Summary of findings 3. Inhaled steroids compared with systemic steroids for BPD (infants randomised at < 72 hours or between 12 and 21 days of age).

Inhaled steroids compared with systemic steroids for BPD (infants randomised at < 72 hours or between 12 and 21 days of age)
Patient or population: Neonates with developing BPD Settings: Neonatal intensive care unit Intervention: Inhaled steroids Comparison: Systemic steroids
Outcomes	Illustrative comparative risks* (95% CI)		Relative effect (95% CI)	No of Participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Systemic steroids	Inhaled steroids
BPD at 36 weeks' postmenstrual age	High risk population		RR 1.08 (95% CI 0.88 to 1.32)	429 (3)	⊕⊕⊝⊝ low	Bias: The risk of bias for these three studies was high. There was blinding of randomisation in all three studies. There was no blinding of the intervention or outcome measurements at all sites in the largest study (Halliday 2001). In Rozycki 2003 there was blinding of the intervention but blinding of outcome assessment was unclear. In Suchomski 2002 there was no blinding of the intervention or outcomes measurements. We downgraded the quality of the evidence by two levels. Heterogeneity/consistency: Heterogeneity was low (I² = 39%). Directness of the evidence: The studies were conducted in the target population of newborn infants. Precision: The precision for the point estimate was high as the sample size was quite large. Presence of publication bias: N/A. We did not create a funnel plot as there were only three trials included in the analysis.
	422 per 1000	485 per 1000 (394 to 776)
Hyperglycaemia	High risk population		RR 0.86 (95% CI 0.61 to 1.22)	429 (3)	⊕⊕⊝⊝ low	Bias: The risk of bias for these three studies was high. There was blinding of randomisation in all three studies. There was no blinding of the intervention or outcome measurements at all sites in the largest study (Halliday 2001). In Rozycki 2003 there was blinding of the intervention but blinding of outcome assessments was unclear. In Suchomski 2002 there was no blinding of the intervention or outcome measurements. We downgraded the quality of the evidence by two levels. Heterogeneity/consistency: There was no heterogeneity (I² = 8%). Directness of the evidence: The studies were conducted in the target population of newborn infants. Precision: The precision for the point estimate was high as the sample size was quite large. Presence of publication bias: N/A. We did not create a funnel plot as there were only three trials included in the analysis.
	255 per 1000	177 per 1000 (0 to 282)
Hypertension	High risk population		RR (RR 0.86, 95% CI 0.73 to 1.01)	429 (3)	⊕⊕⊝⊝ low	Bias: The risk of bias for these three studies was high. There was blinding of randomisation in all three studies. There was no blinding of the intervention or outcome measurements at all sites in the largest study (Halliday 2001). In Rozycki 2003 there was blinding of the intervention but blinding of outcome assessments was unclear. In Suchomski 2002 there was no blinding of the intervention or outcome measurements. We downgraded the quality of the evidence by two steps. Heterogeneity/consistency: There was no heterogeneity (I² = 0%). Directness of the evidence: The studies were conducted in the target population of newborn infants. Precision: The precision for the point estimate was high as the sample size was quite large. Presence of publication bias: N/A. We did not create a funnel plot as there were only three trials included in the analysis.
	604 per 1000	430 per 1000 (130 to 627)
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (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; RR: Risk Ratio; BPD: Bronchopulmonary dysplasia; N/A: Not applicable
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.