Summary of findings for the main comparison. Paracetamol compared to ibuprofen for patent ductus arteriosus in preterm or low birth weight infants.

Paracetamol (oral or IV) compared to ibuprofen (oral or IV) for patent ductus arteriosus in preterm or low birth weight infants
Patient or population: preterm or low birth weight infants with patent ductus arteriosus Settings: hospitals in China (2 studies), Egypt, Jordan, and Turkey Intervention: paracetamol Comparison: ibuprofen
Outcomes	Illustrative comparative risks* (95% CI)		Relative effect (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	Oral ibuprofen	Oral paracetamol
Failure of ductal closure after the first course of treatment Echocardiogram	High risk study population		RR 0.95 (0.75 to 1.21)	559 (5 studies)	⊕⊕⊕⊝ moderate	Bias: we had no concerns for random sequence generation in the 5 included trials but the allocation concealment was unclear in 1 of the studies. We did not downgrade the quality of evidence on this item. There were concerns about blinding of personnel and of blinding of outcome assessments. We downgraded the quality of the evidence by 1 step. Heterogeneity/Consistency: we noted no heterogeneity (I² = 0% for both RR and RD). Directness of evidence: studies were conducted in the target population. Precision: because of the relatively large sample size (559 infants), the point estimate was precise with a narrow 95% CI. Presence of publication bias: Although only 5 studies were included, the funnel plot we constructed was symmetrical.
	329 per 1000	312 per 1000 (200 to 438)
All‐cause mortality during initial hospital stay Clinical assessment, no risk of bias	High risk study population		RR 0.96 (0.55 to 1.67)	272 (3 studies)	⊕⊕⊕⊝ moderate	Bias: we had no concerns regarding the assessment of mortality. Heterogeneity/Consistency: we noted no heterogeneity (I² = 0% for both RR and RD). Directness of evidence: studies were conducted in the target population. Precision: because of the relatively small sample size (272 infants), the point estimate was not precise with a wide 95% CI. We downgraded the quality of the evidence by 1 step. Presence of publication bias: as only 3 studies were included we did not perform a funnel plot.
	152 per 1000	152 per 1000 (125 to 231)
Neurodevelopmental impairment Clinical assessments by assessors blinded to group assignment, no risk of bias	High risk study population		RR 0.93 (0.44 to 1.96)	61 (1 study)	⊕⊕⊝⊝ low	Bias: of the 75 infants eligible for follow‐up at 18 to 24 months' corrected age 61 infants were evaluated (81%).The assessor was blinded to the previous assignment to paracetamol or ibuprofen groups. Heterogeneity/Consistency: as only 1 study was included in the analysis, tests for heterogeneity were not applicable. Precision: because of the small sample size of the only included study (61 infants), the point estimate was not precise with a wide 95% CI. We downgraded the quality of the evidence by 2 steps. Presence of publication bias: as only 1 study was included we did not perform a funnel plot.
	323 per 1000	300 per 1000 (300 ‒ 1 study no range)
Gastrointestinal bleed or stools positive for occult blood	High risk study population		RR 0.28 (0.12 to 0.69)	537 (4 studies)	⊕⊕⊕⊝ moderate	Bias: we had no concerns for random sequence generation in the 4 included trials but the allocation concealment was unclear in 1 of the studies. We did not downgrade the quality of evidence on this item. There were concerns about blinding of personnel and of blinding of outcome assessments. We downgraded the quality of the evidence by 1 step. Heterogeneity/Consistency: we noted no heterogeneity (I² = 0% for both RR and RD). Directness of evidence: studies were conducted in the target population. Precision: because of the relatively large sample size (537 infants), the point estimate was precise with a narrow 95% CI. Presence of publication bias: as only 5 studies were included we did not construct a funnel plot.
	78 per 1000	22 per 1000 (12 to 135)
Serum levels of creatinine after treatment (µmol/L) Serum samples		The weighted mean difference (WMD) for serum levels of creatinine after treatment mmol/L in the intervention (paracetamol) group was 8.92 µmol/L lower (−6.55 to −11.28 lower) than in the ibuprofen group		537 (4 studies)	⊕⊕⊕⊝ moderate	Bias: we had no concerns for random sequence generation in the 4 included trials but the allocation concealment was unclear in 1 of the studies. We did not downgrade the quality of evidence on this item. There were concerns about blinding of personnel and of blinding of outcome assessments. However, for an objective outcome of serum creatinine level we have not downgraded the evidence. Heterogeneity/Consistency: there was high heterogeneity (I² = 84%) for WMD. We downgraded the quality of the evidence by 1 step. Directness of evidence: studies were conducted in the target population. Precision: because of the relatively large sample size (537 infants), the point estimate was precise with a narrow 95% CI. Presence of publication bias: As only 4 studies were included we did not construct a funnel plot.
*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% CI) 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
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.