the optimal type, dosage, or timing of initiation is for the prevention of BPD in preterm infants, based on current level of evidence. Further high quality trials would be needed to establish the optimal systemic postnatal corticosteroid dosage regimen.

PICOs

Population
Intervention
Comparison
Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

Different timing and dosages of corticosteroids to prevent lung injury

Review question

What timing and dosage of corticosteroids (a class of drugs that suppress inflammation) are best for preventing lung injury in babies born very early.

Background

Babies who are born too early have an increased risk of developing lung injury. In medical terms, this is called chronic lung disease (CLD) or bronchopulmonary dysplasia (BPD). Inflammation of the lungs is one of the causes of these lung problems, and for this reason studies have investigated the anti‐inflammatory drugs called corticosteroids. These studies showed that corticosteroid treatment reduced the risk of BPD, but it was also associated with serious side effects on development later in life. To reduce these side effects, doctors have looked for alternative courses of these drugs, such as postponing the start of corticosteroid therapy to a later period in life, lowering the total dose of the drug given, giving the drugs only for some days and then pausing for some time instead of every day, or deciding on the total dose or the length of the course of the drug depending on how the baby is doing instead of using a standard dose for all babies.

What did we do?

We searched electronic databases and found 16 studies investigating two or more different corticosteroid courses in preterm babies. The investigated courses differed in the total dose of the drug that was given, timing of start of the drug, and duration and schedule of therapy.

Main results

We identified 16 studies investigating different timing of initiation and dosages of corticosteroid therapy. The studies comparing a higher‐ versus a lower‐dose course showed no difference in the chance of developing BPD between the two groups, but there are concerns of an increased risk of poor development later in life for infants receiving a lower total dose of the drug. The studies investigating an earlier versus later start of steroids did not show any difference in outcome. Furthermore, courses that gave steroids on some days with pauses in between instead of every day showed a higher chance of BPD compared with everyday treatment. Deciding on the total doses and length of the course depending on how the baby was doing showed no differences compared to using the standard course for all babies.

What are the limitations of the evidence?

We have very limited confidence in the evidence, because most of the studies had limitations in study design. Most studies had a small sample size, and there were considerable differences between the studies that made it hard to compare them. Most of the studies were too short to provide information on the babies' longer‐term development. Therefore, it is not very well known what the best course of therapy is to prevent BPD.

How up to date is this evidence?

This review updates our previous review. The evidence is up to date to September 2022.

Authors' conclusions

Implications for practice

The present review includes all randomized studies to date that have investigated different corticosteroid treatment regimens head‐to‐head; all studies used the corticosteroid dexamethasone. Despite the fact that some studies reported a modulating effect of a specific treatment regimen in favor of higher dosage on the incidence of neurodevelopmental impairment, we cannot draw conclusions on the optimal type, dosage, or timing of initiation for the prevention of bronchopulmonary dysplasia(BPD) in preterm infants, based on current evidence. The evidence is very uncertain about the effects of different corticosteroid regimens on the outcomes mortality, pulmonary morbidity, and long term neurodevelopmental impairment. Furthermore, the results of this review do not justify a change in the recommendation published in international guidelines of corticosteroid use to not use these drugs outside the realm of a well‐designed clinical trial. This review demonstrated that a well‐designed large randomized controlled trial (RCT) is needed to establish the optimal systemic postnatal corticosteroid dosage regimen.

Implications for research

In light of the ongoing use of dexamethasone in the clinical setting, we feel that an RCT on dexamethasone dose and timing is justified and, in fact, urgently needed. A large multicenter study with a factorial design should compare a higher cumulative dexamethasone dose with a lower dose, as well as timing of initiation. Although the current evidence prevents firm recommendations, the present review suggests the trial should compare dexamethasone doses in the higher ranges. Obviously, the trial should be adequately powered to detect small but clinically relevant treatment effects and interaction between dose and timing of initiation. It should include ventilated preterm infants with a high risk for BPD based on the known determinants in the development of BPD. The time window to initiate dexamethasone treatment between seven days and 14 days after birth should be compared with initiation after that time period, as suggested by the recently published network analysis (Ramaswamy 2021). We recommend that data on the following primary outcome parameters be collected in any future comparative study: BPD at 36 weeks' postmenstrual age (PMA), mortality at 36 weeks' PMA and at discharge, and neurodevelopmental outcome using predefined definitions, assessed with standardized, validated, reliable diagnostic and functional outcome measurement instruments at standardized time points. In addition, short‐term benefits (time of extubation, ventilation time, discharge home on oxygen, length of hospital stay) and adverse effects (hypertension, sepsis, hyperglycemia, and the need for tracheostomy) should be collected as secondary outcomes. Various threats to the internal validity of the trial should be recognized and contained, such as potential dilution of treatment effect due to the use of 'rescue' corticosteroids outside the study protocol, or crossing over between trial arms. In any event, additional treatments should be adequately reported in order to assess the possibility of confounding.

Summary of findings

Open in table viewer
Summary of findings 1. Lower compared to higher cumulative dose dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants

Lower compared to higher cumulative dose dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants

Patient or population: preterm infants at risk for bronchopulmonary dysplasia
Intervention: lower cumulative dose dexamethasone regimen
Comparison: higher cumulative dose dexamethasone regimen

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with higher cumulative dose dexamethasone regimen

Risk with lower

Death or bronchopulmonary dysplasia at 36 weeks PMA

Study population

RR 1.03
(0.86 to 1.24)

268
(7 RCTs)

⊕⊕⊝⊝
LOWa,b

651 per 1000

671 per 1000
(560 to 808)

Death or cerebral palsy at 1 to 3 years

Study population

RR 1.74
(0.94 to 3.24)

193
(4 RCTs)

⊕⊕⊝⊝
LOWa,b

308 per 1000

535 per 1000
(289 to 997)

Death or abnormal neurodevelopmental outcome (various definitions) at 1 to 3 years

Study population

RR 1.86
(0.98 to 3.53)

100
(3 RCTs)

⊕⊝⊝⊝
VERY LOWa,b,c

172 per 1000

319 per 1000
(168 to 606)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; PMA: postmenstrual age; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded one level for risk of bias in some included studies
b Downgraded one level for serious imprecision of effect estimates (95% CI around estimate consistent with substantial harm or benefit).
c Downgraded one level for serious inconsistency across studies.

Open in table viewer
Summary of findings 2. Later compared to earlier initiation of dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants

Later compared to earlier initiation of dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants

Patient or population: preterm infants at risk for bronchopulmonary dysplasia
Intervention: later initiation of dexamethasone therapy
Comparison: earlier initiation of dexamethasone therapy

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with earlier initiation of dexamethasone therapy

Risk with later

Death or bronchopulmonary dysplasia at 36 weeks' PMA

Study population

RR 1.06
(0.87 to 1.29)

391
(3 RCTs)

⊕⊕⊝⊝
LOWa,b

476 per 1000

505 per 1000
(414 to 614)

Death or cerebral palsy at 1 to 3 years

Study population

RR 1.12
(0.68 to 1.84)

86
(1 RCT)

⊕⊕⊝⊝
LOWa,b

412 per 1000

461 per 1000
(280 to 758)

Death or abnormal neurodevelopmental outcome (various definitions)

Study population

RR 0.87
(0.63 to 1.21)

167
(2 RCTs)

⊕⊕⊝⊝
LOWa,b

507 per 1000

441 per 1000
(319 to 613)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; PMA: postmenstrual age; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded one level for serious study design limitations (unclear methodology or random sequence allocation, lack of blinding of clinicians and outcome assessment (Bloomfield 1998; Merz 1999; Hingre 1992; Halliday 2001a)
b Downgraded one level for serious imprecision of effect estimate (95% CI around estimate consistent with substantial harm or benefit and number of events < 300).

Open in table viewer
Summary of findings 3. Pulse compared to continuous dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants

Pulse compared to continuous dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants

Patient or population: preterm infants at risk for bronchopulmonary dysplasia
Intervention: pulse dexamethasone therapy
Comparison: continuous dexamethasone therapy

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with continuous dexamethasone therapy

Risk with pulse

Death or bronchopulmonary dysplasia at 36 weeks' PMA

Study population

RR 1.38
(1.02 to 1.88)

197
(2 RCTs)

⊕⊕⊝⊝
LOWa,b

390 per 1000

538 per 1000
(398 to 733)

Death or abnormal neurodevelopmental outcome (various definitions)

Study population

RR 1.23
(0.79 to 1.92)

76
(1 RCT)

⊕⊝⊝⊝
VERY LOWa,b,c

459 per 1000

565 per 1000
(363 to 882)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; PMA: postmenstrual age; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded one level for serious study design limitations (lack of blinding of clinicians and outcome assessment (Bloomfield 1998))
b Downgraded one level for serious imprecision of effect estimate (95% CI around estimate consistent with substantial harm or benefit and number of events < 300).
c Downgraded one level for publication bias, since the study by Barkemeyer 2001 was never published as full text.

Open in table viewer
Summary of findings 4. Individual tailored compared to continuous tapered dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants

Individual tailored compared to continuous tapered dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants

Patient or population: preterm infants at risk for bronchopulmonary dysplasia
Intervention: individual tailored dexamethasone regimen
Comparison: continuous tapered dexamethasone regimen

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with continuous tapered dexamethasone regimen

Risk with individual tailored

Death or bronchopulmonary dysplasia at 36 weeks PMA

Study population

RR 1.06
(0.88 to 1.29)

168
(3 RCTs)

⊕⊕⊝⊝
LOWa,b

639 per 1000

677 per 1000
(562 to 824)

Death or cerebral palsy at 1 to 3 years

Study population

RR 7.24
(0.95 to 55.26)

59
(1 RCT)

⊕⊕⊕⊝
MODERATEb

33 per 1000

241 per 1000
(32 to 1000)

Death or abnormal neurodevelopmental outcome (various definitions)

Study population

RR 1.44
(0.99 to 2.07)

168
(3 RCTs)

⊕⊕⊝⊝
LOWa,b

313 per 1000

451 per 1000
(310 to 648)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; PMA: postmenstrual age; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded one level for serious study design limitations (unclear methodology or random sequence allocation, lack of blinding of clinicians and outcome assessment (Bloomfield 1998; Odd 2004 )
b Downgraded one level for serious imprecision of effect estimate (95% CI around estimate consistent with substatial harm or benefit and number of events < 300).

Background

Description of the condition

The first description of bronchopulmonary dysplasia (BPD) by Northway and colleagues in 1967 was one of severe lung injury in relatively mature preterm infants who were ventilated with high pressures and high concentrations of oxygen before the advent of surfactant therapy (Northway 1967). This so‐called 'classical' BPD is characterized by profound lung parenchymal inflammation, fibrosis, muscle hypertrophy and diffuse airway damage (O'Brodovich 1985). Treatment and survival of the very young has led to a new pattern of lung injury (Coalson 2006; Jobe 1999). This so‐called 'new' BPD is mainly seen in very preterm infants with gestational ages less than 30 weeks. It is characterized by an arrest in lung development with fewer and larger alveoli, and less striking fibrosis and inflammation (Husain 1998). As a result of changes in infant and histological characteristics, the timing at which BPD is diagnosed has shifted from 28 days' postnatal age (PNA) to 36 weeks' postmenstrual age (PMA) (Bancalari 2006). Cohort studies have shown that, compared with 28 days' PNA, diagnosing BPD at 36 weeks' PMA provides a better identification of infants at risk for long‐term pulmonary and neurological sequelae (Ehrenkranz 2005).

BPD, defined as oxygen dependency at 36 weeks' PMA, remains an important complication of preterm birth with a reported incidence ranging from 23% to 73%, depending on the gestational age (Stoll 2010). BPD is characterized by prolonged respiratory support and recurrent respiratory infections during the first years, and compromised lung function lasting into adulthood. Furthermore, BPD is an independent risk factor for neurodevelopmental impairment (Short 2007; Walsh 2005).

BPD is considered a multifactorial disease. Besides genetic susceptibility, intrauterine growth restriction, nutritional deficits, direct mechanical injury caused by artificial ventilation, oxygen toxicity, and pulmonary inflammation has been identified as a key factor in the development of BPD (Carlton 1997; Ferreira 2000; Jobe 2001). Corticosteroids have a strong anti‐inflammatory effect, making them an ideal candidate to attenuate this inflammatory response associated with BPD.

Description of the intervention

Since the 1980s, several randomized controlled trials (RCTs) have investigated the use of corticosteroids, in particular dexamethasone, as a means to reduce the incidence of BPD. Some of these trials started corticosteroid therapy in the first week of life (early), with the aim of preventing progression of the initial acute inflammatory response to BPD (Yeh 1997). Others used corticosteroid therapy in infants who had evolving BPD, starting administration either moderately early (7 to 14 days) or delayed (more than three weeks) after birth (CDTG 1991; Durand 1995).

Current Cochrane Reviews of placebo‐controlled RCTs clearly showed that systemic corticosteroids, mainly dexamethasone, significantly reduced the incidence of BPD and the combined outcome of death or BPD in ventilated preterm infants, independent of the time of postnatal administration (Doyle 2021a; Doyle 2021b). However, at the end of the 1990s the first reports on long‐term neurodevelopmental outcome were published, showing that early postnatal systemic dexamethasone treatment is associated with an increased risk of abnormal neurological development (O'Shea 1999; Yeh 1998).

In response to these reports, the American Academy of Pediatrics, the Canadian Paediatric Society and the European Association of Perinatal Medicine concluded that routine use of systemic dexamethasone in the treatment of evolving BPD can no longer be recommended until further research has established the optimal type, dose and timing of corticosteroid therapy (AAP 2002; Halliday 2001; Watterberg 2010). Following these statements, observational reports have shown a sharp decline in the use of postnatal corticosteroids, a reduction in its cumulative dose, a delay in starting treatment, and a switch to alternative corticosteroids such as hydrocortisone (Kaempf 2003; Shinwell 2003; Walsh 2006).

How the intervention might work

To date, most studies have used a placebo‐controlled design to study the effects of postnatal corticosteroid treatment in preterm infants at risk for BPD. These studies have shown both benefits and harms of corticosteroid treatment. Adjusting the dosage regimen might improve the benefit‐to‐risk ratio of postnatal corticosteroid use. This review identified and analyzed the available randomized trials, using a head‐to‐head comparative design, on five possible treatment regimens.

  1. Alternative corticosteroids The association between systemic dexamethasone treatment and long‐term neurodevelopmental impairment has resulted in the use of alternative anti‐inflammatory corticosteroids, such as hydrocortisone. Animal studies have suggested that, in contrast to dexamethasone, hydrocortisone has no detrimental effect on the brain (Huang 2007). Historical cohort studies have suggested that hydrocortisone treatment is equally effective in reducing death or BPD compared with dexamethasone‐treated infants without increasing the risk of adverse neurological outcome (van der Heide‐Jalving 2003; Karemaker 2006; Lodygensky 2005; Rademaker 2007). Pooled data on placebo‐controlled trials investigating a low hydrocortisone dose initiated at an early treatment onset (< 7 days' PNA) showed reduced rates of mortality, and of the combined outcome of mortality or bronchopulmonary dysplasia, without causing any obvious long‐term harm. However, gastrointestinal perforation was more frequent in the hydrocortisone group (Doyle 2021a). The only large placebo‐controlled randomized trial investigating the use of hydrocortisone after the first week of life in ventilator‐dependent preterm infants showed no improvement in the outcome BPD, or the composite outcome death or BPD (Onland 2019).

  2. Lowering the corticosteroid dose and duration In line with the current opinion of postnatal corticosteroids being 'misguided rockets', clinicians have started to use lower dosage schedules of dexamethasone. The available reviews on placebo‐controlled trials of postnatal corticosteroids stacked information from trials with tremendous heterogeneity in their cumulative dose and duration of therapy (Doyle 2021a; Doyle 2021b). Subgroup analyses using this heterogeneity by dividing the different trials according to the used cumulative dexamethasone dose showed that higher dexamethasone doses reduce the typical risk ratio (RR) for the combined outcome of death or BPD, with the largest treatment effect in trials using a cumulative dose above 4 mg/kg (Onland 2009). No overall effect was found of dosing on the risk of neurodevelopmental sequelae, but in the moderately early treatment studies the risk of death or cerebral palsy (CP) significantly decreased when using a higher cumulative dose (Onland 2009).

  3. Postponing initiation of therapy Besides lowering the cumulative dose, clinicians limited the use of corticosteroids to those infants that do not respond to other supportive therapies and spontaneous improvement over time. As a result, administration of postnatal corticosteroids in those infants is often postponed until the third or fourth week of life. Placebo‐controlled trials administrating dexamethasone after the first week of life differ in their timing of onset. Meta‐analysis dividing the different placebo‐controlled studies according to the timing of initiation used seems to suggest that moderately early administration is more effective in reducing BPD than delayed administration (Onland 2009; Schmidt 2008).

  4. Pulse dose administration To minimize the possible adverse effects associated with continuous corticosteroid use, some have suggested prescribing dexamethasone in a pulse regimen using dexamethasone‐free intervals to minimize the risk of direct toxic effects of dexamethasone, while maintaining the beneficial effects on the lung. One placebo‐controlled trial showed that such a pulse regimen resulted in improved pulmonary outcome without clinically relevant side effects (Brozanski 1995).

  5. Individualized tailored regimen Another approach is to reduce the risk of possible adverse effects of corticosteroids by tailoring the administered cumulative dose to the infant's pulmonary response. For instance, a rapid and clear improvement in respiratory status will allow for a rapid reduction in corticosteroid dose or duration (Bloomfield 1998). To date, there are no placebo‐controlled trials on individualized regime.

Why it is important to do this review

The international neonatal community has discarded the use of early postnatal corticosteroids completely for the reasons stated above. Regarding the use of moderately early or late postnatal systemic corticosteroids, clinicians encounter a dilemma facing those infants at high risk of BPD, since BPD itself is associated with an increased risk of adverse neurological outcome (Ehrenkranz 2005).

It is unknown whether both the beneficial and adverse treatment effects of postnatal corticosteroids can be modulated by the various different dosing regimens described above. Despite all the aforementioned concerns on the long‐term neurodevelopmental sequelae, corticosteroids are still used in approximately 16% of preterm infants (Costeloe 2012). Clinicians remain in doubt as to what the correct drug, cumulative dose, duration and timing of therapy are in terms of the optimal balance between beneficial and adverse effects. Addressing these questions is also important since studies have suggested that restricted use of postnatal corticosteroids resulted in an increased incidence of BPD (Cheong 2013; Shinwell 2007; Yoder 2009).

Objectives

To assess the effects of different corticosteroid treatment regimens on mortality, pulmonary morbidity, and neurodevelopmental outcome in very low birth weight infants.

Methods

Criteria for considering studies for this review

Types of studies

Randomized controlled or quasi‐randomized and cluster‐randomized trials comparing two or more different regimens of systemic corticosteroids in preterm infants at risk for BPD were eligible for this review. Non‐randomized cohort studies were not eligible for this review, given the fact of potential bias of confounding by indication or residual confounding influencing the results of studies with such designs (Fewell 2007; Kyriacou 2016). We excluded studies investigating the effects of one regimen of systemic corticosteroids versus a placebo arm or studies using inhalation corticosteroids. Those studies are included in other Cochrane Reviews.

Types of participants

Eligible participants were preterm infants at risk for BPD, as defined by the original trialists.

Types of interventions

We included trials which randomized infants to treatment with two different regimens of systemic corticosteroids. The following types of intervention were eligible.

  • An alternative corticosteroid (e.g. hydrocortisone) as the experimental arm versus another type of corticosteroid (e.g. dexamethasone) as the control arm. Any type of corticosteroid in either arm was allowed.

  • Lower cumulative corticosteroid dosage (experimental arm) versus higher cumulative corticosteroid dosage (control arm). Both arms of the identified trials were categorized according to the cumulative dosage investigated, 'low' being less than 2 mg/kg, 'moderate' being between 2 and 4 mg/kg, and 'high' using a cumulative dosage greater than 4 mg/kg. For inclusion, all comparisons of low‐, moderate‐ or high‐dosage regimens were allowed. Although arbitrary, these cut‐off values were chosen given the results of a systematic review of placebo‐controlled trials (Onland 2009).

  • Later (experimental arm) versus earlier (control arm) initiation of therapy. We categorized both arms of the identified trials according to the investigated timing of initiation, 'early' being less than 8 days' PNA, 'moderately early' being between 8 and 21 days' PNA, and 'delayed' being greater than 21 days' PNA. Similar to the dosing analyses, all comparisons were allowed. This arbitrary cut‐off point was chosen according to the original Cochrane Reviews on placebo‐controlled trials (Halliday 2003a; Halliday 2003b; Halliday 2003c).

  • Pulse‐dosage regimen (experimental arm) versus continuous‐dosage regimen (control arm). During pulse therapy, the administration of corticosteroids is interrupted for a period longer than the normal interval between corticosteroid doses. Any period of interruption was allowed.

  • Individually tailored regimens (experimental arm) based on the pulmonary response defined by the original trialists versus a standardized (predetermined schedule administered to every infant) dosage regimen independent of the pulmonary response (control arm).

Types of outcome measures

In the previous version of this review, two review authors (WO and ADJ) independently extracted the following outcome parameters for each study. In the current update of this review, one review author entered final data into Review Manager (WO) and a second review author checked the data for accuracy (MvdL).

Primary outcomes

  • Combined outcome of death or BPD at 36 weeks' PMA (BPD defined as the need for respiratory support or oxygen dependency at 36 weeks' PMA).

Secondary outcomes

  • Mortality at 28 days' PNA, 36 weeks' PMA, hospital discharge and during the first year of life.

  • BPD (defined by the need for supplemental oxygen) at 28 days' PNA and 36 weeks' PMA.

  • Failure to extubate at days three and seven after initiating therapy and at the latest reported time point.

  • Days of mechanical ventilation and supplemental oxygen.

  • Complications during primary hospitalization: hypertension, defined as more than two standard deviations (SD) according to local protocols; hyperglycemia, defined as greater than 8.3 mmol/L or requiring insulin therapy; rescue treatment with open‐label corticosteroids within or outside the study period; culture‐confirmed and clinically suspected infection; gastrointestinal bleeding or perforation, spontaneous intestinal perforation (SIP); necrotizing enterocolitis (NEC), following Bell's stages; patent ductus arteriosus (PDA), according to trial protocol and requiring therapy; intraventricular hemorrhage (IVH), any and severe grades; periventricular leukomalacia (PVL); cardiac hypertrophy; and retinopathy of prematurity (ROP), any and severe stages.

  • Long‐term neurodevelopmental sequelae, assessed after at least one year corrected gestational age (CGA) and before a CGA of four years, and at the latest reported time point, including cerebral palsy and Bayley Scales of Infant Development (Mental Development Index, MDI), blindness, and deafness.

Search methods for identification of studies

The Neonatal Group Information Specialist, in consultation with the authors, revised the search strategies for this update to incorporate a more sensitive approach to drug names for interventions (corticosteroids) and bronchopulmonary dysplasia.

Electronic searches

The following databases were searched without language or publication status restrictions. The RCT search was not limited by date; the search for systematic reviews was limited from 2020 forward.

  • Cochrane Central Register of Controlled Trials, Issue 9, 2022 (Wiley)

  • Cochrane Database of Systematic Reviews, Issue 9, 2022 (Wiley)

  • Ovid MEDLINE and Epub Ahead of Print, In‐Process, In‐Data‐Review & Other Non‐Indexed Citations, Daily and Versions 1946 to 9 September 2022

  • Embase (OVID) 1974 to 12 September 2022

Search strategies are available in Appendix 1; Appendix 2; and Appendix 3. Search strategies used in 2016 are available in Appendix 4.

Searching other resources

Two trial registries were searched without date limits.

Search strategies are available in Appendix 5.

We searched the following websites for conference abstracts from 1990 forward.

  • Pediatric Academic Societies (PAS)

  • European Society for Pediatric Research

Electronic searches were supplemented by contacting original authors of all studies to confirm details of reported follow‐up studies or to obtain information about long‐term follow‐up where none were reported.

We checked the reference lists of included studies and of systematic reviews related to the topic of this review.

Data collection and analysis

Selection of studies

We used Cochrane’s Screen4Me (S4M) automatic classifier (Noel‐Storr 2020; Noel‐Storr 2021a; Noel‐Storr 2021b; Screen4Me), and S4M’s Known Assessments and RCT Classifier to exclude known non‐RCTs. Two review authors (WO and MvdL) independently screened remaining title/abstracts and full‐texts. At any point in the screening process, disagreements were resolved by discussion. Search results were managed in Endnote and screened using Covidence software (Covidence). We documented reasons for excluding studies after full‐text review, and noted their characteristics. We documented characteristics of ongoing studies and studies awaiting assessment. We collated multiple reports of the same study so that each study, not reference, is the unit of interest in the review. We recorded the search process in sufficient detail to create a study flow diagram (Liberati 2009).

Data extraction and management

In the current update of this review, one review author entered final data into Review Manager (WO) and a second review author checked the data for accuracy (MvdL). Review authors resolved any discrepancies through discussion. In the previous version of this review, two review authors (WO and ADJ) independently extracted the following data for each study, in addition to the predefined outcome measurements, using a predefined data sheet: infant's characteristics (such as birth weight, gestational age, gender); number of participants randomized; treatment with antenatal corticosteroids and postnatal surfactant; type of corticosteroid and regimens (PNA at start, duration of therapy, cumulative dose; dosing interval (fixed or variable); dose adjustments according to infant's characteristics); and the incidence of open‐label (outside the study protocol) use of corticosteroids in both arms of the studies. The original investigators of the included RCTs were asked to confirm whether the data extraction was accurate and, where necessary, to provide additional (unpublished) data.

Assessment of risk of bias in included studies

Two review authors (WO and ADJ) in the previous version, and two reviewers in the current version (WO and MvdL), independently assessed the risk of bias (low, high, or unclear) of all included trials using the Cochrane risk of bias tool in the Cochrane Handbook for Systematic Reviews of Interventions for the following domains (Higgins 2017).

  1. Selection bias

  2. Performance bias

  3. Detection bias

  4. Attrition bias

  5. Reporting bias

  6. Any other bias

Any disagreements were resolved by discussion or by a third assessor. See Appendix 6 for a more detailed description of risk of bias for each domain.

Measures of treatment effect

We conducted data management using the Cochrane statistical package, Review Manager 5 (Review Manager 2020). Where possible, we calculated treatment effect estimates for dichotomous outcomes in all individual trials expressed as the risk ratio (RR) and risk difference (RD), all with a 95% confidence interval (CI). For continuous outcomes reported in individual studies we used the mean values for treatment and control groups, with the SD. If median and range were given in individual studies, and the study authors were not able to provide the mean value and variance from the original data set, we calculated them according to the method described by Hozo 2005. We calculated the number needed to treat for an additional beneficial outcome (NNTB) and number needed to treat for an additional harmful outcome (NNTH) for each different outcome in case of statistical significance (Graphpad 2021).

Unit of analysis issues

If cluster‐randomized trials had been included in the analyses, we would have adjusted their sample sizes using the methods described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2022).

Dealing with missing data

We asked the study author of each included RCT to confirm whether the data extraction was accurate and, where necessary, to provide additional (unpublished) data.

Assessment of heterogeneity

We assessed heterogeneity between trials by inspecting the forest plots and quantifying the impact of heterogeneity using the I² statistic, using the following categories as defined by the Cochrane Neonatal Review Group.

  1. Less than 25%: no heterogeneity.

  2. 25% to 49%: low heterogeneity.

  3. 50% to 74%: moderate heterogeneity.

  4. 75% or greater: high heterogeneity.

We explored possible causes of statistical heterogeneity using prespecified subgroup analysis (e.g. differences in intervention regimens).

Assessment of reporting biases

We used funnel plots to assess possible reporting or publication biases.

Data synthesis

We performed meta‐analysis of the extracted data using standard Cochrane methods and Review Manager 5 (Review Manager 2020). Treatment effects for dichotomous outcomes were expressed as RR with a 95% CI, RD, and NNTBs or NNTHs in case of significance. We used mean differences (MD) for continuous outcomes. In case of variance of outcome measures (with different SD) measuring the same outcome, we calculated standardized mean differences (SMD) in the meta‐analysis. We used the fixed‐effect model for all meta‐analyses.

Subgroup analysis and investigation of heterogeneity

In case of substantial heterogeneity, we performed subgroup analyses and sensitivity analyses, and, if not appropriate, reconsidered whether an overall summary was meaningful at all. We planned to carry out the following subgroup analyses.

  1. Gestational age using an arbitrary cut‐off point of 26 weeks.

  2. The degree of illness at the start of treatment as defined by mean respiratory index or fractional inspired oxygen, if available, at trial entry.

  3. Ventilated versus non‐ventilated neonates at study entry.

  4. Trials allowing use of open‐label corticosteroids during the study period, by dividing the individual trials according to the percentage of infants treated with open‐label corticosteroids in the experimental arm, using arbitrary cut‐off points of less than 30%, 30% to 50%, and greater than 50% of the included infants; and trials investigating two (or more) of the main comparisons analyzed in both comparisons in subgroups. For example, if a study investigates hydrocortisone at an early initiation versus a dexamethasone regimen at a later treatment onset, this study would be analyzed in both the main comparison type of corticosteroids, as well as the comparison timing of initiation.

Sensitivity analysis

We performed sensitivity analyses when we judged trials to be at high risk of bias, to assess the effect of the bias on the meta‐analysis.

Summary of findings and assessment of the certainty of the evidence

We used the GRADE approach, as outlined in the GRADE Handbook (Schünemann 2013), to assess the quality of evidence for the following (clinically relevant) outcomes: the combined outcome of BPD or death at 36 weeks' PMA, as well as the combined outcomes of death or cerebral palsy, and death or abnormal neurodevelopmental outcome.

Two authors independently assessed the quality of the evidence for each of the outcomes above. We considered evidence from randomized controlled trials as high quality but downgraded the evidence one level for serious (or two levels for very serious) limitations based upon the following: design (risk of bias), consistency across studies, directness of the evidence, precision of estimates and presence of publication bias. We used the GRADEpro GDT Guideline Development Tool to create summary of findings tables to report the quality of the evidence.

The GRADE approach results in an assessment of the quality of a body of evidence in one of four grades.

  1. High: we are very confident that the true effect lies close to that of the estimate of the effect.

  2. Moderate: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.

  3. Low: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.

  4. Very low: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

Results

Description of studies

Results of the search

Searches identified 8758 references (8225 from databases and trial registries; 533 from conference searching). After removing 2371 duplicates, 6387 records were available for screening. We excluded 6351 references (500 via Screen4Me (Figure 1); 5851 by authors).


Screen4Me 13 September 2022

Screen4Me 13 September 2022

We assessed 35 full‐texts and one trial registry record for inclusion. We excluded five studies (five references) with reasons (Characteristics of excluded studies). We placed one study (IRCT20200721048155N1), in Awaiting classification since the current status is that recruitment has been completed but no results have been published yet (Characteristics of studies awaiting classification); and identified one ongoing study (IRCT20201222049802N3).

We included 16 studies (29 references); see Characteristics of included studies. One study, Groneck 1993, could not be included in our quantitative synthesis because outcome data were not presented in the published reports and were not available from the authors of the study. Thus, 15 studies are included in our quantitative synthesis and 16 studies in the qualitative synthesis.

Details of the selection process can be seen in Figure 2.


PRISMA flow chart

PRISMA flow chart

Included studies

Sixteen trials (29 reports) met inclusion criteria; of these, 15 are included in the quantitative synthesis. The 15 trials randomized a total of 1257 infants. A detailed description of participant characteristics of the individual trials can be found in Table 1. Most trials included preterm infants with similar ranges of gestational age and birth weight, yet there was considerable variation in the use of antenatal corticosteroids and exogenous surfactant. Pulmonary illness, assessed by the amount of supplemental oxygen and the level of mean airway pressure at study entry, differed considerably across the trials. Only three trials reported no late rescue treatment with dexamethasone in both treatment groups. The investigated regimens differed in the used cumulative dose, timing of initiation and duration of therapy.

Open in table viewer
Table 1. Participant characteristics of individual trials

Allocation arm

Participants (N)

Birthweight

(grams)

Gestational age

(weeks)

Antenatal

steroids(%)

Surfactant(%)

Starting dose (mg/kg/d)

Cumulative dose (mg/kg)‍‍

Mean age initiation

Total duration(days)

Late rescue corticosteroids

(%)

Entry FiO₂ (%)

Entry MAP

(cmH0)

Lower cumulative dosage (experimental arm) versus higher cumulative dosage (control arm)

Cummings 1989

High

13

818 ± 145

26 ± 2

38

0

0.5

7.9

14

42

0

0.60 ± 0.27

10.4 ± 6.0

Moderate

12

810 ± 208

26 ± 2

25

0

0.5

3.0

18

0

0.51 ± 0.23

8.8 ± 2.7

DeMartini 1999

High

16

741 ± 142

25.5 ± 1.7

62

100

0.5

4.1

?

21

0

0.61 ± 26.9

?

Moderate

14

848 ± 224

26.4 ± 1.6

64

100

0.5

2.7

7

0

0.60 ± 25.2

Marr 2019a

High

30

769 ± 149

25.2 ± 1.2

63

97

0.5

7.96

14 ± 4

42

20

0.72 ± 0.13

10.3 ± 2.0

Moderate

29

785 ± 167

25.2 ± 1.1

62

86

0.5

4.04

13 ± 3

9b

37

0.77 ± 0.16

10.4 ± 1.7

Malloy 2005

Moderate

9c

767 ± 149

25.8 ± 0.9

75

100

0.5

2.7

14.8 ± 6.5

7

88

0.57 ± 0.08

?

Low

8

773 ± 182

26.1 ± 1.8

63

100

0.08

0.6

16.8 ± 5.7

7

50

0.52 ± 0.16

Durand 2002

Moderate

23

932 ± 182

27.1 ± 1.8

52

87

0.5

2.4

11.5 ± 2.2

7

22

0.43 ± 0.11

7.8 ± 2.2

Low

24

858 ± 186

26.9 ± 1.6

50

88

0.2

1.0

11.3 ± 2.7

7

29

0.41 ± 0.10

7.0 ± 1.2

McEvoy 2004

Moderate

29

839 ± 229

26.1 ± 2.0

34

97

0.5

2.4

10.7 ± 3.7

7

55

0.44 ± 0.13

6.8 ± 1.8

Low

33

830 ± 248

26.3 ± 1.8

48

82

0.2

1.0

11.6 ± 4.3

7

39

0.42 ± 0.13

7.4 ± 2.2

Ramanathan 1994

Moderate

15

850 ± 290

27 ± 2

?

67

0.4

1.9d

10 to 14

7

67

?

?

Low

13

817 ± 186

27 ± 2

62

0.2

1.0d

7

54

Da Silva 2002

Moderate

17

821 ± 160

25.4 ± 0.9

?

?

0.5

?

?

7

?

?

?

Low

21

851 ± 465

25.7 ± 1.8

0.1

0.7

7

Later initiation (experimental arm) versus earlier (control arm)

Papile 1998

ME

182

808 ± 187

25.7 ± 1.9

29

91

0.5

3.7

14

14

12

0.54 ± 0.18

8 ± 2

L

189

801 ± 182

25.6 ± 1.6

27

89

28

16

0.54 ± 0.19

8 ± 2

Merz 1999

E

15

980 (710 to 1250)

27 (25 to 29)

87

87

0.5

3.1

7

16

0

0.3 (0.25 to 0.5)

?

ME

15

938 (680 to 1250)

27.5 (24 to 29)

73

73

14

0

0.3 (0.25 to 0.55)

Halliday 2001a

E

135

1017 ± 290

27.4 ± 1.9

61

95

0.5

2.7

3

12

?

?

?

ME

150

1007 ± 283

27.1 ± 1.9

55

92

16

Hingre 1992

E

14

744 ± 144

26 ± 2

?

?

0.5

7.96

5

42

?

?

?

ME

21

730 ± 135e

25 ± 2

14

Pulse dosage regimen (experimental arm) versus continuous dosage regimen (control arm)

Bloomfield 1998

Pulse/Ef

39

776 ± 25

25.8 ± 0.3

95

?

0.5

5.3 (1.5 to 11.8)

7

34 (11 to 73)

?

0.30 ± 0.02

8.0 ± 0.3

Cont/ME

37

793 ± 28

25.8 ± 0.3

73

0.5

7.1 (4.5 to 7.6)

14

42 (42 to 51)

0.30 ± 0.01

7.8 ± 0.3

Barkemeyer 2001

Pulse

58

816

26.1

84

92

0.5

4.5

7 to 21

23

41

?

?

Cont

63

842

26.2

78

88

36

Individualized tailored (experimental arm) versus standard dosage regimen (control arm)

Odd 2004

Indiv

17

669 ± 113

24 (23 to 27)

?

?

0.5

3.8 (2.0 to 5.7)

12 (7 to 16)

42 (5 to 73)

?

0.40 (0.25 to 1.0)

9 (7 to 14)

Cont

16

720 ± 130

24 (23 to 26)

0.5

7.9

10 (7 to 23)

42

0.40 (0.21 to 1.0)

9 (7 to 13)

a Marr not only in higher versus lower comparison, but also in individual tailored versus standard dosage regimen
b 19 of the 29 infants (66%) received one course, 5 infants (17%) 2 courses, and 5 infants (17%) 3 courses of dexamethasone
c Including one patient in high dose group who died on the second day of treatment
d Estimated cumulative dose based on abstract data
e participant characteristics calculated on 16 participants in the moderately early group
f Bloomfield not only pulse versus continuous comparison, but also in early versus later initiation and in individualized tailored versus standard dosage regimens comparison

Mean ± standard deviation or median (interquartile range)

Abbreviations: FiO2: fractional inspired oxygen; MAP: mean airway pressure; E: Early initiation (≤ 7 days' PNA); ME: Moderately early initiation (7 to 14 days' PNA); L: Late initiation (> 14 days' PNA); Pulse: Pulse dosage regimen; Cont: Continuous tapered dosage regimen; Indiv: Individual tailored regimen.

The trial by Marr 2019 investigated a high‐dosage tapered regimen of corticosteroids with a duration of 42 days versus a moderate‐dosage regimen of corticosteroids during nine days. Whereas the high‐dosage regimen was only given once, participants on the moderate‐dosage regimen of nine days could receive a second, or even a third course of nine days moderate dose of dexamethasone if entry respiratory criteria were met after completion of the previous course. Based on this design, we used the trial for two comparisons in this review: higher versus lower dosage regimens of corticosteroid treatment, and individualized versus standardized dosing. The trial by Bloomfield 1998 allocated infants to a group receiving a pulse dose of corticosteroids initiated early or a group receiving a continuous tapering dose of corticosteroids started moderately early. In addition, the duration of the pulse dose, but not the continuous tapering dose, was dependent on the pulmonary response of the infant. Based on this design, we used the trial for three comparisons in this review: earlier versus later initiation of corticosteroid treatment, pulse versus continuous dosing, and individualized versus standardized dosing. We included all other studies in one comparison each.

Eight of the 15 original investigators provided the authors with additional data on methodology, intervention, infant characteristics or missing outcome parameters.

Alternative corticosteroids

We did not identify any trials investigating two or more different types of corticosteroids. In fact, all trials included in this review used dexamethasone in both treatment arms.

Lowering the corticosteroid dose and duration

The timing of the eight eligible trials investigating this comparison was moderately early (7 to 21 days). The cumulative dexamethasone doses in the studies ranged from 0.6 to 3.0 mg/kg in the lower‐dosage regimens (experimental arm) to 1.9 to 7.9 mg/kg in the high‐dosage regimens (control arm). Only two dosage comparisons were identified during this review, high (> 4 mg/kg cumulative dose) versus moderate dose (between 2 and 4 mg/kg cumulative dose) and moderate‐ versus low‐ (< 2 mg/kg cumulative dose) dosage regimens. Three trials compared a high dose (control arm) to a moderate dose (Cummings 1989; DeMartini 1999; Marr 2019); and five trials a moderate dose to a low dose (Da Silva 2002; Durand 2002; Malloy 2005; McEvoy 2004; Ramanathan 1994). We analyzed and reported these two comparisons separately.

Postponing initiation of therapy

Five RCTs investigated the effect of timing on the dexamethasone treatment effects in preterm infants (Bloomfield 1998; Halliday 2001a; Hingre 1992; Merz 1999; Papile 1998). Only two comparisons were identified, namely late versus moderately early initiation, and moderately early versus early initiation of corticosteroid therapy. Papile 1998 compared late (> 21 days' PNA (experimental arm)) to moderately early (between 8 and 21 days (control arm)) initiation of treatment. The other four trials contrasted early (≤ 7 days' PNA) to moderately early (experimental arm) initiation of treatment. We analyzed these two comparisons separately. The comparison of moderately early versus early initiation included the trial performed by Halliday 2001a. This RCT used a factorial design with four allocation arms. Two arms administered corticosteroids by inhalation, and we excluded the data of the infants treated with inhalation corticosteroids from this review. The other two arms administered dexamethasone systemically, starting either early or moderately early, and we therefore included these in the analysis.

Pulse dose administration

Two trials compared pulse therapy of dexamethasone (experimental arm) with a continuous tapering dosage regimen (control arm) (Barkemeyer 2001; Bloomfield 1998). Both trials used a pulse dexamethasone therapy (0.5 mg/kg/day) for three consecutive days followed by seven days of no corticosteroid therapy. One trial administered similar cumulative doses of dexamethasone in both allocation arms (Barkemeyer 2001). However, in the other trial the duration of the pulse‐dosage regimen varied, depending on the infant’s pulmonary condition and level of respiratory support (Bloomfield 1998). The continuous tapering dosage regimen in this study, however, was the same for every infant allocated to this arm.

Individualized tailored regimen

Three trials allocated the infants to either an individualized dosage regimen (experimental arm), or a tapering dosage regimen. Two studies initiated the intervention at the same postnatal age (Odd 2004; Marr 2019), whereas the other study initiated the pulse therapy at day seven of life, comparing it to a tapering continuous‐dosage regimen commencing at day 14 of life (Bloomfield 1998).

Excluded studies

We excluded five trials after reading the full text. One had a retrospective study design (DeCastro 2009); one was a placebo controlled trial of early dexamethasone administration (Shipalana 1994); one was a report of a web‐based survey on corticosteroids (Singh 2022); one trial registration was identified without any report of completion of the study or results (Ahrens 2000); and one investigated two different dexamethasone regimens, but in a placebo controlled design (Romagnoli 1999).

Awaiting classification

One study is awaiting classification (IRCT20200721048155N1) (see Characteristics of studies awaiting classification).

Ongoing studies

We identified one trial registration title as an ongoing trial (IRCT20201222049802N3) (see Characteristics of ongoing studies).

Risk of bias in included studies

We deemed the overall risk of bias of the 15 trials to range from unclear to low (Figure 3; Figure 4). Four trials were only published as abstracts, and therefore had insufficient data to allow a proper methodological assessment (Da Silva 2002; DeMartini 1999; Hingre 1992; Ramanathan 1994).


Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included studies.

Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included studies.


Risk of bias summary: review authors' judgments about each risk of bias item for each included study.

Risk of bias summary: review authors' judgments about each risk of bias item for each included study.

Allocation

Five trials described the random sequence generation insufficiently, whereas four trials did not mention the method of allocation concealment. Therefore, only nine trials described both these items properly, and we judged them as having low risk of selection bias.

Blinding

Six trials did not attempt to blind the intervention; thus caregivers, parents and outcome assessors were not blinded. We judged these trials as being at high risk for performance and detection bias. In two trials no information on blinding was available, making it impossible to assess bias (Hingre 1992; Ramanathan 1994).

Incomplete outcome data

We judged the RCTs to be at low risk of attrition bias, because all but two trials reported data on 'loss to follow‐up' or participant selection, or both. Malloy 2005 excluded one infant who died during the study course, and for this reason was assessed as being at high risk of attrition bias. However, this infant was included in the current analyses. Hingre 1992 excluded five deceased infants in the late treatment group from the analysis, but these infants were included in the current analysis for mortality.

Selective reporting

None of the included trials published a study protocol. Four included studies were published only as abstracts (Da Silva 2002; DeMartini 1999; Hingre 1992; Ramanathan 1994); therefore, this item could not be assessed. All studies reported sufficiently on the predefined outcome parameters.

Other potential sources of bias

We judged most trials as having an unclear risk for other potential sources of bias, because the authors did not state in the manuscript if and how the studies were funded (Bloomfield 1998; Cummings 1989; Da Silva 2002; DeMartini 1999; Durand 2002; Marr 2019; Merz 1999; Odd 2004; Ramanathan 1994). Malloy 2005 was terminated prematurely; and in Halliday 2001a, a large proportion of the infants randomized to delayed selective treatment either died or did not fulfill the entry criteria. We judged the other trials to be at low risk.

Effects of interventions

See: Summary of findings 1 Lower compared to higher cumulative dose dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants; Summary of findings 2 Later compared to earlier initiation of dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants; Summary of findings 3 Pulse compared to continuous dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants; Summary of findings 4 Individual tailored compared to continuous tapered dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants

Comparison 1. Lower (experimental arm) versus higher (control arm) cumulative dosage regimens of dexamethasone

Primary outcome
Combined outcome of death or BPD at 36' weeks PMA

Meta‐analysis suggests that a lower compared with a higher cumulative corticosteroid dosage may not affect the risk of death or BPD at 36' weeks PMA (RR 1.03, 95% CI 0.86 to 1.24; P = 0.73; I²= 7%; 7 trials, 268 participants; Analysis 1.1). There was no evidence of a subgroup effect by 'moderate versus high cumulative dose regimen' (Chi2 = 1.09, df = 1 (P = 0.30), I2 = 8%). We graded the certainty of the evidence low using GRADE methods because of the small number of events, and the risk of selection, attrition and reporting bias (summary of findings Table 1). We were unable to assess potential publication bias in a funnel plot as fewer than 10 eligible RCTs reported this outcome. Furthermore, the planned subgroup or sensitivity analyses by gestational age, severity of illness, mode of ventilation, and use of open‐label corticosteroids were not possible because of paucity of available data. Meta‐analysis including only those trials without high risk of bias in any domains suggests that a lower compared with a higher cumulative corticosteroid dosage may not affect the risk of death or BPD at 36' weeks PMA (RR 1.09, 95% CI 0.93 to 1.29; I² = 41%; 4 trials, 142 participants).

Secondary outcomes
Mortality at 28 days' PNA, at 36 weeks' PMA and at hospital discharge.

No data were retrieved on mortality at 28 days' PNA. Meta‐analysis suggests that a lower compared with a higher cumulative corticosteroid dosage may not affect the risk of the outcome of death at 36 weeks' PMA, and at hospital discharge (RR 0.68, 95% CI 0.29 to 1.60; P = 0.38; I² = 0%; 7 trials, 268 participants; subgroup differences Chi2 = 1.03, df = 1 (P = 0.31), I2 = 2.9%; Analysis 1.2; and RR 0.93, 95% CI 0.48 to 1.81; P = 0.84; I² = 0%; 7 trials, 268 participants; subgroup differences Chi2 = 2.14, df = 1 (P = 0.14), I2 = 53.3%; Analysis 1.3, respectively). Subgroups for moderate‐ versus high‐ and low‐ versus moderate‐dosage regimens also showed no evidence of a difference.

BPD at 28 days' PNA and 36 weeks' PMA

No data were retrieved on the outcome of BPD at 28 days' PNA. Meta‐analysis suggests that a lower compared with a higher cumulative corticosteroid dosage may not affect the risk of the outcome BPD at 36 weeks' PMA (RR 1.12, 95% CI 0.91 to 1.37; P = 0.28; I² = 31%; 7 trials, 268 participants; subgroup differences Chi2 = 0.00, df = 1 (P = 0.99), I2 = 0%; Analysis 1.4). Subgroups moderate versus high and low versus moderate dosage regimen also showed no evidence of a difference.

Failure to extubate

Meta‐analysis suggests that a lower compared with a higher cumulative corticosteroid dosage may not affect the outcome of 'failure to extubate day three after initiation therapy' (RR 1.12, 95% CI 0.98 to 1.29; P = 0.09; I² = 0%; 5 trials, 209 participants; subgroup differences Chi2 = 0.04, df = 1 (P = 0.85), I2 = 0%; Analysis 1.5). However, compared to the infants allocated to the higher dosage regimens, the infants allocated to the lower‐dose regimen had a higher incidence of failing extubation at day seven after initiation of therapy (RR 1.32, 95% CI 1.08 to 1.60; P = 0.006; I²= 5%; NNTH 6, 95% CI 3 to 17; 5 trials, 210 participants; Analysis 1.6). Although there was no evidence of a subgroup difference between the high versus moderate dosage regimen compared with the moderate versus low dosage regimen (Chi2 = 1.03, df = 1 (P = 0.31), I2 = 3.2%), a larger effect was found in the higher dosage regimen subgroup (RR 1.49, 95% CI 1.10 to 2.02; P = 0.01; I²=54%; NNTH 4, 95% CI 2 to 12; 2 trials, 58 participants; Analysis 1.6.1). For the subgroup low versus moderate dosage regimen there was no evidence of a difference in failure to extubate seven days after initiation of therapy.

Days of mechanical ventilation and supplemental oxygen

Meta‐analysis suggests that a lower compared with a higher cumulative corticosteroid dosage may not affect the duration of mechanical ventilation (MD 4.50, 95% CI ‐0.68 to 9.67; P = 0.09; I² = 68%; 6 trials, 218 participants; Analysis 1.7). Although there was no evidence of a subgroup difference (Chi2 = 1.40, df = 1 (P = 0.24), I2 = 28.7%) in the high‐dosage regimen compared to the moderate‐dosage regimen, duration of mechanical ventilation was shorter (MD 8.09, 95% CI 0.21 to 15.96; P = 0.04; I² = 85%; 3 trials; 112 participants; Analysis 1.7.1). No evidence of a difference was seen in the outcome 'days of supplemental oxygen' (MD 0.30, 95% CI ‐20.14 to 20.74; P = 0.98; I² = 0%; 2 trials, 80 participants; subgroup differences Chi2 = 0.16, df = 1 (P = 0.69), I2 = 0%; Analysis 1.8).

Complications during primary hospitalization

Compared to the infants allocated to the higher‐dosage regimen, the infants allocated to the lower‐corticosteroid regimen showed a lower incidence of the short‐term adverse effect of hypertension (RR 0.31, 95% CI 0.12 to 0.77; P = 0.01; I² = 0%; NNTB 10, 95% CI 5.9 to 32; 6 trials, 240 participants; Analysis 1.9). Although there was no evidence of a subgroup difference between the high versus moderate dosage regimen and the moderate versus low dosage regimen (Chi2 = 0.01, df =1 (P = 0.91), I2 = 0%), a smaller confidence interval was found in the lower dosage regimen subgroup (RR 0.31, 95% CI 0.11 to 0.87; P = 0.03; I² = 0%; NNTH 7, 95% CI 4 to 30; 3 trials, 126 participants; Analysis 1.9.2).

For hyperglycemia a similar result was found for higher versus lower dosage regimen (RR 0.60, 95% CI 0.37 to 0.97; P = 0.04; I² = 13%; NNTB 10, 95% CI 4.8 to 113; 6 trials, 240 participants; Analysis 1.10). Although there was no evidence of a subgroup difference between the high versus moderate dosage regimen compared with the moderate versus low dosage regimen (Chi2 = 1.93, df =1 (P = 0.17), I2 = 48.1%), a larger effect was seen in the low versus moderate dosage regimen (RR 0.40, 95% CI 0.17 to 0.93; P = 0.03; I² = 0%; NNTH 7, 95% CI 4 to 41; 3 trials, 126 participants; Analysis 1.10.2). No differences were seen between the moderate and high dosage comparison.

There was no evidence of a difference between the different dosage regimens for the following outcomes:

  • incidence of late ‘rescue’ therapy with open‐label corticosteroids (RR 0.93, 95% CI 0.68 to 1.28; P = 0.66; I² = 10%; 7 trials, 268 participants; subgroup differences Chi2 = 2.50, df = 1 (P = 0.11), I2 = 59.9%; Analysis 1.11);

  • culture confirmed infection (RR 0.96, 95% CI 0.67 to 1.39; P = 0.72; I² = 0%; 7 trials, 289 participants; subgroup differences Chi2 = 1.31, df = 1 (P = 0.25), I2 = 23.7%; Analysis 1.12);

  • clinical suspected infection (RR 1.03, 95% CI 0.62 to 1.70; P = 0.44; I² = 0%; 3 trials, 131 participants; subgroup differences Chi2 = 0.52, df = 1 (P = 0.47), I2 = 0%; Analysis 1.13);

  • gastrointestinal hemorrhage (no events in either allocation arm; 3 trials, 101 participants), gastrointestinal perforation (RR 0.92, 95% CI 0.13 to 6.28; P = 0.96; I² = 0%; 4 trials, 185 participants; subgroup differences not applicable; Analysis 1.14);

  • NEC (RR 0.53, 95% CI 0.18 to 1.56; P = 0.57; I² = 0%; 4 trials, 198 participants; subgroup differences Chi2 = 1.05, df = 1 (P = 0.31), I2 = 4.9%; Analysis 1.15);

  • severe IVH (RR 1.68, 95% CI 0.65 to 4.37; P = 0.63; I² = 0%; 3 trials, 101 participants; subgroup differences Chi2 = 0.00, df = 1 (P = 1.00), I2 = 0%; Analysis 1.16);

  • PVL (RR 0.93, 95% CI 0.20 to 4.39; P = 0.92; I² = 0%; 2 trials, 121 participants; subgroup differences Chi2 = 0.01, df = 1 (P = 0.92), I2 = 0%; Analysis 1.17); or

  • severe ROP (RR 0.64, 95% CI 0.32 to 1.28; P = 0.83; I² = 0%; 5 trials, 176 participants; subgroup differences Chi2 = 0.02, df = 1 (P = 0.89), I2 = 0%; Analysis 1.18)

No data were retrieved on the outcomes PDA and cardiac hypertrophy.

Long‐term neurodevelopmental sequelae

Four trials reported the long‐term neurodevelopmental outcomes of cerebral palsy, visual impairment or the Bayley MDI in survivors, including 66% to 100% of their randomized infants. Malloy 2005 performed long‐term neurodevelopmental assessment, but used the modified Gesell Developmental Appraisal, which was deemed not to be comparable with the Bayley MDI reported in the other studies. Analysis showed a difference in the incidence of cerebral palsy. Compared to the infants allocated to the higher‐dosage regimens, the infants allocated to the lower‐dose regimen had a higher incidence of cerebral palsy assessed between one and three years of age (RR 2.64, 95% CI 1.02 to 6.83; P = 0.04; I² = 3%; NNTH 9, 95% CI 4.5 to 87; 4 trials, 149 participants; Analysis 1.19). Although there was no evidence of a subgroup difference for the higher versus lower dosage regimens comparisons (Chi2 = 2.91, df = 1 (P = 0.09), I2 = 65.7%), a larger effect was seen in the subgroup analysis of moderate‐dosage regimens versus high‐dosage regimens (RR 6.85, 95% CI 1.29 to 36.36; P = 0.02; I² = 0%; NNTH 5, 95% CI 2.6 to 12.7; 2 trials, 74 participants; Analysis 1.19.1).

Meta‐analysis suggests that a lower compared with a higher cumulative corticosteroid dosage may not affect the composite outcome of death or cerebral palsy of all trials (Analysis 1.20). However, there was evidence of a subgroup difference for higher versus lower dosage regimens (Chi2 = 4.25, df = 1 (P = 0.04), I2 = 76.5%). A larger effect was seen in the group of infants allocated to the high dosage regimen comparing a moderate dosage regimen (RR 3.20, 95% 1.35 to 7.58; P = 0.008; I² = 25%; NNTH 5, 95% CI 2.6 to 13.1; 2 trials, 84 participants; Analysis 1.20.1).

There were no differences in the number of infants with Bayley MDI less than ‐2 SD in the higher versus lower dosage comparison. However, there was evidence of a subgroup difference for higher versus lower dosage regimens (Chi2 = 4.26, df = 1 (P = 0.04), I2 = 76.5%). In participants treated with a moderate‐dosage regimen compared to a high‐dosage regimen, an increased risk was found on the Bayley MDI < ‐2 SD (RR 3.99, 95% CI 1.06 to 15.08; P = 0.04; I² = 0%; NNTB 6, 95% CI 3 to 27; 2 trials, 72 participants; Analysis 1.21.1).

Meta‐analysis suggests that a lower compared with a higher cumulative corticosteroid dosage may not affect the outcome visual impairment (RR 0.66, 95% CI 0.17 to 2.66; P = 0.69; I² = 0%; 5 trials, 166 participants; subgroup differences Chi2 = 1.47, df = 1 (P = 0.22), I2 = 32.1%; Analysis 1.22).

Three studies reported on the incidence of abnormal neurodevelopmental outcome as defined by the trialists. The meta‐analyses of low‐ versus moderate‐dosage regimens did not reveal any differences. However, there was evidence of a subgroup difference for higher versus lower dosage regimens (Chi2 = 6.39, df = 1 (P = 0.01), I2 = 84.3%). Compared to the infants allocated to a high‐dosage regimen, a higher incidence of abnormal neurodevelopmental outcome was seen in the group of infants allocated to the moderate‐dosage regimen (RR 7.60, 95% CI 1.45 to 39.78; P = 0.02; I² = 0%; NNTH 4, 95% CI 2.4 to 9.8; 2 trials, 74 participants; Analysis 1.23.1). Meta‐analysis suggests that a lower compared with a higher cumulative corticosteroid dosage may not affect the composite outcome of abnormal neurodevelopmental outcome or death assessed between one and three years of age (RR 1.86, 95% CI 0.98 to 3.53; P = 0.06; I² = 73%; 3 trials, 100 participants; Analysis 1.24). However, there was evidence of a subgroup difference for higher versus lower dosage regimens (Chi2 = 7.12, df = 1 (P = 0.008), I2 = 85.9%). The composite outcome of abnormal neurodevelopmental outcome or death showed the same benefits in favor of the high‐dosage group (RR 3.41, 95% CI 1.44 to 8.07; P = 0.005; I² = 41%; NNTH 4, 95% CI 2.2 to 10.4; 2 trials, 84 participants; moderate certainty; Analysis 1.24.1).

We graded the certainty of the evidence for the composite outcomes death or CP, and death or abnormal neurodevelopmental outcome as very low because of the inconsistency across studies, the small number of events, and the risk of performance, detection and attrition bias (summary of findings Table 1). Cummings 1989 and Marr 2019) also reported the neurodevelopmental outcome at age seven years and 15 years, respectively. No meta‐analysis was possible given the single study results for each time of assessment, but these individual studies showed similar results in the reduced risk of neurodevelopmental impairment in the group of infants allocated to high dosage regimen, compared to the infants treated with moderate dosage regimens.

Comparison 2. Later (experimental arm) versus earlier (control arm) initiation of dexamethasone

Primary outcome
Combined outcome of death or BPD at 36 weeks' PMA

Meta‐analysis suggests that a later compared with an earlier cumulative corticosteroid dosage may not affect the combined outcome of death or BPD at 36 weeks' PMA (RR 1.06, 95% CI 0.87 to 1.29; P = 0.57; I² = 17%; 3 trials, 391 participants; subgroup differences not applicable; Analysis 2.1). We graded the certainty of the evidence as low because of the small number of events, and the risk of performance and detection bias in all three trials and unclear selection bias in one trial (summary of findings Table 2). We were unable to assess potential publication bias in a funnel plot as fewer than 10 eligible RCTs reported this outcome. We were unable to undertake the planned subgroup or sensitivity analyses by gestational age, severity of illness, mode of ventilation, and use of open‐label corticosteroids, because of paucity of available data. We were not able to perform a meta‐analysis including only those trials without high risk of bias in any domains, because all were judged to have a high risk of bias.

Secondary outcomes
Mortality at 28 days' PNA, 36 weeks' PMA and at hospital discharge

Meta‐analysis suggests that a later compared with an earlier cumulative corticosteroid dosage may not affect mortality at 28 days' PNA (RR 1.01, 95% CI 0.69 to 1.47; P = 0.97; I² = 59%; 4 trials, 762 participants; subgroup differences Chi2 = 4.42, df = 1 (P = 0.04), I2 = 77.4%; Analysis 2.2), 36 weeks' PMA (RR 0.93, 95% CI 0.68 to 1.28; P = 0.66; I² = 54%; 4 trials, 762 participants; subgroup differences Chi2 = 3.98, df = 1 (P = 0.05), I2 = 74.9%; Analysis 2.3) and mortality at hospital discharge (RR 1.00, 95% CI 0.75 to 1.33; P = 0.30; I² = 18%; 5 trials, 797 participants; subgroup differences Chi2 = 2.67, df = 1 (P = 0.10), I2 = 62.5%; Analysis 2.4).

BPD at 28 days' PNA and 36 weeks' PMA

Compared to the infants who were allocated to earlier initiation, the infants allocated to later initiation had a higher incidence of the outcome BPD at 28 days' PNA (RR 1.12, 95% CI 1.02 to 1.23; P = 0.02; I² = 49%; NNTH 14, 95% CI 7 to 90; 4 trials, 762 participants; Analysis 2.5). Although there was no evidence of a subgroup difference between the earlier versus later initiation regimens (Chi2 = 0.33, df =1 (P = 0.56), I2 = 0%), a larger effect was found in the late versus moderate early initiation regimen subgroup. Compared to the infants who were allocated to moderately early initiation, the infants allocated to delayed initiation had a higher incidence of the outcome BPD at 28 days' PNA (RR 1.15, 95% CI 1.05 to 1.26; P = 0.004; I² = not applicable; NNTH 9, 95% CI 5 to 26; 1 trial, 371 participants; Analysis 2.5.1). Later versus earlier initiation of corticosteroids showed no effect on the outcome BPD at 36 weeks' PMA. Although there was no evidence of a subgroup difference between the earlier versus later initiation regimens (Chi2 = 3.10, df =1 (P = 0.08), I2 = 67.7%), a larger effect was found in the moderate early versus early initiation regimen subgroup. Compared to the infants allocated in the early administration, the infants who were allocated in the moderately early group had a higher incidence of BPD at 36 weeks' PMA (RR 1.38, 95% CI 1.01 to 1.90; P = 0.05; I² = 17%; NNTH 11, 95% CI 6 to 333; 3 trials, 391 participants; Analysis 2.6.2).

Failure to extubate

Compared to late initiation, moderately early initiation resulted in a reduction in the number of infants failing extubation at day three and day seven in the only trial reporting this outcome (RR 1.10, 95% CI 1.05 to 1.15; P < 0.0001; I² = not applicable; 1 trial, 371 participants; subgroup differences not applicable; Analysis 2.7; RR 1.22, 95% CI 1.14 to 1.32; P < 0.00001; I² = not applicable; 1 trial, 378 participants; subgroup differences not applicable; Analysis 2.8).

Days of mechanical ventilation and supplemental oxygen

The trials publishing data on the duration of mechanical ventilation showed a difference between early administration and moderately early administration (MD 12.71, 95% CI 4.44 to 20.99; P = 0.003; I² = 0%; 2 trials, 60 participants; subgroup differences not applicable; Analysis 2.9), and the only trial reporting the outcome supplemental days of oxygen showed that compared to moderately early initiation, early initiation of dexamethasone results in a reduction in days on supplemental oxygen (RR 29.00, 95% CI 7.10 to 50.90; P = 0.009; I² = not applicable; 1 trial, 30 participants; subgroup differences not applicable; Analysis 2.10).

Complications during primary hospitalization

Meta‐analysis suggests that a later compared with an earlier cumulative corticosteroid dosage may not affect the incidence of hypertension (RR 0.99, 95% CI 0.67 to 1.47; P = 0.98; I² = 34%; 4 trials, 762 participants; subgroup differences Chi2 = 1.52, df = 1 (P = 0.22), I2 = 34.2%; Analysis 2.11).

Compared to the infants allocated to the earlier administration arm, the infants allocated to the later initiation arm had a lower incidence of hyperglycemia (RR 0.66, 95% CI 0.53 to 0.82; P = 0.0001; I² = 21%; NNTB 8, 95% CI 5.0 to 15.7; 4 trials, 726 participants; Analysis 2.12). No subgroup differences were seen (Chi2 = 0.00, df = 1 (P = 1.00), I2 = 0%). These effects were seen both in the subgroup late versus moderately early, and moderately early versus early initiation of therapy (RR 0.66, 95% CI 0.46 to 0.95; P = 0.03; I² = not applicable; NNTB 10, 95% CI 5 to 73; 1 trial, 371 participants; Analysis 2.12.1 and RR 0.66, 95% CI 0.51 to 0.85; P = 0.002; I² = 47%; NNTB 6, 95% CI 4 to 15; 3 trials, 355 participants; Analysis 2.12.2, respectively).

The incidence of ‘rescue’ therapy with open‐label corticosteroids was higher in the group of infants allocated to late initiation (RR 1.71, 95% CI 1.04 to 2.81; P = 0.03; NNTH 25, 95% CI 12.50 to 462.4; I² = 66%; 3 trials, 732 participants; subgroup differences Chi2 = 2.67, df = 1 (P = 0.10), I2 = 62.6%; Analysis 2.13). Overall, no evidence of a difference was seen in the outcome culture‐proven infection between later and earlier initiation of therapy (Analysis 2.14). However, there was some evidence of a subgroup difference for earlier versus later initiation of therapy in both outcomes (Chi2 = 6.39, df = 1 (P = 0.01), I2 = 84.3%). Compared to the infants allocated to the moderately early dexamethasone initiation, the infants allocated to late initiation showed a lower incidence of the outcomes of culture‐proven infection (RR 0.67, 95% CI 0.54 to 0.84; P = 0.0005; I² = not applicable; NNTH 6, 95% CI 3.50 to 12.00; 1 trial, 175 participants; Analysis 2.14.1). No data were reported on the outcome clinically suspected infection.

Furthermore, no evidence of a difference was seen in the outcome gastrointestinal hemorrhage between later and earlier initiation of therapy (Analysis 2.15). Although no difference was found between the subgroups (Chi2 = 0.57, df = 1 (P = 0.45), I2 = 0%), a higher incidence of gastrointestinal hemorrhage was found in the subgroup comparing late versus moderate early initiation of therapy (RR 0.60, 95% CI 0.38 to 0.95; P = 0.04; I² = 0%; NNTH 12, 95% CI 6.0 to 98.5; 4 trials, 762 participants; Analysis 2.15.1). Meta‐analysis suggests that a later compared with an earlier cumulative corticosteroid dosage may not affect for the outcomes gastrointestinal perforation (RR 0.75, 95% CI 0.23 to 2.40; P = 0.63; I² = not applicable; 2 trials, 315 participants; subgroup differences not applicable; Analysis 2.16), and NEC (RR 1.44, 95% CI 0.82 to 2.55; P = 0.21; I² = 17%; 4 trials, 725 participants; subgroup differences Chi2 = 0.16, df = 1 (P = 0.69), I2 = 0%; Analysis 2.17).

Compared to the infants allocated to the early initiation group, the infants allocated to the moderately early initiation arm had an increased risk of a PDA requiring therapy (RR 1.74, 95% CI 1.32 to 2.29; P < 0.0001; I² = not applicable; NNTH 5, 95% CI 2.80 to 7.60; 1 trial, 285 participants; Analysis 2.18). Meta‐analysis suggests that a later compared with an earlier cumulative corticosteroid dosage may not affect the outcomes IVH (RR 2.37, 95% CI 0.49 to 11.48; P = 0.28; I² = not applicable; 1 trial, 76 participants; subgroup differences not applicable; Analysis 2.19); ROP any grade (RR 0.80, 95% CI 0.52 to 1.23; P = 0.31; I² = 0%; 2 trials, 324 participants; subgroup differences not applicable; Analysis 2.20); or severe ROP (RR 1.50, 95% CI 0.63 to 3.53; P = 0.58; I² = 0%; 3 trials, 391 participants; subgroup differences not applicable; Analysis 2.21). No data were reported on the outcome PVL.

Long‐term neurodevelopmental sequelae

Two trials investigating moderately early versus early initiation of dexamethasone reported long‐term neurodevelopmental outcomes using various definitions. No data were reported on the Mental Developmental Index of the Bayley Scales of Infant Development in these trials. Meta‐analysis showed no evidence of a difference in the incidence of cerebral palsy in survivors assessed between both allocation arms (RR 1.95, 95% CI 0.43 to 8.86; P = 0.39; I² = not applicable; 1 trial, 61 participants; subgroup differences not applicable; Analysis 2.22), and the composite outcome death or cerebral palsy (RR 1.12, 95% CI 0.68 to 1.84; P = 0.65; I² = not applicable; 1 trial, 86 participants; subgroup differences not applicable; Analysis 2.23). None of the infants had severe blindness, regardless of the allocation group (1 trial, 61 participants). Analyses showed no evidence of a difference in the incidence of abnormal neurodevelopmental outcome in survivors assessed between both allocation arms (RR 1.06, 95% CI 0.66 to 1.69; P = 0.82; I² = 0%; 2 trials, 155 participants; subgroup differences not applicable; Analysis 2.24), or the composite outcome of death or long‐term neurodevelopmental outcome (RR 0.87, 95% CI 0.63 to 1.21; P = 0.42; I² = 0%; 2 trials, 167 participants; subgroup differences not applicable; Analysis 2.25). The certainty of the evidence was low because of the small number of events, and the risk of performance and detection bias and unclear selection bias (summary of findings Table 2).

Comparison 3. Pulse therapy (experimental arm) versus continuous tapered (control arm) dosage regimens of dexamethasone

Primary outcome
Combined outcome death or BPD at 36 weeks' PMA

Compared to the infants allocated to the continuous tapered dosage regimen, the infants allocated to pulse therapy showed an increase in the incidence of the combined outcome of death or BPD at 36 weeks' PMA (RR 1.38, 95% CI 1.02 to 1.88; P = 0.04; I² = not applicable; NNTH 7, 95% CI 4 to 155; 1 trial, 197 participants; Analysis 3.1). We graded the certainty of the evidence as low because of the small number of events, and the risk of performance and detection bias in one trial (summary of findings Table 3). We were unable to assess potential publication bias in a funnel plot as fewer than 10 eligible RCTs reported this outcome. Furthermore, the planned subgroup or sensitivity analyses by gestational age, severity of illness, mode of ventilation, and use of open‐label corticosteroids were not possible because of paucity of available data. Meta‐analysis including the only trial without high risk of bias in any domains suggests that a pulse compared with a continuous tapered corticosteroid regimen may not affect the risk of death or BPD at 36' weeks PMA (RR 1.42, 95% CI 0.99 to 2.05; I² not applicable; 1 trials, 121 participants).

Secondary outcomes
Mortality at 28 days, 36 weeks' PMA and at hospital discharge

Meta‐analysis suggests that a pulse compared with a continuous tapered corticosteroid regimen may not affect the outcome of mortality at any time point (RR 2.85, 95% CI 0.31 to 26.15; P = 0.36; I² = not applicable; 1 trial, 76 participants; Analysis 3.2; RR 2.04, 95% CI 0.72 to 5.78; P = 0.18; I² = 0%; 2 trials, 197 participants; Analysis 3.3; RR 2.04, 95% CI 0.72 to 5.78; P = 0.18; I² = 0%; 2 trials, 197 participants; Analysis 3.4, respectively).

BPD at 28 days' PNA and at 36 weeks' PMA

Meta‐analysis suggests that a pulse compared with a continuous tapered corticosteroid regimen may not affect the outcomes of BPD at 28 days' PNA or 36 weeks' PMA (RR 1.40, 95% CI 0.92 to 2.13; P = 0.12; I² = not applicable; 1 trial, 76 participants; Analysis 3.5; RR 1.29, 95% CI 0.90 to 1.83; P = 0.16; I² = 0%; 2 trials, 197 participants; Analysis 3.6, respectively).

Failure to extubate

No data could be retrieved on the outcomes of failure to extubate.

Days of mechanical ventilation and supplemental oxygen

No data could be retrieved on the days of mechanical ventilation or supplemental oxygen.

Complications during primary hospitalization

No evidence of a difference between the two allocation arms was found for the outcomes hypertension (RR 0.50, 95% CI 0.20 to 1.23; P = 0.13; I² = not applicable; 2 trials, 197 participants; Analysis 3.7), or hyperglycemia (RR 1.08, 95% CI 0.71 to 1.65; P = 0.73; I² = 31%; 2 trials, 160 participants; Analysis 3.8). The use of open‐label corticosteroids was similar in both groups in the trial providing this information (RR 0.97, 95% CI 0.64 to 1.47; P = 0.87; I² = not applicable; 2 trials, 197 participants; Analysis 3.9). No evidence of a difference was seen for the following outcomes:

  • culture‐proven infection (RR 1.32, 95% CI 0.87 to 2.01; P = 0.19; I² = not applicable; 1 trial, 121 participants; Analysis 3.10);

  • clinically suspected infection (RR 1.21, 95% CI 0.70 to 2.10; P = 0.49; I² = not applicable; 1 trial, 121 participants; Analysis 3.11);

  • gastrointestinal hemorrhage (RR 0.65, 95% CI 0.25 to 1.68; P = 0.38; I² = not applicable; 2 trials, 197 participants; Analysis 3.12);

  • NEC (RR 0.78, 95% CI 0.33 to 1.83; P = 0.57; I² = 60%; 2 trials, 160 participants; Analysis 3.13);

  • IVH above grade II (RR 2.37, 95% CI 0.49 to 11.48; P = 0.28; I² = not applicable; 1 trial, 76 participants; Analysis 3.14);

  • ROP (RR 0.53, 95% CI 0.05 to 5.34; P = 0.59; I² = not applicable; 1 trial, 39 participants; Analysis 3.15); and

  • severe ROP (RR 0.24, 95% CI 0.05 to 1.07; P = 0.06; I² = not applicable; 1 trial, 121 participants; Analysis 3.16).

Long‐term neurodevelopmental sequelae

Follow‐up was only performed in one trial. No data were reported on Bayley Scales of Infant Development or cerebral palsy outcomes in this trial. No evidence of a difference was found for the outcome abnormal neurodevelopmental outcome alone (RR 0.88, 95% CI 0.54 to 1.44; P = 0.62; I² = not applicable; 1 trial, 64 participants; Analysis 3.17), or combined with death (RR 1.23, 95% CI 0.79 to 1.92; P = 0.37; I² = not applicable; 1 trial, 76 participants; Analysis 3.18). We graded the certainty of the evidence as very low because of the small number of events, and the risk of performance and detection bias in one trial and potential publication bias of one trial (summary of findings Table 3).

Comparison 4. Individual tailored (experimental arm) versus continuous tapered (control arm) dosage regimens of dexamethasone

Primary outcome
Combined outcome death or BPD at 36 weeks' PMA

Meta‐analysis suggests that an individual tailored regimen compared with a continuous tapered corticosteroid regimen may not affect the outcome of combined death or BPD at 36 weeks' PMA (RR 1.06, 95% CI 0.88 to 1.29; P = 0.54; I² = 10%; 3 trials, 168 participants; Analysis 4.1). We graded the certainty of the evidence as low because of the small number of events, and the risk of performance and detection bias (summary of findings Table 4). We were unable to assess potential publication bias in a funnel plot as fewer than 10 eligible RCTs reported this outcome. We were unable to undertake the planned subgroup or sensitivity analyses by gestational age, severity of illness, mode of ventilation, use of open‐label corticosteroids, because of paucity of available data. Meta‐analysis including the only trial without high risk of bias in any domains suggests that an individual tailored regimen compared with a continuous tapered corticosteroid regimen may not affect the risk of death or BPD at 36' weeks PMA (RR 0.96, 95% CI 0.82 to 1.12; I² = not applicable; 1 trial, 59 participants).

Secondary outcomes
Mortality at 28 days' PNA, 36 weeks' PMA and at hospital discharge

Meta‐analysis suggests that an individual tailored regimen compared with a continuous tapered corticosteroid regimen may not affect mortality at 28 days' PNA and 36 weeks' PMA (RR 2.83, 95% CI 0.60 to 13.32; P = 0.19; I² = 0%; 2 trials, 109 participants; Analysis 4.2; RR 1.54, 95% CI 0.63 to 3.79; P = 0.35; I² = 0%; 3 trials, 168 participants; Analysis 4.3; RR 1.84, 95% CI 0.77 to 4.37; P = 0.17; I² = 0%; 3 trials, 168 participants; Analysis 4.4, respectively).

BPD at 28 days' PNA and 36 weeks' PMA

Meta‐analysis suggests that an individual tailored compared with a continuous tapered corticosteroid regimen may not affect the outcome BPD at 28 days' PNA or 36 weeks' PMA (RR 1.15, 95% CI 0.88 to 1.50; P = 0.30; I² = 81%; 2 trials, 109 participants; Analysis 4.5; RR 0.99, 95% CI 0.79 to 1.25; P = 0.96; I² = 0%; 3 trials, 168 participants; Analysis 4.6, respectively).

Failure to extubate

Meta‐analysis suggests that an individual tailored compared with a continuous tapered corticosteroid regimen may not affect the outcomes of failure to extubate three days after initiation of therapy (RR 1.16, 95% CI 0.95 to 1.43; P = 0.15; I² = not applicable; 1 trial, 59 participants; Analysis 4.7). Compared to the infants allocated to the continuous tapered regimen, the infants who were allocated to the individual tailored dosage regimen had an increased risk of failure to extubate seven days after initiation of therapy (RR 1.72, 95% CI 1.17 to 2.54, NNTH 3, 95% CI 2 to 7; P = 0.006; I² = not applicable; 1 trial, 59 participants; Analysis 4.8).

Days of mechanical ventilation and supplemental oxygen

Compared to the infants allocated to the continuous tapered regimen, the infants who were allocated to the individual tailored dosage regimen had a decreased duration of mechanical ventilation (MD 9.26, 95% CI 4.32 to 14.21; I2 = 69%; 2 studies, 90 participants; Analysis 4.9). Meta‐analysis suggests that an individual tailored regimen compared with a continuous tapered corticosteroid regimen may not affect the days of supplemental oxygen (MD 8.00, 95% CI ‐34.64 to 50.64; P = 0.71; I² = not applicable; 1 trial, 52 participants; Analysis 4.10).

Complications during primary hospitalization

Meta‐analysis suggests that an individual tailored compared with a continuous tapered corticosteroid regimen may not affect the following outcomes:

  • hypertension (RR 0.34, 95% CI 0.01 to 8.13; P = 0.51; I² = not applicable; 2 trials, 135 participants; Analysis 4.11);

  • hyperglycemia (RR 0.66, 95% CI 0.26 to 1.66; P = 0.37; I² = not applicable; 2 trials, 98 participants; Analysis 4.12);

  • open‐label corticosteroids (RR 1.72, 95% CI 0.72 to 4.13; P = 0.22; I² = not applicable; 2 trials, 135 participants; Analysis 4.13);

  • culture‐proven infection (RR 1.27, 95% CI 0.58 to 2.76; P = 0.55; I² = not applicable; 2 trials, 92 participants; Analysis 4.14);

  • clinically suspected infection (no events in either allocation arms; 1 trial, 59 participants), gastrointestinal hemorrhage and perforation (no events in either allocation arms; 2 trials, 135 participants; no events in either allocation arms; 1 trial, 59 participants, respectively);

  • NEC (RR 1.75, 95% CI 0.48 to 6.35; P = 0.39; I² = not applicable; 2 trials, 98 participants; Analysis 4.15);

  • IVH above grade II (RR 2.00, 95% CI 0.78 to 5.18; P = 0.15; I² = 0%; 3 trials, 168 participants; Analysis 4.16);

  • PVL (RR 1.03, 95% CI 0.07 to 15.77; P = 0.98; I² = not applicable; 1 trial, 59 participants; Analysis 4.17); and

  • ROP (RR 0.83, 95% CI 0.24 to 2.92; P = 0.77; I² = 0%; 2 trials, 98 participants; Analysis 4.18).

Long‐term neurodevelopmental sequelae

The included trials reporting in this comparison did not show any evidence of difference in the following outcomes:

  • abnormal neurodevelopmental outcome, defined as either the presence of severe cerebral palsy alone (RR 4.62, 95% CI 0.55 to 38.74; P = 0.16; I² = not applicable; 1 trial, 56 participants; Analysis 4.19), or in combination with death (RR 7.24, 95% CI 0.95 to 55.26; P = 0.06; I² = not applicable; 1 trial, 59 participants; Analysis 4.20);

  • a Bayley mental score greater than 2 SD below the mean (RR 2.69, 95% CI 0.57 to 12.69; P = 0.21; I² = not applicable; 1 trial, 54 participants; Analysis 4.21);

  • bilateral blindness (RR 3.44, 95% CI 0.15 to 81.09; P = 0.44; I² = not applicable; 1 trial, 56 participants; Analysis 4.22); or

  • abnormal neurodevelopmental outcome in survivors alone (RR 1.09, 95% CI 0.71 to 1.70; P = 0.69; I² = 39%; 3 trials, 143 participants; Analysis 4.23), or in combination with death (RR 1.44, 95% CI 0.99 to 2.07; P = 0.05; I² = 52%; 3 trials, 168 participants; Analysis 4.24)

The certainty of the evidence was moderate for the outcome death or cerebral palsy and low for the outcome death or abnormal neurodevelopmental outcome because of the small number of events, and the risk of performance and detection bias (summary of findings Table 4).

Discussion

It has been proven in RCTs that corticosteroids reduce the combined outcome of death or BPD at 36 weeks' PMA. However, concerns have arisen about negative long‐term neurodevelopmental effects of this therapy. Despite the firm recommendations of several pediatric societies to stop using postnatal systemic dexamethasone outside the realm of randomized clinical trials, clinicians are still using dexamethasone to treat ventilator‐dependent preterm infants. Therefore, attempts to identify the optimal corticosteroid treatment regimen remain clinically relevant and important. Questions that need to be answered are: 1) what is the optimal time to start corticosteroid treatment; 2) what is optimal cumulative dose; 3) what is the optimal duration of therapy; 4) what is the optimal corticosteroid to use? This systematic review summarizes all published studies that have investigated the comparative effects of different corticosteroid treatment regimens head‐to‐head on the incidence of the combined outcome of death or BPD and the risk of adverse effects on neurodevelopment.

Summary of main results

This review examines four different types of postnatal corticosteroid regimens. The first intervention, investigated in eight RCTs (n = 306) compares a lower versus a higher dose of dexamethasone. The absolute dexamethasone dose used to contrast a higher versus a lower dose varied considerably between the included trials. This heterogeneity in dose contrast precluded a pooled analysis of all available trials. For this reason, we divided the studies into a high‐range contrast subgroup, comparing a high cumulative dose (> 4 mg/kg) to a moderate dose (2 to 4 mg/kg) and a low‐range contrast subgroup, comparing a moderate to a low cumulative dose (< 2 mg/kg). We would like to emphasize that the terms 'high', 'moderate', and 'low' should be interpreted from a relative perspective, because compared to the physiological levels of corticosteroids all reported doses are supraphysiological (i.e. 'high'). The analyses showed no evidence of outcome differences when contrasting a low to a moderate dexamethasone dose, except the analyses of the outcomes hypertension and hyperglycemia. However, these adverse effects seen in the group of infants treated with a moderate dosage regimen might be outweighed by beneficial effect on long term outcomes. The analyses also showed no evidence of a difference in the outcome BPD at 36 weeks' PMA, and the combined outcome of death or BPD at 36 weeks' PMA when contrasting a moderate to a high dexamethasone dose. However, compared to a moderate dose, a high dexamethasone dose reduced the risk of failure to extubate, prolonged duration of mechanical ventilation, cerebral palsy or abnormal neurodevelopmental outcome assessed between one and three years of age, and the composite outcome death or cerebral palsy or abnormal neurodevelopmental outcome.

In contrast with a previous meta‐analysis assessing the impact of (different) cumulative dexamethasone doses used in placebo‐controlled trials (Onland 2009), no evidence of a difference was found for the primary outcome death or BPD at 36 weeks' PMA in the meta‐analyses of studies investigating a higher versus lower dosage regimen. This lack of evidence of a difference changed compared with the previous review published 2017 (Onland 2017). However, the current review was able to include the full publication of the trial conducted by Marr 2019. No differences were found in that trial comparing a moderate versus high dosage regimen of dexamethasone, hence the effect estimate changed from borderline significant (RR 1.35, 95% CI 1.00 to 1.82) in the previous version of this review towards no significant difference in the current version (Analysis 1.1). We can only speculate on the possible explanations for this change in findings compared to the previous version of this review (Onland 2017). First, the a priori risk of BPD might have been different between the different studies and comparisons. One of the studies in the high‐range contrast comparison was performed in the pre‐surfactant era (Cummings 1989), and another study in this comparison included infants with a quite low birth weight and gestational age with considerable supplemental oxygen and mean airway pressure at trial entry (Marr 2019). Both factors are known BPD risk factors and might explain the statistical heterogeneity found in these analyses. Second, the use of additional ('rescue') dexamethasone treatment outside the study protocol by infants in both allocation arms was only observed in the studies comparing a low to moderate cumulative dose and the Marr 2019 study. This could well have resulted in an underestimation of the true treatment effect in these trials (Onland 2010). Finally, these results may also suggest that a relatively low cumulative dexamethasone dose as used in the low‐range contrast comparison is, in a pharmacodynamic sense, not sufficient to change the rate of BPD and hence any contrast in this dosing range will not result in a group difference in BPD.

In contrast with the analyses of the primary outcome showing no difference, the short term pulmonary outcomes, namely duration of mechanical ventilation and failure to extubate seven days after initiation of therapy in the studies investigating a moderate versus a high dosage regimen did show a difference between groups in favor of the high‐dosage regimens. Furthermore, this review suggests that the benefit of high‐dose dexamethasone on these short‐term pulmonary outcomes is not outweighed by an increased risk of neurodevelopmental impairment. It even suggests that, compared to a moderate cumulative dose, neurodevelopment might be improved in the infants treated with a high dose, although this finding should be interpreted cautiously for the following reasons. First, the low a priori chance of adverse neurodevelopmental outcomes in combination with the relatively small number of included infants in this review might not be sufficient to detect small but clinically relevant treatment effects on these outcomes. Second, the number of infants lost to follow‐up was more than 10% in two of the three studies, which might have biased the results, since children with cerebral palsy are especially difficult to follow up. A possible benefit of high‐dose dexamethasone on neurodevelopmental outcome might be mediated by the reduced duration of mechanical ventilation. This outcome is associated with an increased risk of neurodevelopmental impairment and may, in the high‐risk infant, override a possible direct toxic effect of dexamethasone on the brain (Doyle 2014; Vliegenthart 2017; Vliegenthart 2019; Walsh 2005). In line with the aforementioned results of these meta‐analyses and reasoning, a recent network analysis including 62 RCTs evaluating 14 different inhaled and systemic administered corticosteroid regimens showed that moderately early initiation of a systemic moderate‐dosage regimen of dexamethasone might be the most appropriate treatment for the prevention of BPD and mortality (Ramaswamy 2021). However, more high‐quality evidence is needed to support or refute this hypothesis.

The second intervention, contrasting a later versus earlier initiation of therapy, showed conflicting results. The subgroup analyses comparing trials that started corticosteroids within the first week to trials starting after the first week of life showed a reduction of ventilation days and supplemental oxygen, as well as a decreased risk of BPD when treatment was initiated earlier. This beneficial effect of early treatment did not come at the expense of an increased risk of adverse neurodevelopmental outcome, as reported in the meta‐analysis of placebo‐controlled trials starting dexamethasone in the first week of life (Doyle 2021a). However, it is important to emphasize that only three studies performed a head‐to‐head comparison of moderately early versus early dexamethasone treatment reporting these outcomes, and included a small number of participants.

Analyses of primary comparisons including trials investigating late‐initiated dexamethasone versus initiation in the moderately early period revealed no benefits on long‐term pulmonary outcomes. Although postponing the start of dexamethasone treatment did reduce the risk of hypertension and culture‐proven sepsis, data on long‐term neurodevelopmental outcomes were not reported. These results are in contrast with the meta‐analyses of the placebo‐controlled trials, showing a lower number needed to treat for an additional beneficial outcome (NNTB) for reducing BPD when starting treatment moderately early compared to delayed administration (Onland 2009; Schmidt 2008).

The third intervention summarized in this review involved studies exploring the effect of a pulse‐dosing regimen on both the beneficial and adverse effects of dexamethasone treatment. These analyses showed that a pulse‐dosing regimen increased the risk of the combined outcome death or BPD compared with a continuous‐dosing regimen. Although speculative, it might be that the ongoing inflammatory response causing the development of BPD will not be suppressed by a pulse therapy regimen, which incorporated a seven‐day treatment pause.

Finally, tailoring the dexamethasone dose to the individual pulmonary response of the infant seems a logical approach, since there is a wide spectrum of lung damage in preterm infants. More inflamed and damaged lungs could theoretically benefit from a higher cumulative corticosteroid dose. To date, only three trials including a small number of infants have investigated this contrast, with no difference in the primary or secondary outcomes.

Overall completeness and applicability of evidence

We were not able to perform funnel plot analyses of the primary outcomes to identify potential publication bias, because less then 10 RCTs were identified per comparison. Therefore, we cannot rule out that other small RCTs were performed, but not published. Several studies were only published as abstracts, limiting methodological assessment and data on the primary and secondary outcomes. Another major problem that this review uncovered is that even when full text was published, not every trial reported on our stated primary and secondary outcomes. Specifically, few studies reported on neurodevelopmental outcome parameters, and those that did used various definitions or assessed neurodevelopment at different points in time. Although we pooled the data as if they were statistically homogeneous, this clinical heterogeneity might compromise the validity of the results of our meta‐analysis. It remains unclear how this influences the conclusions of this review. We could not perform the previously mentioned subgroup analyses, i.e. according to gestational age and respiratory status at trial entry, due to lack of data or heterogeneity between the trials on these clinical characteristics.

Quality of the evidence

Except for the trials that were only published as abstracts, and for which assessment of potential biases was not possible, we deemed the risk of bias in the trials as unclear to very low and we believe that bias has no large influence on the results. However, the overall quality of the evidence provided by the meta‐analyses using the GRADE approach for each outcome was moderate to very low due to several severe study limitations, such as risk of bias; potential publication bias; and imprecision of effect estimates. First, as mentioned earlier, the overall sample size in these analyses was small, resulting in inadequate power to detect small but clinically relevant differences in some of the important outcome parameters. Second, although most studies contrasted two dosing regimens of dexamethasone, there was considerable diversity in the study designs, like the cumulative dexamethasone dose that was used, the starting dose and the duration of therapy. It remains unclear whether and how these differences affect the observed treatment effect in the different interventions. Third, the use of late 'rescue' corticosteroids outside the study protocol was considerable in the majority of the trials; this may have confounded (contaminated) the true dexamethasone treatment effect.

Potential biases in the review process

The 2021 search did not include independent searches of EMBASE, ClinicalTrials.gov, or the World Health Organization’s International Clinical Trials Registry Platform (ICTRP). Although records from these sources are included in Cochrane CENTRAL, their omission may have reduced sensitivity of the search. Subsequent updates of this review will include these sources to minimize potential bias.

Agreements and disagreements with other studies or reviews

The first systematic review investigating the effect of different dosage regimens on the outcome BPD was published in 2008 (Onland 2008). The conclusion on the pulmonary outcome remains unchanged. However, that review did not include the abstract of Marr 2011, relating to the now‐published Marr 2019 study. Including the data from that study into the 2017 version of this review (Onland 2017) changed the cumulative effect estimate of long‐term neurodevelopmental outcome, showing a significantly reduced risk when administrating a higher‐dosage regimen. Since the Onland 2017 version of this review, the full publication of the study by Marr became available (Marr 2019), and the effect estimate of the outcome death or BPD showed no difference comparing a high‐ versus moderate‐dosage regimen of dexamethasone (Analysis 1.1). However, although the evidence is very uncertain, the meta‐analyses showing the long‐term neurodevelopmental outcomes suggest a beneficial effect in favor of higher dosage regimens including this trial (Analysis 1.20; Analysis 1.24), compared to a moderate‐dosage regimen of dexamethasone. Therefore, the results of the current review do not support the recommendation of international guidelines proclaiming that steroids should be dosed as short and low as possible (AAP 2002; Watterberg 2010).

Screen4Me 13 September 2022

Figures and Tables -
Figure 1

Screen4Me 13 September 2022

PRISMA flow chart

Figures and Tables -
Figure 2

PRISMA flow chart

Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included studies.

Figures and Tables -
Figure 3

Risk of bias graph: review authors' judgments about each risk of bias item presented as percentages across all included studies.

Risk of bias summary: review authors' judgments about each risk of bias item for each included study.

Figures and Tables -
Figure 4

Risk of bias summary: review authors' judgments about each risk of bias item for each included study.

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 1: Death or bronchopulmonary dysplasia at 36 weeks PMA

Figures and Tables -
Analysis 1.1

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 1: Death or bronchopulmonary dysplasia at 36 weeks PMA

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 2: Mortality at 36 weeks' PMA

Figures and Tables -
Analysis 1.2

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 2: Mortality at 36 weeks' PMA

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 3: Mortality at hospital discharge

Figures and Tables -
Analysis 1.3

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 3: Mortality at hospital discharge

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 4: Bronchopulmonary dysplasia at 36 weeks' PMA

Figures and Tables -
Analysis 1.4

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 4: Bronchopulmonary dysplasia at 36 weeks' PMA

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 5: Failure to extubate 3 days after initiation

Figures and Tables -
Analysis 1.5

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 5: Failure to extubate 3 days after initiation

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 6: Failure to extubate 7 days after initiation

Figures and Tables -
Analysis 1.6

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 6: Failure to extubate 7 days after initiation

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 7: Days of mechanical ventilation

Figures and Tables -
Analysis 1.7

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 7: Days of mechanical ventilation

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 8: Days on supplemental oxygen

Figures and Tables -
Analysis 1.8

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 8: Days on supplemental oxygen

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 9: Hypertension

Figures and Tables -
Analysis 1.9

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 9: Hypertension

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 10: Hyperglycemia

Figures and Tables -
Analysis 1.10

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 10: Hyperglycemia

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 11: Open‐label corticosteroids

Figures and Tables -
Analysis 1.11

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 11: Open‐label corticosteroids

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 12: Culture confirmed infection

Figures and Tables -
Analysis 1.12

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 12: Culture confirmed infection

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 13: Clinical suspected infection

Figures and Tables -
Analysis 1.13

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 13: Clinical suspected infection

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 14: Gastrointestinal perforation

Figures and Tables -
Analysis 1.14

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 14: Gastrointestinal perforation

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 15: Necrotizing enterocolitis

Figures and Tables -
Analysis 1.15

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 15: Necrotizing enterocolitis

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 16: Intraventricular hemorrhage (> grade II)

Figures and Tables -
Analysis 1.16

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 16: Intraventricular hemorrhage (> grade II)

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 17: Periventricular leukomalacia (PVL)

Figures and Tables -
Analysis 1.17

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 17: Periventricular leukomalacia (PVL)

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 18: Severe retinopathy of prematurity

Figures and Tables -
Analysis 1.18

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 18: Severe retinopathy of prematurity

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 19: Cerebral palsy in survivors assessed at 1‐3 years

Figures and Tables -
Analysis 1.19

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 19: Cerebral palsy in survivors assessed at 1‐3 years

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 20: Death or cerebral palsy at 1‐3 years

Figures and Tables -
Analysis 1.20

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 20: Death or cerebral palsy at 1‐3 years

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 21: Bayley's MDI < 2 SD in survivors assessed

Figures and Tables -
Analysis 1.21

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 21: Bayley's MDI < 2 SD in survivors assessed

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 22: Severe blindness in survivors assessed

Figures and Tables -
Analysis 1.22

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 22: Severe blindness in survivors assessed

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 23: Abnormal neurodevelopmental outcome in survivors assessed (various definitions) at 1‐3 years

Figures and Tables -
Analysis 1.23

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 23: Abnormal neurodevelopmental outcome in survivors assessed (various definitions) at 1‐3 years

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 24: Death or abnormal neurodevelopmental outcome (various definitions) at 1‐3 years

Figures and Tables -
Analysis 1.24

Comparison 1: Lower versus higher cumulative dose dexamethasone regimen, Outcome 24: Death or abnormal neurodevelopmental outcome (various definitions) at 1‐3 years

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 1: Death or bronchopulmonary dysplasia at 36 weeks' PMA

Figures and Tables -
Analysis 2.1

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 1: Death or bronchopulmonary dysplasia at 36 weeks' PMA

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 2: Mortality at 28 days' PNA

Figures and Tables -
Analysis 2.2

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 2: Mortality at 28 days' PNA

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 3: Mortality at 36 weeks' PMA

Figures and Tables -
Analysis 2.3

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 3: Mortality at 36 weeks' PMA

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 4: Mortality at hospital discharge

Figures and Tables -
Analysis 2.4

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 4: Mortality at hospital discharge

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 5: Bronchopulmonary dysplasia at 28 days' PNA

Figures and Tables -
Analysis 2.5

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 5: Bronchopulmonary dysplasia at 28 days' PNA

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 6: Bronchopulmonary dysplasia at 36 weeks' PMA

Figures and Tables -
Analysis 2.6

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 6: Bronchopulmonary dysplasia at 36 weeks' PMA

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 7: Failure to extubate 3 days after initiation

Figures and Tables -
Analysis 2.7

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 7: Failure to extubate 3 days after initiation

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 8: Failure to extubate 7 days after initiation

Figures and Tables -
Analysis 2.8

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 8: Failure to extubate 7 days after initiation

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 9: Days of mechanical ventilation

Figures and Tables -
Analysis 2.9

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 9: Days of mechanical ventilation

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 10: Days of supplemental oxygen

Figures and Tables -
Analysis 2.10

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 10: Days of supplemental oxygen

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 11: Hypertension

Figures and Tables -
Analysis 2.11

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 11: Hypertension

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 12: Hyperglycemia

Figures and Tables -
Analysis 2.12

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 12: Hyperglycemia

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 13: Open‐label corticosteroids

Figures and Tables -
Analysis 2.13

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 13: Open‐label corticosteroids

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 14: Culture confirmed infection

Figures and Tables -
Analysis 2.14

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 14: Culture confirmed infection

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 15: Gastrointestinal hemorrhage

Figures and Tables -
Analysis 2.15

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 15: Gastrointestinal hemorrhage

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 16: Gastrointestinal perforation

Figures and Tables -
Analysis 2.16

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 16: Gastrointestinal perforation

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 17: Necrotizing enterocolitis

Figures and Tables -
Analysis 2.17

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 17: Necrotizing enterocolitis

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 18: Patent ductus arteriosus requiring therapy

Figures and Tables -
Analysis 2.18

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 18: Patent ductus arteriosus requiring therapy

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 19: Intraventricular hemorrhage (> grade II)

Figures and Tables -
Analysis 2.19

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 19: Intraventricular hemorrhage (> grade II)

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 20: Retinopathy of prematurity (any)

Figures and Tables -
Analysis 2.20

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 20: Retinopathy of prematurity (any)

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 21: Severe retinopathy of prematurity

Figures and Tables -
Analysis 2.21

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 21: Severe retinopathy of prematurity

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 22: Cerebral palsy in survivors assessed

Figures and Tables -
Analysis 2.22

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 22: Cerebral palsy in survivors assessed

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 23: Death or cerebral palsy

Figures and Tables -
Analysis 2.23

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 23: Death or cerebral palsy

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 24: Abnormal neurodevelopmental outcome in survivors assessed (various definitions)

Figures and Tables -
Analysis 2.24

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 24: Abnormal neurodevelopmental outcome in survivors assessed (various definitions)

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 25: Death or abnormal neurodevelopmental outcome (various definitions)

Figures and Tables -
Analysis 2.25

Comparison 2: Later versus earlier initiation of dexamethasone therapy, Outcome 25: Death or abnormal neurodevelopmental outcome (various definitions)

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 1: Death or bronchopulmonary dysplasia at 36 weeks PMA

Figures and Tables -
Analysis 3.1

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 1: Death or bronchopulmonary dysplasia at 36 weeks PMA

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 2: Mortality at 28 days PNA

Figures and Tables -
Analysis 3.2

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 2: Mortality at 28 days PNA

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 3: Mortality at 36 weeks PMA

Figures and Tables -
Analysis 3.3

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 3: Mortality at 36 weeks PMA

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 4: Mortality at hospital discharge

Figures and Tables -
Analysis 3.4

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 4: Mortality at hospital discharge

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 5: Bronchopulmonary dysplasia at 28 days PNA

Figures and Tables -
Analysis 3.5

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 5: Bronchopulmonary dysplasia at 28 days PNA

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 6: Bronchopulmonary dysplasia at 36 weeks PMA

Figures and Tables -
Analysis 3.6

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 6: Bronchopulmonary dysplasia at 36 weeks PMA

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 7: Hypertension

Figures and Tables -
Analysis 3.7

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 7: Hypertension

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 8: Hyperglycemia

Figures and Tables -
Analysis 3.8

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 8: Hyperglycemia

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 9: Open‐label corticosteroids

Figures and Tables -
Analysis 3.9

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 9: Open‐label corticosteroids

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 10: Culture confirmed infection

Figures and Tables -
Analysis 3.10

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 10: Culture confirmed infection

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 11: Clinical suspected infection

Figures and Tables -
Analysis 3.11

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 11: Clinical suspected infection

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 12: Gastrointestinal hemorrhage

Figures and Tables -
Analysis 3.12

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 12: Gastrointestinal hemorrhage

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 13: Necrotizing enterocolitis

Figures and Tables -
Analysis 3.13

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 13: Necrotizing enterocolitis

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 14: Intraventricular hemorrhage (> grade II)

Figures and Tables -
Analysis 3.14

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 14: Intraventricular hemorrhage (> grade II)

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 15: Retinopathy of prematurity (any)

Figures and Tables -
Analysis 3.15

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 15: Retinopathy of prematurity (any)

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 16: Severe retinopathy of prematurity

Figures and Tables -
Analysis 3.16

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 16: Severe retinopathy of prematurity

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 17: Abnormal neurodevelopmental outcome in survivors assessed (various definitions)

Figures and Tables -
Analysis 3.17

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 17: Abnormal neurodevelopmental outcome in survivors assessed (various definitions)

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 18: Death or abnormal neurodevelopmental outcome (various definitions)

Figures and Tables -
Analysis 3.18

Comparison 3: Pulse versus continuous dexamethasone therapy, Outcome 18: Death or abnormal neurodevelopmental outcome (various definitions)

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 1: Death or bronchopulmonary dysplasia at 36 weeks PMA

Figures and Tables -
Analysis 4.1

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 1: Death or bronchopulmonary dysplasia at 36 weeks PMA

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 2: Mortality at 28 days PNA

Figures and Tables -
Analysis 4.2

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 2: Mortality at 28 days PNA

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 3: Mortality at 36 weeks PMA

Figures and Tables -
Analysis 4.3

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 3: Mortality at 36 weeks PMA

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 4: Mortality at hospital discharge

Figures and Tables -
Analysis 4.4

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 4: Mortality at hospital discharge

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 5: Bronchopulmonary dysplasia at 28 days PNA

Figures and Tables -
Analysis 4.5

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 5: Bronchopulmonary dysplasia at 28 days PNA

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 6: Bronchopulmonary dysplasia at 36 weeks PMA

Figures and Tables -
Analysis 4.6

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 6: Bronchopulmonary dysplasia at 36 weeks PMA

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 7: Failure to extubate 3 days after initiation

Figures and Tables -
Analysis 4.7

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 7: Failure to extubate 3 days after initiation

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 8: Failure to extubate 7 days after initiation

Figures and Tables -
Analysis 4.8

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 8: Failure to extubate 7 days after initiation

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 9: Days of mechanical ventilation

Figures and Tables -
Analysis 4.9

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 9: Days of mechanical ventilation

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 10: Days on supplemental oxygen

Figures and Tables -
Analysis 4.10

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 10: Days on supplemental oxygen

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 11: Hypertension

Figures and Tables -
Analysis 4.11

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 11: Hypertension

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 12: Hyperglycemia

Figures and Tables -
Analysis 4.12

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 12: Hyperglycemia

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 13: Open‐label corticosteroids

Figures and Tables -
Analysis 4.13

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 13: Open‐label corticosteroids

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 14: Culture confirmed infection

Figures and Tables -
Analysis 4.14

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 14: Culture confirmed infection

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 15: Necrotizing enterocolitis

Figures and Tables -
Analysis 4.15

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 15: Necrotizing enterocolitis

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 16: Intraventricular hemorrhage (> grade II)

Figures and Tables -
Analysis 4.16

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 16: Intraventricular hemorrhage (> grade II)

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 17: Periventricular leucomalacia

Figures and Tables -
Analysis 4.17

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 17: Periventricular leucomalacia

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 18: Retinopathy of prematurity (any)

Figures and Tables -
Analysis 4.18

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 18: Retinopathy of prematurity (any)

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 19: Cerebral palsy in survivors assessed at 1‐3 years

Figures and Tables -
Analysis 4.19

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 19: Cerebral palsy in survivors assessed at 1‐3 years

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 20: Death or cerebral palsy at 1‐3 years

Figures and Tables -
Analysis 4.20

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 20: Death or cerebral palsy at 1‐3 years

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 21: Bayley's MDI < ‐2 SD

Figures and Tables -
Analysis 4.21

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 21: Bayley's MDI < ‐2 SD

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 22: Severe blindness

Figures and Tables -
Analysis 4.22

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 22: Severe blindness

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 23: Abnormal neurodevelopmental outcome in survivors assessed (various definitions)

Figures and Tables -
Analysis 4.23

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 23: Abnormal neurodevelopmental outcome in survivors assessed (various definitions)

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 24: Death or abnormal neurodevelopmental outcome (various definitions)

Figures and Tables -
Analysis 4.24

Comparison 4: Individual tailored versus continuous tapered dexamethasone regimen, Outcome 24: Death or abnormal neurodevelopmental outcome (various definitions)

Summary of findings 1. Lower compared to higher cumulative dose dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants

Lower compared to higher cumulative dose dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants

Patient or population: preterm infants at risk for bronchopulmonary dysplasia
Intervention: lower cumulative dose dexamethasone regimen
Comparison: higher cumulative dose dexamethasone regimen

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with higher cumulative dose dexamethasone regimen

Risk with lower

Death or bronchopulmonary dysplasia at 36 weeks PMA

Study population

RR 1.03
(0.86 to 1.24)

268
(7 RCTs)

⊕⊕⊝⊝
LOWa,b

651 per 1000

671 per 1000
(560 to 808)

Death or cerebral palsy at 1 to 3 years

Study population

RR 1.74
(0.94 to 3.24)

193
(4 RCTs)

⊕⊕⊝⊝
LOWa,b

308 per 1000

535 per 1000
(289 to 997)

Death or abnormal neurodevelopmental outcome (various definitions) at 1 to 3 years

Study population

RR 1.86
(0.98 to 3.53)

100
(3 RCTs)

⊕⊝⊝⊝
VERY LOWa,b,c

172 per 1000

319 per 1000
(168 to 606)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; PMA: postmenstrual age; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded one level for risk of bias in some included studies
b Downgraded one level for serious imprecision of effect estimates (95% CI around estimate consistent with substantial harm or benefit).
c Downgraded one level for serious inconsistency across studies.

Figures and Tables -
Summary of findings 1. Lower compared to higher cumulative dose dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants
Summary of findings 2. Later compared to earlier initiation of dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants

Later compared to earlier initiation of dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants

Patient or population: preterm infants at risk for bronchopulmonary dysplasia
Intervention: later initiation of dexamethasone therapy
Comparison: earlier initiation of dexamethasone therapy

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with earlier initiation of dexamethasone therapy

Risk with later

Death or bronchopulmonary dysplasia at 36 weeks' PMA

Study population

RR 1.06
(0.87 to 1.29)

391
(3 RCTs)

⊕⊕⊝⊝
LOWa,b

476 per 1000

505 per 1000
(414 to 614)

Death or cerebral palsy at 1 to 3 years

Study population

RR 1.12
(0.68 to 1.84)

86
(1 RCT)

⊕⊕⊝⊝
LOWa,b

412 per 1000

461 per 1000
(280 to 758)

Death or abnormal neurodevelopmental outcome (various definitions)

Study population

RR 0.87
(0.63 to 1.21)

167
(2 RCTs)

⊕⊕⊝⊝
LOWa,b

507 per 1000

441 per 1000
(319 to 613)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; PMA: postmenstrual age; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded one level for serious study design limitations (unclear methodology or random sequence allocation, lack of blinding of clinicians and outcome assessment (Bloomfield 1998; Merz 1999; Hingre 1992; Halliday 2001a)
b Downgraded one level for serious imprecision of effect estimate (95% CI around estimate consistent with substantial harm or benefit and number of events < 300).

Figures and Tables -
Summary of findings 2. Later compared to earlier initiation of dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants
Summary of findings 3. Pulse compared to continuous dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants

Pulse compared to continuous dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants

Patient or population: preterm infants at risk for bronchopulmonary dysplasia
Intervention: pulse dexamethasone therapy
Comparison: continuous dexamethasone therapy

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with continuous dexamethasone therapy

Risk with pulse

Death or bronchopulmonary dysplasia at 36 weeks' PMA

Study population

RR 1.38
(1.02 to 1.88)

197
(2 RCTs)

⊕⊕⊝⊝
LOWa,b

390 per 1000

538 per 1000
(398 to 733)

Death or abnormal neurodevelopmental outcome (various definitions)

Study population

RR 1.23
(0.79 to 1.92)

76
(1 RCT)

⊕⊝⊝⊝
VERY LOWa,b,c

459 per 1000

565 per 1000
(363 to 882)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; PMA: postmenstrual age; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded one level for serious study design limitations (lack of blinding of clinicians and outcome assessment (Bloomfield 1998))
b Downgraded one level for serious imprecision of effect estimate (95% CI around estimate consistent with substantial harm or benefit and number of events < 300).
c Downgraded one level for publication bias, since the study by Barkemeyer 2001 was never published as full text.

Figures and Tables -
Summary of findings 3. Pulse compared to continuous dexamethasone therapy for prevention of bronchopulmonary dysplasia in preterm infants
Summary of findings 4. Individual tailored compared to continuous tapered dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants

Individual tailored compared to continuous tapered dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants

Patient or population: preterm infants at risk for bronchopulmonary dysplasia
Intervention: individual tailored dexamethasone regimen
Comparison: continuous tapered dexamethasone regimen

Outcomes

Anticipated absolute effects* (95% CI)

Relative effect
(95% CI)

№ of participants
(studies)

Certainty of the evidence
(GRADE)

Comments

Risk with continuous tapered dexamethasone regimen

Risk with individual tailored

Death or bronchopulmonary dysplasia at 36 weeks PMA

Study population

RR 1.06
(0.88 to 1.29)

168
(3 RCTs)

⊕⊕⊝⊝
LOWa,b

639 per 1000

677 per 1000
(562 to 824)

Death or cerebral palsy at 1 to 3 years

Study population

RR 7.24
(0.95 to 55.26)

59
(1 RCT)

⊕⊕⊕⊝
MODERATEb

33 per 1000

241 per 1000
(32 to 1000)

Death or abnormal neurodevelopmental outcome (various definitions)

Study population

RR 1.44
(0.99 to 2.07)

168
(3 RCTs)

⊕⊕⊝⊝
LOWa,b

313 per 1000

451 per 1000
(310 to 648)

*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; PMA: postmenstrual age; RR: risk ratio

GRADE Working Group grades of evidence
High certainty: We are very confident that the true effect lies close to that of the estimate of the effect
Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different
Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect
Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect

a Downgraded one level for serious study design limitations (unclear methodology or random sequence allocation, lack of blinding of clinicians and outcome assessment (Bloomfield 1998; Odd 2004 )
b Downgraded one level for serious imprecision of effect estimate (95% CI around estimate consistent with substatial harm or benefit and number of events < 300).

Figures and Tables -
Summary of findings 4. Individual tailored compared to continuous tapered dexamethasone regimen for prevention of bronchopulmonary dysplasia in preterm infants
Table 1. Participant characteristics of individual trials

Allocation arm

Participants (N)

Birthweight

(grams)

Gestational age

(weeks)

Antenatal

steroids(%)

Surfactant(%)

Starting dose (mg/kg/d)

Cumulative dose (mg/kg)‍‍

Mean age initiation

Total duration(days)

Late rescue corticosteroids

(%)

Entry FiO₂ (%)

Entry MAP

(cmH0)

Lower cumulative dosage (experimental arm) versus higher cumulative dosage (control arm)

Cummings 1989

High

13

818 ± 145

26 ± 2

38

0

0.5

7.9

14

42

0

0.60 ± 0.27

10.4 ± 6.0

Moderate

12

810 ± 208

26 ± 2

25

0

0.5

3.0

18

0

0.51 ± 0.23

8.8 ± 2.7

DeMartini 1999

High

16

741 ± 142

25.5 ± 1.7

62

100

0.5

4.1

?

21

0

0.61 ± 26.9

?

Moderate

14

848 ± 224

26.4 ± 1.6

64

100

0.5

2.7

7

0

0.60 ± 25.2

Marr 2019a

High

30

769 ± 149

25.2 ± 1.2

63

97

0.5

7.96

14 ± 4

42

20

0.72 ± 0.13

10.3 ± 2.0

Moderate

29

785 ± 167

25.2 ± 1.1

62

86

0.5

4.04

13 ± 3

9b

37

0.77 ± 0.16

10.4 ± 1.7

Malloy 2005

Moderate

9c

767 ± 149

25.8 ± 0.9

75

100

0.5

2.7

14.8 ± 6.5

7

88

0.57 ± 0.08

?

Low

8

773 ± 182

26.1 ± 1.8

63

100

0.08

0.6

16.8 ± 5.7

7

50

0.52 ± 0.16

Durand 2002

Moderate

23

932 ± 182

27.1 ± 1.8

52

87

0.5

2.4

11.5 ± 2.2

7

22

0.43 ± 0.11

7.8 ± 2.2

Low

24

858 ± 186

26.9 ± 1.6

50

88

0.2

1.0

11.3 ± 2.7

7

29

0.41 ± 0.10

7.0 ± 1.2

McEvoy 2004

Moderate

29

839 ± 229

26.1 ± 2.0

34

97

0.5

2.4

10.7 ± 3.7

7

55

0.44 ± 0.13

6.8 ± 1.8

Low

33

830 ± 248

26.3 ± 1.8

48

82

0.2

1.0

11.6 ± 4.3

7

39

0.42 ± 0.13

7.4 ± 2.2

Ramanathan 1994

Moderate

15

850 ± 290

27 ± 2

?

67

0.4

1.9d

10 to 14

7

67

?

?

Low

13

817 ± 186

27 ± 2

62

0.2

1.0d

7

54

Da Silva 2002

Moderate

17

821 ± 160

25.4 ± 0.9

?

?

0.5

?

?

7

?

?

?

Low

21

851 ± 465

25.7 ± 1.8

0.1

0.7

7

Later initiation (experimental arm) versus earlier (control arm)

Papile 1998

ME

182

808 ± 187

25.7 ± 1.9

29

91

0.5

3.7

14

14

12

0.54 ± 0.18

8 ± 2

L

189

801 ± 182

25.6 ± 1.6

27

89

28

16

0.54 ± 0.19

8 ± 2

Merz 1999

E

15

980 (710 to 1250)

27 (25 to 29)

87

87

0.5

3.1

7

16

0

0.3 (0.25 to 0.5)

?

ME

15

938 (680 to 1250)

27.5 (24 to 29)

73

73

14

0

0.3 (0.25 to 0.55)

Halliday 2001a

E

135

1017 ± 290

27.4 ± 1.9

61

95

0.5

2.7

3

12

?

?

?

ME

150

1007 ± 283

27.1 ± 1.9

55

92

16

Hingre 1992

E

14

744 ± 144

26 ± 2

?

?

0.5

7.96

5

42

?

?

?

ME

21

730 ± 135e

25 ± 2

14

Pulse dosage regimen (experimental arm) versus continuous dosage regimen (control arm)

Bloomfield 1998

Pulse/Ef

39

776 ± 25

25.8 ± 0.3

95

?

0.5

5.3 (1.5 to 11.8)

7

34 (11 to 73)

?

0.30 ± 0.02

8.0 ± 0.3

Cont/ME

37

793 ± 28

25.8 ± 0.3

73

0.5

7.1 (4.5 to 7.6)

14

42 (42 to 51)

0.30 ± 0.01

7.8 ± 0.3

Barkemeyer 2001

Pulse

58

816

26.1

84

92

0.5

4.5

7 to 21

23

41

?

?

Cont

63

842

26.2

78

88

36

Individualized tailored (experimental arm) versus standard dosage regimen (control arm)

Odd 2004

Indiv

17

669 ± 113

24 (23 to 27)

?

?

0.5

3.8 (2.0 to 5.7)

12 (7 to 16)

42 (5 to 73)

?

0.40 (0.25 to 1.0)

9 (7 to 14)

Cont

16

720 ± 130

24 (23 to 26)

0.5

7.9

10 (7 to 23)

42

0.40 (0.21 to 1.0)

9 (7 to 13)

a Marr not only in higher versus lower comparison, but also in individual tailored versus standard dosage regimen
b 19 of the 29 infants (66%) received one course, 5 infants (17%) 2 courses, and 5 infants (17%) 3 courses of dexamethasone
c Including one patient in high dose group who died on the second day of treatment
d Estimated cumulative dose based on abstract data
e participant characteristics calculated on 16 participants in the moderately early group
f Bloomfield not only pulse versus continuous comparison, but also in early versus later initiation and in individualized tailored versus standard dosage regimens comparison

Mean ± standard deviation or median (interquartile range)

Abbreviations: FiO2: fractional inspired oxygen; MAP: mean airway pressure; E: Early initiation (≤ 7 days' PNA); ME: Moderately early initiation (7 to 14 days' PNA); L: Late initiation (> 14 days' PNA); Pulse: Pulse dosage regimen; Cont: Continuous tapered dosage regimen; Indiv: Individual tailored regimen.

Figures and Tables -
Table 1. Participant characteristics of individual trials
Comparison 1. Lower versus higher cumulative dose dexamethasone regimen

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

1.1 Death or bronchopulmonary dysplasia at 36 weeks PMA Show forest plot

7

268

Risk Ratio (M‐H, Fixed, 95% CI)

1.03 [0.86, 1.24]

1.1.1 Moderate versus high cumulative dose regimen

3

114

Risk Ratio (M‐H, Fixed, 95% CI)

1.11 [0.95, 1.30]

1.1.2 Low versus moderate cumulative dose regimen

4

154

Risk Ratio (M‐H, Fixed, 95% CI)

0.83 [0.50, 1.40]

1.2 Mortality at 36 weeks' PMA Show forest plot

7

268

Risk Ratio (M‐H, Fixed, 95% CI)

0.68 [0.29, 1.60]

1.2.1 Moderate versus high cumulative dose regimen

3

114

Risk Ratio (M‐H, Fixed, 95% CI)

1.07 [0.34, 3.41]

1.2.2 Low versus moderate cumulative dose regimen

4

154

Risk Ratio (M‐H, Fixed, 95% CI)

0.43 [0.11, 1.63]

1.3 Mortality at hospital discharge Show forest plot

7

268

Risk Ratio (M‐H, Fixed, 95% CI)

0.93 [0.48, 1.81]

1.3.1 Moderate versus high cumulative dose regimen

3

114

Risk Ratio (M‐H, Fixed, 95% CI)

1.36 [0.62, 2.99]

1.3.2 Low versus moderate cumulative dose regimen

4

154

Risk Ratio (M‐H, Fixed, 95% CI)

0.43 [0.11, 1.63]

1.4 Bronchopulmonary dysplasia at 36 weeks' PMA Show forest plot

7

268

Risk Ratio (M‐H, Fixed, 95% CI)

1.12 [0.91, 1.37]

1.4.1 Moderate versus high cumulative dose regimen

3

114

Risk Ratio (M‐H, Fixed, 95% CI)

1.12 [0.93, 1.35]

1.4.2 Low versus moderate cumulative dose regimen

4

154

Risk Ratio (M‐H, Fixed, 95% CI)

1.12 [0.65, 1.95]

1.5 Failure to extubate 3 days after initiation Show forest plot

5

209

Risk Ratio (M‐H, Fixed, 95% CI)

1.12 [0.98, 1.29]

1.5.1 Moderate versus high cumulative dose regimen

2

84

Risk Ratio (M‐H, Fixed, 95% CI)

1.11 [0.93, 1.32]

1.5.2 Low versus moderate dose regimen

3

125

Risk Ratio (M‐H, Fixed, 95% CI)

1.14 [0.93, 1.39]

1.6 Failure to extubate 7 days after initiation Show forest plot

5

210

Risk Ratio (M‐H, Fixed, 95% CI)

1.32 [1.08, 1.60]

1.6.1 Moderate versus high cumulative dose regimen

2

84

Risk Ratio (M‐H, Fixed, 95% CI)

1.49 [1.10, 2.02]

1.6.2 Low versus moderate cumulative dose regimen

3

126

Risk Ratio (M‐H, Fixed, 95% CI)

1.21 [0.94, 1.56]

1.7 Days of mechanical ventilation Show forest plot

6

218

Mean Difference (IV, Fixed, 95% CI)

4.50 [‐0.68, 9.67]

1.7.1 Moderate versus high cumulative dose regimen

3

112

Mean Difference (IV, Fixed, 95% CI)

8.09 [0.21, 15.96]

1.7.2 Low versus moderate cumulative dose regimen

3

106

Mean Difference (IV, Fixed, 95% CI)

1.77 [‐5.09, 8.64]

1.8 Days on supplemental oxygen Show forest plot

2

80

Mean Difference (IV, Fixed, 95% CI)

0.30 [‐20.14, 20.74]

1.8.1 Moderate versus high cumulative dose regimen

1

52

Mean Difference (IV, Fixed, 95% CI)

8.00 [‐34.64, 50.64]

1.8.2 Low versus moderate cumulative dose regimen

1

28

Mean Difference (IV, Fixed, 95% CI)

‐2.00 [‐25.29, 21.29]

1.9 Hypertension Show forest plot

6

240

Risk Ratio (M‐H, Fixed, 95% CI)

0.31 [0.12, 0.77]

1.9.1 Moderate versus high cumulative dose regimen

3

114

Risk Ratio (M‐H, Fixed, 95% CI)

0.27 [0.03, 2.34]

1.9.2 Low versus moderate cumulative dose regimen

3

126

Risk Ratio (M‐H, Fixed, 95% CI)

0.31 [0.11, 0.87]

1.10 Hyperglycemia Show forest plot

6

240

Risk Ratio (M‐H, Fixed, 95% CI)

0.60 [0.37, 0.97]

1.10.1 Moderate versus high cumulative dose regimen

3

114

Risk Ratio (M‐H, Fixed, 95% CI)

0.82 [0.47, 1.46]

1.10.2 Low versus moderate cumulative dose regimen

3

126

Risk Ratio (M‐H, Fixed, 95% CI)

0.40 [0.17, 0.93]

1.11 Open‐label corticosteroids Show forest plot

7

268

Risk Ratio (M‐H, Fixed, 95% CI)

0.93 [0.68, 1.28]

1.11.1 Moderate versus high cumulative dose regimen

3

114

Risk Ratio (M‐H, Fixed, 95% CI)

1.72 [0.72, 4.13]

1.11.2 Low versus moderate cumulative dose regimen

4

154

Risk Ratio (M‐H, Fixed, 95% CI)

0.81 [0.57, 1.14]

1.12 Culture confirmed infection Show forest plot

7

289

Risk Ratio (M‐H, Fixed, 95% CI)

0.96 [0.67, 1.39]

1.12.1 Moderate versus high cumulative dose regimen

3

114

Risk Ratio (M‐H, Fixed, 95% CI)

1.20 [0.71, 2.01]

1.12.2 Low versus moderate cumulative dose regimen

4

175

Risk Ratio (M‐H, Fixed, 95% CI)

0.78 [0.46, 1.32]

1.13 Clinical suspected infection Show forest plot

3

131

Risk Ratio (M‐H, Fixed, 95% CI)

1.03 [0.62, 1.70]

1.13.1 Moderate versus high cumulative dose regimen

2

84

Risk Ratio (M‐H, Fixed, 95% CI)

1.22 [0.71, 2.09]

1.13.2 Low versus moderate cumulative dose regimen

1

47

Risk Ratio (M‐H, Fixed, 95% CI)

0.82 [0.32, 2.08]

1.14 Gastrointestinal perforation Show forest plot

4

185

Risk Ratio (M‐H, Fixed, 95% CI)

0.92 [0.13, 6.28]

1.14.1 Moderate versus high cumulative dose regimens

1

59

Risk Ratio (M‐H, Fixed, 95% CI)

Not estimable

1.14.2 Low versus moderate cumulative dose regimen

3

126

Risk Ratio (M‐H, Fixed, 95% CI)

0.92 [0.13, 6.28]

1.15 Necrotizing enterocolitis Show forest plot

4

198

Risk Ratio (M‐H, Fixed, 95% CI)

0.53 [0.18, 1.56]

1.15.1 Moderate versus high cumulative dose regimen

2

89

Risk Ratio (M‐H, Fixed, 95% CI)

0.86 [0.23, 3.19]

1.15.2 Low versus moderate cumulative dose regimen

2

109

Risk Ratio (M‐H, Fixed, 95% CI)

0.23 [0.03, 1.97]

1.16 Intraventricular hemorrhage (> grade II) Show forest plot

3

101

Risk Ratio (M‐H, Fixed, 95% CI)

1.68 [0.65, 4.37]

1.16.1 Moderate versus high cumulative dose regimen

2

84

Risk Ratio (M‐H, Fixed, 95% CI)

1.68 [0.49, 5.72]

1.16.2 Low versus moderate cumulative dose regimen

1

17

Risk Ratio (M‐H, Fixed, 95% CI)

1.69 [0.37, 7.67]

1.17 Periventricular leukomalacia (PVL) Show forest plot

2

121

Risk Ratio (M‐H, Fixed, 95% CI)

0.93 [0.20, 4.39]

1.17.1 Moderate versus high cumulative dose regimens

1

59

Risk Ratio (M‐H, Fixed, 95% CI)

1.03 [0.07, 15.77]

1.17.2 Low versus moderate cumulative dose regimen

1

62

Risk Ratio (M‐H, Fixed, 95% CI)

0.88 [0.13, 5.85]

1.18 Severe retinopathy of prematurity Show forest plot

5

176

Risk Ratio (M‐H, Fixed, 95% CI)

0.64 [0.32, 1.28]

1.18.1 Moderate versus high cumulative dose regimen

2

84

Risk Ratio (M‐H, Fixed, 95% CI)

0.60 [0.19, 1.92]

1.18.2 Low versus moderate cumulative dose regimen

3

92

Risk Ratio (M‐H, Fixed, 95% CI)

0.66 [0.28, 1.57]

1.19 Cerebral palsy in survivors assessed at 1‐3 years Show forest plot

4

149

Risk Ratio (M‐H, Fixed, 95% CI)

2.64 [1.02, 6.83]

1.19.1 Moderate versus high cumulative dose regimen

2

74

Risk Ratio (M‐H, Fixed, 95% CI)

6.85 [1.29, 36.36]

1.19.2 Low versus moderate cumulative dose regimen

2

75

Risk Ratio (M‐H, Fixed, 95% CI)

1.08 [0.29, 4.00]

1.20 Death or cerebral palsy at 1‐3 years Show forest plot

4

193

Risk Ratio (M‐H, Fixed, 95% CI)

1.74 [0.94, 3.24]

1.20.1 Moderate versus high cumulative dose regimen

2

84

Risk Ratio (M‐H, Fixed, 95% CI)

3.20 [1.35, 7.58]

1.20.2 Low versus moderate dose regimen

2

109

Risk Ratio (M‐H, Fixed, 95% CI)

0.78 [0.28, 2.18]

1.21 Bayley's MDI < 2 SD in survivors assessed Show forest plot

4

147

Risk Ratio (M‐H, Fixed, 95% CI)

1.44 [0.71, 2.92]

1.21.1 Moderate versus high cumulative dose regimen

2

72

Risk Ratio (M‐H, Fixed, 95% CI)

3.99 [1.06, 15.08]

1.21.2 Low versus moderate cumulative dose regimen

2

75

Risk Ratio (M‐H, Fixed, 95% CI)

0.72 [0.29, 1.83]

1.22 Severe blindness in survivors assessed Show forest plot

5

166

Risk Ratio (M‐H, Fixed, 95% CI)

0.66 [0.17, 2.66]

1.22.1 Moderate versus high cumulative dose regimen

2

74

Risk Ratio (M‐H, Fixed, 95% CI)

3.44 [0.15, 81.09]

1.22.2 Low versus moderate cumulative dose regimen

3

92

Risk Ratio (M‐H, Fixed, 95% CI)

0.36 [0.06, 2.19]

1.23 Abnormal neurodevelopmental outcome in survivors assessed (various definitions) at 1‐3 years Show forest plot

3

89

Risk Ratio (M‐H, Fixed, 95% CI)

2.22 [0.86, 5.76]

1.23.1 Moderate versus high cumulative dose regimen

2

74

Risk Ratio (M‐H, Fixed, 95% CI)

7.60 [1.45, 39.78]

1.23.2 Low versus moderate cumulative dose regimen

1

15

Risk Ratio (M‐H, Fixed, 95% CI)

0.30 [0.05, 1.97]

1.24 Death or abnormal neurodevelopmental outcome (various definitions) at 1‐3 years Show forest plot

3

100

Risk Ratio (M‐H, Fixed, 95% CI)

1.86 [0.98, 3.53]

1.24.1 Moderate versus high cumulative dose regimen

2

84

Risk Ratio (M‐H, Fixed, 95% CI)

3.41 [1.44, 8.07]

1.24.2 Low versus moderate cumulative dose regimen

1

16

Risk Ratio (M‐H, Fixed, 95% CI)

0.43 [0.12, 1.51]

Figures and Tables -
Comparison 1. Lower versus higher cumulative dose dexamethasone regimen
Comparison 2. Later versus earlier initiation of dexamethasone therapy

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

2.1 Death or bronchopulmonary dysplasia at 36 weeks' PMA Show forest plot

3

391

Risk Ratio (M‐H, Fixed, 95% CI)

1.06 [0.87, 1.29]

2.1.1 Moderate early versus early initiation

3

391

Risk Ratio (M‐H, Fixed, 95% CI)

1.06 [0.87, 1.29]

2.2 Mortality at 28 days' PNA Show forest plot

4

762

Risk Ratio (M‐H, Fixed, 95% CI)

1.01 [0.69, 1.47]

2.2.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

2.20 [0.93, 5.23]

2.2.2 Moderate early versus early initiation

3

391

Risk Ratio (M‐H, Fixed, 95% CI)

0.78 [0.51, 1.20]

2.3 Mortality at 36 weeks' PMA Show forest plot

4

762

Risk Ratio (M‐H, Fixed, 95% CI)

0.93 [0.68, 1.28]

2.3.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

1.47 [0.83, 2.62]

2.3.2 Moderate early versus early initiation

3

391

Risk Ratio (M‐H, Fixed, 95% CI)

0.73 [0.49, 1.07]

2.4 Mortality at hospital discharge Show forest plot

5

797

Risk Ratio (M‐H, Fixed, 95% CI)

1.00 [0.75, 1.33]

2.4.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

1.47 [0.83, 2.62]

2.4.2 Moderate early versus early initiation

4

426

Risk Ratio (M‐H, Fixed, 95% CI)

0.85 [0.61, 1.18]

2.5 Bronchopulmonary dysplasia at 28 days' PNA Show forest plot

4

762

Risk Ratio (M‐H, Fixed, 95% CI)

1.12 [1.02, 1.23]

2.5.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

1.15 [1.05, 1.26]

2.5.2 Moderate early versus early initiation

3

391

Risk Ratio (M‐H, Fixed, 95% CI)

1.08 [0.91, 1.29]

2.6 Bronchopulmonary dysplasia at 36 weeks' PMA Show forest plot

4

762

Risk Ratio (M‐H, Fixed, 95% CI)

1.11 [0.97, 1.28]

2.6.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

1.01 [0.88, 1.17]

2.6.2 Moderate early versus early initiation

3

391

Risk Ratio (M‐H, Fixed, 95% CI)

1.38 [1.01, 1.90]

2.7 Failure to extubate 3 days after initiation Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.7.1 Late versus moderate early initiation

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.8 Failure to extubate 7 days after initiation Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.8.1 Late versus moderate early initiation

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.9 Days of mechanical ventilation Show forest plot

2

60

Mean Difference (IV, Fixed, 95% CI)

12.71 [4.44, 20.99]

2.9.1 Moderate early versus early initiation

2

60

Mean Difference (IV, Fixed, 95% CI)

12.71 [4.44, 20.99]

2.10 Days of supplemental oxygen Show forest plot

1

Mean Difference (IV, Fixed, 95% CI)

Totals not selected

2.10.1 Moderate early versus early initiation

1

Mean Difference (IV, Fixed, 95% CI)

Totals not selected

2.11 Hypertension Show forest plot

4

762

Risk Ratio (M‐H, Fixed, 95% CI)

0.99 [0.67, 1.47]

2.11.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

1.30 [0.72, 2.36]

2.11.2 Moderate early versus early initiation

3

391

Risk Ratio (M‐H, Fixed, 95% CI)

0.79 [0.47, 1.34]

2.12 Hyperglycemia Show forest plot

4

726

Risk Ratio (M‐H, Fixed, 95% CI)

0.66 [0.53, 0.82]

2.12.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

0.66 [0.46, 0.95]

2.12.2 Moderate early versus early initiation

3

355

Risk Ratio (M‐H, Fixed, 95% CI)

0.66 [0.51, 0.85]

2.13 Open‐label corticosteroids Show forest plot

3

732

Risk Ratio (M‐H, Fixed, 95% CI)

1.71 [1.04, 2.81]

2.13.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

1.38 [0.82, 2.31]

2.13.2 Moderate early versus early initiation

2

361

Risk Ratio (M‐H, Fixed, 95% CI)

15.31 [0.89, 262.78]

2.14 Culture confirmed infection Show forest plot

3

732

Risk Ratio (M‐H, Fixed, 95% CI)

0.82 [0.68, 0.98]

2.14.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

0.67 [0.54, 0.84]

2.14.2 Moderate early versus early initiation

2

361

Risk Ratio (M‐H, Fixed, 95% CI)

1.16 [0.83, 1.63]

2.15 Gastrointestinal hemorrhage Show forest plot

4

762

Risk Ratio (M‐H, Fixed, 95% CI)

0.66 [0.45, 0.97]

2.15.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

0.60 [0.38, 0.95]

2.15.2 Moderate early versus early initiation

3

391

Risk Ratio (M‐H, Fixed, 95% CI)

0.84 [0.41, 1.71]

2.16 Gastrointestinal perforation Show forest plot

2

315

Risk Ratio (M‐H, Fixed, 95% CI)

0.75 [0.23, 2.40]

2.16.1 Moderate early versus early initiation

2

315

Risk Ratio (M‐H, Fixed, 95% CI)

0.75 [0.23, 2.40]

2.17 Necrotizing enterocolitis Show forest plot

4

725

Risk Ratio (M‐H, Fixed, 95% CI)

1.44 [0.82, 2.55]

2.17.1 Late versus moderate early initiation

1

371

Risk Ratio (M‐H, Fixed, 95% CI)

1.73 [0.59, 5.07]

2.17.2 Moderate early versus early initiation

3

354

Risk Ratio (M‐H, Fixed, 95% CI)

1.33 [0.68, 2.61]

2.18 Patent ductus arteriosus requiring therapy Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.18.1 Moderate early versus early initiation

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.19 Intraventricular hemorrhage (> grade II) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.19.1 Moderate early versus early initiation

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.20 Retinopathy of prematurity (any) Show forest plot

2

324

Risk Ratio (M‐H, Fixed, 95% CI)

0.80 [0.52, 1.23]

2.20.1 Moderate early versus early initiation

2

324

Risk Ratio (M‐H, Fixed, 95% CI)

0.80 [0.52, 1.23]

2.21 Severe retinopathy of prematurity Show forest plot

3

391

Risk Ratio (M‐H, Fixed, 95% CI)

1.50 [0.63, 3.53]

2.21.1 Moderate early versus early initiation

3

391

Risk Ratio (M‐H, Fixed, 95% CI)

1.50 [0.63, 3.53]

2.22 Cerebral palsy in survivors assessed Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.22.1 Moderate early versus early initiation

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.23 Death or cerebral palsy Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.23.1 Moderate early versus early initiation

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

2.24 Abnormal neurodevelopmental outcome in survivors assessed (various definitions) Show forest plot

2

155

Risk Ratio (M‐H, Fixed, 95% CI)

1.06 [0.66, 1.69]

2.24.1 Moderate early versus early initiation

2

155

Risk Ratio (M‐H, Fixed, 95% CI)

1.06 [0.66, 1.69]

2.25 Death or abnormal neurodevelopmental outcome (various definitions) Show forest plot

2

167

Risk Ratio (M‐H, Fixed, 95% CI)

0.87 [0.63, 1.21]

2.25.1 Moderate early versus early

2

167

Risk Ratio (M‐H, Fixed, 95% CI)

0.87 [0.63, 1.21]

Figures and Tables -
Comparison 2. Later versus earlier initiation of dexamethasone therapy
Comparison 3. Pulse versus continuous dexamethasone therapy

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

3.1 Death or bronchopulmonary dysplasia at 36 weeks PMA Show forest plot

2

197

Risk Ratio (M‐H, Fixed, 95% CI)

1.38 [1.02, 1.88]

3.2 Mortality at 28 days PNA Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3.3 Mortality at 36 weeks PMA Show forest plot

2

197

Risk Ratio (M‐H, Fixed, 95% CI)

2.04 [0.72, 5.78]

3.4 Mortality at hospital discharge Show forest plot

2

197

Risk Ratio (M‐H, Fixed, 95% CI)

2.04 [0.72, 5.78]

3.5 Bronchopulmonary dysplasia at 28 days PNA Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3.6 Bronchopulmonary dysplasia at 36 weeks PMA Show forest plot

2

197

Risk Ratio (M‐H, Fixed, 95% CI)

1.29 [0.90, 1.83]

3.7 Hypertension Show forest plot

2

197

Risk Ratio (M‐H, Fixed, 95% CI)

0.50 [0.20, 1.23]

3.8 Hyperglycemia Show forest plot

2

160

Risk Ratio (M‐H, Fixed, 95% CI)

1.08 [0.71, 1.65]

3.9 Open‐label corticosteroids Show forest plot

2

197

Risk Ratio (M‐H, Fixed, 95% CI)

0.97 [0.64, 1.47]

3.10 Culture confirmed infection Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3.11 Clinical suspected infection Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3.12 Gastrointestinal hemorrhage Show forest plot

2

197

Risk Ratio (M‐H, Fixed, 95% CI)

0.65 [0.25, 1.68]

3.13 Necrotizing enterocolitis Show forest plot

2

160

Risk Ratio (M‐H, Fixed, 95% CI)

0.78 [0.33, 1.83]

3.14 Intraventricular hemorrhage (> grade II) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3.15 Retinopathy of prematurity (any) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3.16 Severe retinopathy of prematurity Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3.17 Abnormal neurodevelopmental outcome in survivors assessed (various definitions) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

3.18 Death or abnormal neurodevelopmental outcome (various definitions) Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

Figures and Tables -
Comparison 3. Pulse versus continuous dexamethasone therapy
Comparison 4. Individual tailored versus continuous tapered dexamethasone regimen

Outcome or subgroup title

No. of studies

No. of participants

Statistical method

Effect size

4.1 Death or bronchopulmonary dysplasia at 36 weeks PMA Show forest plot

3

168

Risk Ratio (M‐H, Fixed, 95% CI)

1.06 [0.88, 1.29]

4.2 Mortality at 28 days PNA Show forest plot

2

109

Risk Ratio (M‐H, Fixed, 95% CI)

2.83 [0.60, 13.32]

4.3 Mortality at 36 weeks PMA Show forest plot

3

168

Risk Ratio (M‐H, Fixed, 95% CI)

1.54 [0.63, 3.79]

4.4 Mortality at hospital discharge Show forest plot

3

168

Risk Ratio (M‐H, Fixed, 95% CI)

1.84 [0.77, 4.37]

4.5 Bronchopulmonary dysplasia at 28 days PNA Show forest plot

2

109

Risk Ratio (M‐H, Fixed, 95% CI)

1.15 [0.88, 1.50]

4.6 Bronchopulmonary dysplasia at 36 weeks PMA Show forest plot

3

168

Risk Ratio (M‐H, Fixed, 95% CI)

0.99 [0.79, 1.25]

4.7 Failure to extubate 3 days after initiation Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

4.8 Failure to extubate 7 days after initiation Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

4.9 Days of mechanical ventilation Show forest plot

2

90

Mean Difference (IV, Fixed, 95% CI)

9.26 [4.32, 14.21]

4.10 Days on supplemental oxygen Show forest plot

1

Mean Difference (IV, Fixed, 95% CI)

Totals not selected

4.11 Hypertension Show forest plot

2

135

Risk Ratio (M‐H, Fixed, 95% CI)

0.34 [0.01, 8.13]

4.12 Hyperglycemia Show forest plot

2

98

Risk Ratio (M‐H, Fixed, 95% CI)

0.66 [0.26, 1.66]

4.13 Open‐label corticosteroids Show forest plot

2

135

Risk Ratio (M‐H, Fixed, 95% CI)

1.72 [0.72, 4.13]

4.14 Culture confirmed infection Show forest plot

2

92

Risk Ratio (M‐H, Fixed, 95% CI)

1.27 [0.58, 2.76]

4.15 Necrotizing enterocolitis Show forest plot

2

98

Risk Ratio (M‐H, Fixed, 95% CI)

1.75 [0.48, 6.35]

4.16 Intraventricular hemorrhage (> grade II) Show forest plot

3

168

Risk Ratio (M‐H, Fixed, 95% CI)

2.00 [0.78, 5.18]

4.17 Periventricular leucomalacia Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

4.18 Retinopathy of prematurity (any) Show forest plot

2

98

Risk Ratio (M‐H, Fixed, 95% CI)

0.83 [0.24, 2.92]

4.19 Cerebral palsy in survivors assessed at 1‐3 years Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

4.20 Death or cerebral palsy at 1‐3 years Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

4.21 Bayley's MDI < ‐2 SD Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

4.22 Severe blindness Show forest plot

1

Risk Ratio (M‐H, Fixed, 95% CI)

Totals not selected

4.23 Abnormal neurodevelopmental outcome in survivors assessed (various definitions) Show forest plot

3

143

Risk Ratio (M‐H, Fixed, 95% CI)

1.09 [0.71, 1.70]

4.24 Death or abnormal neurodevelopmental outcome (various definitions) Show forest plot

3

168

Risk Ratio (M‐H, Fixed, 95% CI)

1.44 [0.99, 2.07]

Figures and Tables -
Comparison 4. Individual tailored versus continuous tapered dexamethasone regimen