Interventions for patent ductus arteriosus (PDA) in preterm infants: an overview of Cochrane Systematic Reviews

Abstract Background Patent ductus arteriosus (PDA) is associated with significant morbidity and mortality in preterm infants. Several non‐pharmacological, pharmacological, and surgical approaches have been explored to prevent or treat a PDA. Objectives To summarise Cochrane Neonatal evidence on interventions (pharmacological or surgical) for the prevention of PDA and related complications, and interventions for the management of asymptomatic and symptomatic PDA in preterm infants. Methods We searched the Cochrane Database of Systematic Reviews on 20 October 2022 for ongoing and published Cochrane Reviews on the prevention and treatment of PDA in preterm (< 37 weeks' gestation) or low birthweight (< 2500 g) infants. We included all published Cochrane Reviews assessing the following categories of interventions: pharmacological therapy using prostaglandin inhibitor drugs (indomethacin, ibuprofen, and acetaminophen), adjunctive pharmacological interventions, invasive PDA closure procedures, and non‐pharmacological interventions. Two overview authors independently checked the eligibility of the reviews retrieved by the search, and extracted data from the included reviews using a predefined data extraction form. Any disagreements were resolved by discussion with a third overview author. Two overview authors independently assessed the methodological quality of the included reviews using the AMSTAR 2 (A MeaSurement Tool to Assess systematic Reviews) tool. We reported the GRADE certainty of evidence as assessed by the respective review authors using summary of findings tables. Main results We included 16 Cochrane Reviews, corresponding to 138 randomised clinical trials (RCT) and 11,856 preterm infants, on the prevention and treatment of PDA in preterm infants. One of the 16 reviews had no included studies, and therefore, did not contribute to the results. Six reviews reported on prophylactic interventions for the prevention of PDA and included pharmacological prophylaxis with prostaglandin inhibitor drugs, prophylactic surgical PDA ligation, and non‐pharmacologic interventions (chest shielding during phototherapy and restriction of fluid intake); one review reported on the use of indomethacin for the management of asymptomatic PDA; nine reviews reported on interventions for the management of symptomatic PDA, and included pharmacotherapy with prostaglandin inhibitor drugs in various routes and dosages, surgical PDA ligation, and adjunct therapies (use of furosemide and dopamine in conjunction with indomethacin). The quality of reviews varied. Two reviews were assessed to be high quality, seven reviews were of moderate quality, five of low quality, while two reviews were deemed to be of critically low quality. For prevention of PDA, prophylactic indomethacin reduces severe intraventricular haemorrhage (IVH; relative risk (RR) 0.66, 95% confidence interval (CI) 0.53 to 0.82; 14 RCTs, 2588 infants), and the need for invasive PDA closure (RR 0.51, 95% CI 0.37 to 0.71; 8 RCTs, 1791 infants), but it does not appear to affect the composite outcome of death or moderate/severe neurodevelopmental disability (RR 1.02, 95% CI 0.90 to 1.15; 3 RCTs, 1491 infants). Prophylactic ibuprofen probably marginally reduces severe IVH (RR 0.67, 95% CI 0.45 to 1.00; 7 RCTs, 925 infants; moderate‐certainty evidence), and the need for invasive PDA closure (RR 0.46, 95% CI 0.22 to 0.96; 7 RCTs, 925 infants; moderate‐certainty evidence). The evidence is very uncertain on the effect of prophylactic acetaminophen on severe IVH (RR 1.09, 95% CI 0.07 to 16.39; 1 RCT, 48 infants). Necrotising enterocolitis (NEC) was lower with both prophylactic surgical ligation (RR 0.25, 95% CI 0.08 to 0.83; 1 RCT, 84 infants), and fluid restriction (RR 0.43, 95% CI 0.21 to 0.87; 4 RCTs, 526 infants). For treatment of asymptomatic PDA, indomethacin appears to reduce the development of symptomatic PDA post‐treatment (RR 0.36, 95% CI 0.19 to 0.68; 3 RCTs, 97 infants; quality of source review: critically low). For treatment of symptomatic PDA, all available prostaglandin inhibitor drugs appear to be more effective in closing a PDA than placebo or no treatment (indomethacin: RR 0.30, 95% CI 0.23 to 0.38; 10 RCTs, 654 infants; high‐certainty evidence; ibuprofen: RR 0.62, 95% CI 0.44 to 0.86; 2 RCTs, 206 infants; moderate‐certainty evidence; early administration of acetaminophen: RR 0.35, 95% CI 0.23 to 0.53; 2 RCTs, 127 infants; low‐certainty evidence). Oral ibuprofen appears to be more effective in PDA closure than intravenous (IV) ibuprofen (RR 0.38, 95% CI 0.26 to 0.56; 5 RCTs, 406 infants; moderate‐certainty evidence). High‐dose ibuprofen appears to be more effective in PDA closure than standard‐dose ibuprofen (RR 0.37, 95% CI 0.22 to 0.61; 3 RCTs, 190 infants; moderate‐certainty evidence). With respect to adverse outcomes, compared to indomethacin administration, NEC appears to be lower with ibuprofen (any route; RR 0.68, 95% CI 0.49 to 0.94; 18 RCTs, 1292 infants; moderate‐certainty evidence), oral ibuprofen (RR 0.41, 95% CI 0.23 to 0.73; 7 RCTs, 249 infants; low‐certainty evidence), and with acetaminophen (RR 0.42, 95% CI 0.19 to 0.96; 4 RCTs, 384 infants; low‐certainty evidence). However, NEC appears to be increased with a prolonged course of indomethacin versus a shorter course (RR 1.87, 95% CI 1.07 to 3.27; 4 RCTs, 310 infants). Authors' conclusions This overview summarised the evidence from 16 Cochrane Reviews of RCTs regarding the effects of interventions for the prevention and treatment of PDA in preterm infants. Prophylactic indomethacin reduces severe IVH, but does not appear to affect the composite outcome of death or moderate/severe neurodevelopmental disability. Prophylactic ibuprofen probably marginally reduces severe IVH (moderate‐certainty evidence), while the evidence is very uncertain on the effect of prophylactic acetaminophen on severe IVH. All available prostaglandin inhibitor drugs appear to be effective in symptomatic PDA closure compared to no treatment (high‐certainty evidence for indomethacin; moderate‐certainty evidence for ibuprofen; low‐certainty evidence for early administration of acetaminophen). Oral ibuprofen appears to be more effective in PDA closure than IV ibuprofen (moderate‐certainty evidence). High dose ibuprofen appears to be more effective in PDA closure than standard‐dose ibuprofen (moderate‐certainty evidence). There are currently two ongoing reviews, one on fluid restriction for symptomatic PDA, and the other on invasive management of PDA in preterm infants.


A B S T R A C T Background
Patent ductus arteriosus (PDA) is associated with significant morbidity and mortality in preterm infants. Several non-pharmacological, pharmacological, and surgical approaches have been explored to prevent or treat a PDA.

Objectives
To summarise Cochrane Neonatal evidence on interventions (pharmacological or surgical) for the prevention of PDA and related complications, and interventions for the management of asymptomatic and symptomatic PDA in preterm infants.

Methods
We searched the Cochrane Database of Systematic Reviews on 20 October 2022 for ongoing and published Cochrane Reviews on the prevention and treatment of PDA in preterm (< 37 weeks' gestation) or low birthweight (< 2500 g) infants. We included all published Cochrane Reviews assessing the following categories of interventions: pharmacological therapy using prostaglandin inhibitor drugs (indomethacin, ibuprofen, and acetaminophen), adjunctive pharmacological interventions, invasive PDA closure procedures, and nonpharmacological interventions. Two overview authors independently checked the eligibility of the reviews retrieved by the search, and extracted data from the included reviews using a predefined data extraction form. Any disagreements were resolved by discussion with a third overview author. Two overview authors independently assessed the methodological quality of the included reviews using the AMSTAR 2 (A MeaSurement Tool to Assess systematic Reviews) tool. We reported the GRADE certainty of evidence as assessed by the respective review authors using summary of findings tables.

Main results
We included 16 Cochrane Reviews, corresponding to 138 randomised clinical trials (RCT) and 11,856 preterm infants, on the prevention and treatment of PDA in preterm infants. One of the 16 reviews had no included studies, and therefore, did not contribute to the results. Six reviews reported on prophylactic interventions for the prevention of PDA and included pharmacological prophylaxis with prostaglandin inhibitor drugs, prophylactic surgical PDA ligation, and non-pharmacologic interventions (chest shielding during phototherapy and restriction of fluid intake); one review reported on the use of indomethacin for the management of asymptomatic PDA; nine reviews reported on interventions for the management of symptomatic PDA, and included pharmacotherapy with prostaglandin inhibitor drugs

Description of the condition
The ductus arteriosus is a blood vessel that connects the main pulmonary artery to the proximal descending aorta. It plays an important role in maintaining foetal circulation by allowing a significant proportion of right ventricular output to bypass the pulmonary circulation (Gournay 2011). Following birth, with establishment of respiration and separation of low-resistance placenta, closure of the ductus arteriosus begins. This closure is triggered by physiological mechanisms, such as increased oxygen tension and decreased circulating prostaglandin (PGE ) and prostacyclin (PGI (Hundscheid 2019)). Functional closure of the ductus arteriosus occurs over the next 24 to 72 hours in term infants (Benitz 2016). In preterm infants, closure is o en delayed, leading to the ductus arteriosus remaining patent beyond the first few days of life. In healthy preterm neonates, born at > 30 weeks' gestation, the patent ductus arteriosus (PDA) closes by day four in 90%, and by day seven in 98% of infants (Clyman 2012). In extremely preterm infants, born at < 24 weeks' gestation, spontaneous PDA closure rates are only about 8% by day four and 13% by day seven (Clyman 2012).
Therefore, a PDA o en persists beyond the first few days of life in a preterm neonate, but may remain asymptomatic, without inducing any adverse haemodynamic consequences in the neonate. However, with progressive decline in pulmonary vascular resistance, blood flow from the aorta into the pulmonary arteries is increased through the PDA. Consequently, the proportion of aortic blood flow that is diverted into the pulmonary circulation is correspondingly increased (Benitz 2016). This 'ductal steal' may result in excessive blood flow through the lungs, predisposing the development of pulmonary congestion, pulmonary oedema, worsening respiratory failure, and eventually, chronic lung disease (CLD (Benitz 2016)). At the same time, diversion of blood flow away from the systemic circulation may lead to systemic hypoperfusion, resulting in compromised perfusion to the bowel, kidney, and brain. When a PDA is associated with clinical and echocardiographic signs of pulmonary hyperperfusion and systemic hypoperfusion, this is labelled a symptomatic PDA, or a haemodynamically significant PDA. A persistently symptomatic PDA may be associated with numerous adverse outcomes, including higher rates of death (Dice 2007), bronchopulmonary dysplasia (BPD (Brown 1979)), necrotising enterocolitis (NEC (Dollberg 2005)), impaired renal function (Benitz 2016), intraventricular haemorrhage (IVH (Ballabh 2010)), periventricular leukomalacia (PVL (Chung 2005)), and cerebral palsy (Drougia 2007). However, the causal link between these associations has not been demonstrated (Benitz 2010).

Description of the interventions
A PDA can be closed through medical or surgical interventions. Pharmacotherapeutic agents include non-steroidal anti-inflammatory drugs (NSAIDs), such as ibuprofen or indomethacin, and acetaminophen, which is a derivative of acetanilide with anti-inflammatory properties. Surgical interventions include surgical ligation and transcatheter occlusion. In this overview, we will focus on both pharmacotherapeutic and surgical interventions for prevention and treatment of preterm infants with PDA.
NSAIDs act by inhibiting the cyclo-oxygenase (COX) enzyme, thereby leading to down regulation of PGE , a potent relaxant of the PDA (Mahony 1982). However, use of indomethacin in preterm infants has been associated with transient or permanent derangement of renal function (Seyberth 1983), NEC (Coombs 1990), gastrointestinal haemorrhage or perforation (Wolf 1989), alteration of platelet function (Friedman 1976), and impairment of cerebral blood flow/cerebral blood flow velocity (Ohlsson 1993). Therefore, variations in indomethacin therapy have been attempted to mitigate the said adverse e ects while maximising therapeutic benefit. These include using continuous infusion of indomethacin rather than intermittent bolus doses, which may reduce its adverse e ects on cerebral oxygenation (Hammerman 1995), and use of a prolonged course of indomethacin, which may provide increased therapeutic benefit compared to a short course of indomethacin (Rennie 1991).
Ibuprofen exerts its action through inhibition of the COX enzyme, but appears to be associated with lower risk of NEC and transient renal insu iciency, compared to indomethacin (Ohlsson 2020b). However, variation in dosage and route of administration of ibuprofen may impact medication e ectiveness. It has been demonstrated that to achieve optimal concentrations of ibuprofen for successful PDA closure, irrespective of gestational age, progressively higher doses are required with increasing postnatal age (Hirt 2008). Similar pharmacokinetic studies have shown that peak serum concentrations following oral ibuprofen therapy are significantly higher than previously demonstrated intravenous levels, suggesting a potential for greater responsiveness to oral ibuprofen compared to the intravenous formulation (Barzilay 2012).
Acetaminophen is postulated to exert its action through inhibition of the peroxidase enzyme, thereby leading to down regulation of PGE production (Gillam-Krakauer 2018;Grèen 1989). No shortterm adverse e ects have been noted with acetaminophen. However, data on the safety and long-term neurodevelopmental e ects of acetaminophen in preterm infants are limited (Ohlsson 2020; Van den Anker 2018).
With increasing emphasis on conservative management, surgical PDA ligation is primarily reserved for infants with persistent symptomatic PDA following the failure of medical management. Surgical PDA ligation is associated with reduced mortality, but surviving infants were found to be at increased risk of neurodevelopmental impairment, which could be due to lack of studies addressing survival bias and confounding by indication (Weisz 2014).
reduce the incidence of symptomatic PDA, surgical PDA ligation, and the incidence of severe intraventricular haemorrhage, but has no e ect on mortality, nor on a composite of death or severe neurodevelopmental disability, compared to placebo or no treatment (Fowlie 2010). Prophylactic ibuprofen, compared to placebo or no intervention, has also been shown to reduce the need for rescue treatment with COX inhibitors, and for surgical PDA closure (Ohlsson 2020a). However, both prophylactic indomethacin and prophylactic ibuprofen have been shown to be associated with increased risk of oliguria (Fowlie 2010; Ohlsson 2020a). Prophylactic ibuprofen is further associated with increased risk of gastrointestinal haemorrhage (Ohlsson 2020a). Therefore, interest in expectant management of the PDA in preterm infants is growing, and the safety of this approach remains to be established through large randomised controlled trials (Hundscheid 2018).
On the other hand, treatment of PDA entails pharmacotherapeutic or surgical closure of a PDA, the diagnosis of which was based on characteristic clinical symptoms, echocardiographic findings, or both. NSAIDs and acetaminophen have been shown to be more e ective in closing a symptomatic PDA compared to placebo (Mitra 2018). Ibuprofen appears to be as e ective as indomethacin in closing a symptomatic PDA, while reducing the risk of NEC and transient renal insu iciency (Ohlsson 2020b). Moderate-quality evidence shows that acetaminophen is as e ective as ibuprofen, and low-quality evidence suggests that acetaminophen is as e ective as indomethacin in closing a symptomatic PDA (Ohlsson 2020). Data are inconclusive regarding the e icacy and safety of surgery as the initial modality of treatment for a symptomatic PDA in a preterm infant compared to pharmacotherapeutic management (Malviya 2013).

Why it is important to do this overview
Management of the PDA is one of the most controversial topics in neonatal medicine. Prophylactic treatment with indomethacin reduces the need for surgical PDA ligation and severe periventricular and intraventricular haemorrhage, but does not improve the rate of survival without neurosensory impairment at 18 months (Schmidt 2001). There are also concerns about the increased incidence of spontaneous gastrointestinal perforation with prophylactic indomethacin (Stavel 2017). On the other hand, prophylactic use of ibuprofen for PDA in preterm infants has been associated with severe hypoxaemia, pulmonary hypertension, and gastrointestinal haemorrhage (Gournay 2002). Therefore, debate on whether NSAIDs should be routinely used to prevent PDA in preterm infants is ongoing. To date, four Cochrane Neonatal Reviews have examined prophylactic medical or surgical management of PDA in preterm infants (Fowlie 2010;Mosalli 2008;Ohlsson 2020;Ohlsson 2020a). There is also debate about whether treatment of an asymptomatic PDA before the development of a significant le -to-right shunt improves clinical outcomes. One Cochrane Neonatal Review explored the question of treatment for asymptomatic PDA (Cooke 2003). Similarly, when it comes to treatment for a symptomatic PDA, the availability of multiple management strategies contributes to the dilemma among clinicians. In a recent systematic review and network metaanalysis, 15 di erent pharmacotherapeutic options were identified that have been explored in randomised clinical trials for the management of symptomatic PDA (Mitra 2018). The Cochrane Reviews published so far on this topic tackled the problem from a narrow perspective, as all of them compared only two out of several available interventions against each other. Some of these reviews lacked an assessment of the quality of the evidence, using GRADE, and reviews showed variation in the definitions of symptomatic PDA, interventions, and outcomes described. Therefore, an overview of available Cochrane Neonatal Reviews was justified, as it helped to summarise the evidence generated so far on the management strategies available for PDA in preterm infants with respect to the most important outcomes, including the quality of the evidence, and also highlighted important gaps in knowledge that may guide future research on PDA management.

Is an overview the right approach?
We followed the Editorial Decision Tree proposed by the Cochrane Comparing Multiple Intervention Methods Group to establish whether our review would better fit an overview format or an intervention review format. We decided that for the purposes of this review, to (Shepherd 2018): 1. review only systematic reviews published in the Cochrane Database of Systematic Reviews, instead of individual trials; 2. not compare multiple interventions with the intention of drawing inferences about the comparative e ectiveness of these interventions, as we cannot draw conclusions on the transitivity assumption from systematic reviews only; and 3. present a map of evidence from systematic reviews, but with no attempt to rank the interventions.
On the basis of these points, the Editorial Decision Tree recommended that an overview was the appropriate format for this review.

O B J E C T I V E S
To summarise Cochrane Neonatal evidence on: 1. interventions (pharmacological or surgical) for prevention of patent ductus arteriosus and related complications in preterm infants; and 2. interventions (pharmacological or surgical) for management of patent ductus arteriosus in preterm infants, including: a. interventions for management of asymptomatic patent ductus arteriosus in preterm infants; and b. interventions for management of symptomatic (haemodynamically significant) patent ductus arteriosus in preterm infants.

Types of studies
In this overview of systematic reviews, we included only published Cochrane Systematic Reviews on the management of patent ductus arteriosus (PDA) in a preterm infant.

Library
Trusted evidence. Informed decisions. Better health.

For objective 2 (management of PDA)
Preterm (gestational age < 37 weeks at birth) or low-birthweight infants (< 2500 g) with PDA diagnosed clinically, or via echocardiography, or both, in the neonatal period (< 28 days).
We defined an asymptomatic PDA clinically by the presence of a precordial murmur or echocardiographically (presence of leto-right PDA shunt), without clinical signs of a moderate-to high-volume le -to-right shunt (hyperdynamic precordial impulse, tachycardia, bounding pulses, widened pulse pressure, worsening respiratory status, hypotension, or cardiac failure).
We defined a symptomatic PDA clinically by the presence of a precordial murmur, along with one or more of the following signs: hyperdynamic precordial impulse, tachycardia, bounding pulses, widened pulse pressure, worsening respiratory status, hypotension, or cardiac failure. We defined a symptomatic PDA echocardiographically by a moderate to large transductal diameter, with or without evidence of pulmonary over-circulation, with or without evidence of systemic hypoperfusion. We also defined a symptomatic PDA as a combination of le -to-right PDA shunt on echocardiography, along with clinical signs of a high-volume le -to-right shunt (hyperdynamic precordial impulse, tachycardia, bounding pulses, widened pulse pressure, worsening respiratory status, hypotension, or cardiac failure).

Types of interventions
In this overview, we specifically included reviews of therapies primarily intended to prevent or manage a PDA.

For objective 1 (prevention of PDA)
Interventions included prophylactic (not guided by knowledge of PDA status) pharmacological or surgical treatment of PDA within 24 hours of birth. Pharmacological treatments included indomethacin, ibuprofen, and acetaminophen compared against each other, or placebo, or no treatment. There were no restrictions on dose, route, or duration of treatment. Surgical interventions included surgical or transcatheter PDA closure compared against medical treatment, or placebo, or no treatment.

For objective 2 (management of PDA)
Interventions included pharmacological and surgical treatments for an asymptomatic or a symptomatic PDA. Pharmacological treatments included indomethacin, ibuprofen, and acetaminophen compared against each other, or placebo, or no treatment. There were no restrictions on dose, route, or duration of treatment. Surgical interventions included surgical or transcatheter PDA closure compared against medical treatment, or placebo, or no treatment.

For objective 1 (prevention of PDA)
Primary outcomes 1. Severe intraventricular haemorrhage (IVH; grade III/IV (Papile 1978)) 2. Death or moderate/severe neurodevelopmental disability (assessed by a standardised and validated assessment tool, a child developmental specialist, or both) at any age reported (outcome data grouped at 12, 18, and 24 months, if available).
Individual components of a neurodevelopmental outcome, defined in individual reviews, were reported if data were available.

PDA-related outcomes
1. Symptomatic PDA confirmed on echocardiogram 2. Proportion of infants receiving open-label medical treatment (cyclo-oxygenase inhibitor or paracetamol/acetaminophen dosing, or both) 3. Proportion of infants requiring surgical ligation or transcatheter occlusion Other outcomes 1. Chronic lung disease (CLD), defined as oxygen requirement at 36 weeks' postmenstrual age (Ehrenkranz 2005) 2. Intraventricular haemorrhage (IVH; grade I to IV (Papile 1978)) 3. Pulmonary haemorrhage, defined as blood-stained respiratory secretions with a significant change in respiratory requirements and chest X-ray (CXR) changes in the presence of echocardiographic evidence of significant le -to-right ductal shunting (Kluckow 2014) 4. Retinopathy of prematurity (ROP), defined according to the international classification of ROP (ICCROP 2005) 5. Duration of hospitalisation, defined as total length of hospitalisation from birth to discharge home or mortality, in days 6. Moderate/severe neurodevelopmental disability, assessed by a standardised and validated assessment tool, a child developmental specialist, or both, at any age reported (outcome data grouped at 12, 18, and 24 months, if available). Individual components of a neurodevelopmental outcome were reported if data were available. 7. All-cause mortality any time before neonatal intensive care unit (NICU) discharge

Safety outcomes
1. Necrotising enterocolitis (NEC; stage 2 or greater (Bell 1978)) 2. Gastrointestinal perforation, defined by the presence of free air in peritoneal cavity on an abdominal X-ray (Ohlsson 2020a) 3. Gastrointestinal bleeding within seven days of the first dose of pharmacotherapy 4. Oliguria, defined as less than 1 mL/kg/hour 5. Serum/plasma levels of creatinine (μmol/L) a er treatment 6. Increase in serum/plasma levels of creatinine (μmol/L) a er treatment 7. Serum/plasma levels of bilirubin (μmol/L) a er treatment 8. Increase in serum/plasma levels of bilirubin (μmol/L) a er treatment 1. Necrotising enterocolitis (NEC; stage 2 or greater (Bell 1978)) 2. Gastrointestinal perforation, defined by the presence of free air in peritoneal cavity on an abdominal X-ray (Ohlsson 2020a) 3. Gastrointestinal bleeding within seven days of the first dose of pharmacotherapy 4. Oliguria, defined as less than 1 mL/kg/hour 5. Serum/plasma levels of creatinine (μmol/L) a er treatment 6. Increase in serum/plasma levels of creatinine (μmol/L) a er treatment 7. Serum/plasma levels of bilirubin (μmol/L) a er treatment 8. Increase in serum/plasma levels of bilirubin (μmol/L) a er treatment

Search methods for identification of reviews
We searched the Cochrane Database of Systematic Reviews, using the term 'patent ductus arteriosus', on 20 October 2022. We used the search term to search 'all text', not limited to 'title, abstract, or keywords'. We did not apply any language or date restrictions. We did not search any other databases.

Data collection and analysis
We used the standard methods of Cochrane Neonatal.

Selection of reviews
Two overview authors (SM and DW) independently assessed for inclusion, all potential systematic reviews identified by the search. Disagreements were resolved through discussion, or if required, a third member of the overview team was consulted (PS).

Data extraction and management
Two overview authors (SM and DW) independently extracted data from the reviews, using a standardised form developed in Microso Excel. Discrepancies were resolved through discussion, or if needed, through consultation with a third overview author (PS).
In the event information regarding review outcomes was unclear or missing, individual studies were accessed for further details.
We extracted data on the following.
1. Review characteristics. a. Review title and authors b. Date that the review was last assessed as up-to-date c. Number of included trials and numbers of participants in the trials and their characteristics d. Risk of bias of the included trials, as reported by the review authors; see Quality of studies included within reviews, under Assessment of methodological quality of included reviews e. Interventions and comparisons relevant to this overview f. All prespecified outcomes relevant to this overview (their definitions, and whether they were primary or secondary outcomes in the included reviews) g. Any other characteristics required to assess and report on review quality; see Quality of included reviews, under Assessment of methodological quality of included reviews 2. Statistical summaries.
a. Summary intervention e ects, including pooled e ects (e.g. risk ratios (RRs), odds ratios (ORs), mean di erences (MDs), as reported in the individual reviews), 95% confidence intervals (CIs), numbers of studies and participants contributing data to each pooled e ect, from comparisons, and for outcomes relevant to this overview, including relevant subgroup analyses b. Information required to assess and report on the quality of evidence for the intervention e ects extracted; see Quality of evidence in included reviews, under Assessment of methodological quality of included reviews

Assessment of methodological quality of included reviews
We assessed the methodological quality of each systematic review using the updated AMSTAR 2 (A Measurement Tool to Assess Reviews) instrument (Shea 2017). AMSTAR 2 evaluates the methods used in a review against 16 distinct criteria and assesses the degree to which review methods are unbiased. These criteria are as follows.
1. Did the research questions and inclusion criteria for the review include the components of PICO (Participants, Intervention, Comparison, Outcomes)? 2. Did the report of the review contain an explicit statement that the review methods were established prior to the conduct of the review, and did the report justify any significant deviations from the protocol? 3. Did the review authors explain their selection of the study designs for inclusion in the review? 4. Did the review authors use a comprehensive literature search strategy? 5. Did the review authors perform study selection in duplicate? 6. Did the review authors perform data extraction in duplicate? 7. Did the review authors provide a list of excluded studies and justify the exclusions? 8. Did the review authors describe the included studies in adequate detail? 9. Did the review authors use a satisfactory technique for assessing the risk of bias (RoB) in individual studies that were included in the review? 10.Did the review authors report on the sources of funding for the studies included in the review? 11.If meta-analysis was performed, did the review authors use appropriate methods for statistical combination of results? 12.If meta-analysis was performed, did the review authors assess the potential impact of RoB in individual studies on the results of the meta-analysis or other evidence synthesis? 13.Did the review authors account for RoB in individual studies when interpreting/discussing the results of the review? 14.Did the review authors provide a satisfactory explanation for, and discussion of, any heterogeneity observed in the results of the review? 15.If they performed quantitative synthesis, did the review authors carry out an adequate investigation of publication bias (smallstudy bias) and discuss its likely impact on the results of the review? 16.Did the review authors report any potential sources of conflict of interest, including any funding they received for conducting the review?
Two overview authors (SM and PS) independently assessed the quality of the included reviews using the online AMSTAR 2 tool (Shea 2017). A third overview author (WdB) verified the assessment. We resolved di erences through discussion.

Quality of included studies within reviews
We did not reassess the risk of bias of included studies within reviews. Instead, we reported study quality according to the review authors' assessment. When individual studies were included in two or more Cochrane Reviews, we reported any variation in the review authors' assessments of study quality.

Certainty of evidence in included reviews
We used the GRADE approach, as outlined in the GRADE Handbook to assess the certainty of evidence for the following (clinically relevant) outcomes (Schünemann 2013). We reported the certainty of evidence as assessed by the review authors (who were in the best position to assess certainty given their familiarity with the study level data), using summary of findings tables from the reviews if provided.
The GRADE approach results in an assessment of the certainty of a body of evidence as one of four grades.
1. High certainty: further research is very unlikely to change our confidence in the estimate of e ect 2. Moderate certainty: further research is likely to have an important impact on our confidence in the estimate of e ect, and may change the estimate 1. Gestational age (less than 28 weeks, 28 weeks or more) 2. Birthweight (less than 1000 g, 1000 g or more) 3. Timing of initiation of treatment for asymptomatic PDA (less than 24 hours, 24 hours or longer) 4. Timing of initiation of treatment for symptomatic PDA (less than 72 hours, 72 hours or longer) 5. Method used to diagnose a symptomatic PDA (by echocardiographic criteria or only by clinical criteria) 6. Degree of haemodynamic significance of the PDA (based on echocardiographic criteria)

R E S U L T S
Our search (October 2022) identified 17 relevant Cochrane Reviews and two Cochrane Review protocols under review ( Figure 1). Out of these 19 reviews, we included 16 reviews in this overview. The review by Mitra 2022 was identified to be a Bayesian network metaanalysis of data obtained from studies, most of which were already included in the reviews of the respective pharmacoprophylactic interventions (Fowlie 2010; Jasani 2022; Ohlsson 2020a). Since the latter three reviews were already included in this overview, and to avoid duplication while summarising the results, we excluded the network meta-analysis by Mitra 2022. One of the 16 reviews had no included studies and therefore did not contribute to the results (Anabrees 2011). Cochrane Database of Systematic Reviews

Description of included reviews
We included the following reviews in this overview (Table 1). gestation or < 1000 g at birth who were on assisted ventilation, or supplemental oxygen, or both, without clinical signs of a haemodynamically significant PDA, and compared prophylactic surgical ligation of the PDA (i.e. procedure done during the first 72 hours) versus no prophylactic intervention or medical prophylaxis (cyclooxygenase inhibitors) without dose specification. 15.Ohlsson 2020a (9 RCTs, 1070 infants) included preterm infants (< 37 weeks' gestational age) and low birthweight infants (< 2500 g) in their first 72 hours of life (three days), and compared prophylactic use of ibuprofen for prevention of PDA versus control, consisting of no intervention, placebo, other cyclooxygenase inhibitor drugs (indomethacin, mefenamic acid), or rescue treatment with ibuprofen. 16.Ohlsson 2020b (39 RCTs, 2843 infants) included preterm infants (< 37 weeks' gestational age) or low birthweight infants (< 2500 g) with a PDA, diagnosed either clinically or by echocardiography in the neonatal period (less than 28 days), and compared ibuprofen (in di erent routes and dosages) versus indomethacin, other cyclo-oxygenase inhibitor(s), placebo, or no intervention.

Methodological quality of included reviews
The AMSTAR 2 assessment of the quality of the included reviews is presented in Table 2. Risk of bias in the included trials, as assessed by the respective review authors, is reported in Table 3. The certainty of the evidence for the primary outcomes of this overview (as available from the respective reviews) is summarised in Table 4 and Table 5.

Interventions (pharmacological or surgical) for prevention of PDA and related complications in preterm infants
The results for the following outcomes are summarised in Table 6 Severe intraventricular haemorrhage (IVH; grade III/IV) Five Cochrane Reviews reported on the outcome of severe IVH. They included the following interventions.

Prophylactic acetaminophen
Prophylactic acetaminophen versus placebo: the review by Jasani 2022 showed that there was no evidence of a di erence between prophylactic acetaminophen and placebo or no intervention for severe IVH (RR 1.09, 95% CI 0.07 to 16.39; 1 RCT, 48 infants).

Prophylactic surgical PDA ligation
Prophylactic surgical PDA ligation versus control (prophylactic cyclooxygenase inhibitor drugs only): the review by Mosalli 2008 showed that there was no evidence of a di erence between prophylactic surgical PDA ligation and control (prophylactic cyclooxygenase inhibitor drugs) for severe IVH (RR 0.81, 95% CI 0.52 to 1.28; 1 RCT, 76 infants).

Chest shielding during phototherapy
Chest shielding during phototherapy versus control: the review by Bhola 2015 showed that there was no evidence of a di erence between chest shielding during phototherapy and control for severe IVH (RR 0.64, 95% CI 0.22 to 1.85; 2 RCTs, 128 infants).

Death or moderate/severe neurodevelopmental disability
Only one Cochrane Review reported on the composite outcome of death or moderate/severe neurodevelopmental disability. It included the following intervention.

Prophylactic indomethacin
Prophylactic indomethacin versus placebo: the review by Fowlie 2010 showed that there was no evidence of a di erence between prophylactic indomethacin and control for the composite outcome of death or moderate/severe neurodevelopmental disability (RR 1.02, 95% CI 0.90 to 1.15; 3 RCTs, 1491 infants).

PDA confirmed on echocardiogram
Five Cochrane Reviews reported on the outcome of echocardiogram-confirmed PDA post-prophylaxis. They included the following interventions.

Prophylactic ibuprofen
Prophylactic ibuprofen versus placebo: the review by Ohlsson 2020a showed that compared to placebo or no intervention, prophylactic ibuprofen reduced the presence of PDA (RR 0.39, 95% CI 0.31 to 0.48; 9 RCTs, 1029 infants; moderate-certainty evidence).

Prophylactic acetaminophen
Prophylactic acetaminophen versus placebo: the review by Jasani 2022 showed that compared to placebo or no intervention, prophylactic acetaminophen reduced the presence of PDA (RR 0.27, 95% CI 0.18 to 0.42; 3 RCTs, 240 infants).

Chest shielding during phototherapy
Chest shielding during phototherapy versus control: the review by Bhola 2015 showed that there was no evidence of a di erence between chest shielding during phototherapy and control for any PDA detected by echocardiography (RR 0.92, 95% CI 0.52 to 1.64; 1 RCT, 54 infants), or detection of a haemodynamically significant PDA (RR 0.23, 95% CI 0.05 to 1.01; 1 RCT, 74 infants).

Restriction of fluid intake
Restricted versus liberal fluid intake: the review by Bell 2014 showed that compared to liberal fluid intake, restriction of predominantly intravenous (IV) fluids reduced the persistence of PDA (RR 0.52, 95% CI 0.37 to 0.73; 4 RCTs, 526 infants).

Proportion of infants receiving open-label medical treatment
Two Cochrane Reviews reported on the outcome of receipt of open-label medical treatment for PDA. They included the following interventions.

Library
Trusted evidence. Informed decisions. Better health.

Proportion of infants requiring surgical ligation or transcatheter occlusion
Three Cochrane Reviews reported on the outcome invasive PDA closure by surgical ligation or transcatheter occlusion. They included the following interventions.

Prophylactic indomethacin
Prophylactic indomethacin versus placebo: the review by Fowlie 2010 showed that compared to control, prophylactic indomethacin reduced the need for invasive PDA closure (RR 0.51, 95% CI 0.37 to 0.71; 8 RCTs, 1791 infants).

Prophylactic ibuprofen
Prophylactic ibuprofen versus placebo: the review by Ohlsson 2020a showed that compared to placebo or no intervention, prophylactic ibuprofen reduced the need for invasive PDA closure (RR 0.46, 95% CI 0.22 to 0.96; 7 RCTs, 925 infants; moderatecertainty evidence).

Chest shielding during phototherapy
Chest shielding during phototherapy versus control: the review by Bhola 2015 showed that there was no evidence of a di erence between chest shielding during phototherapy and control for invasive PDA closure (RR 0.35, 95% CI 0.01 to 8.36; 1 RCT, 74 infants).

Chronic lung disease (CLD)
Five Cochrane Reviews reported on the outcome of CLD (all definitions included). They included the following interventions.

Prophylactic indomethacin
Prophylactic indomethacin versus placebo: the review by Fowlie 2010 showed that there was no evidence of a di erence between prophylactic indomethacin and control for CLD, defined as oxygen requirement at 28 postnatal days (RR 1.08, 95% CI 0.92 to 1.26; 9 RCTs, 1022 infants), or CLD, defined as oxygen requirement at 36 weeks' postmenstrual age (RR 1.06, 95% CI 0.92 to 1.22; 1 RCT, 999 infants).

Prophylactic ibuprofen
Prophylactic ibuprofen versus placebo: the review by Ohlsson 2020a showed that there was no evidence of a di erence between prophylactic ibuprofen and placebo or no intervention for CLD, defined as oxygen requirement at 28 postnatal days (

Restriction of fluid intake
Restricted versus liberal fluid intake: the review by Bell 2014 showed that there was no evidence of a di erence between restriction of predominantly IV fluids and liberal fluid intake for CLD (no definition specified for CLD; RR 0.85, 95% CI 0.63 to 1.14; 4 RCTs, 526 infants).

Intraventricular haemorrhage (IVH; grades I to IV)
Four Cochrane Reviews reported on the outcome of IVH (grades I to IV). They included the following interventions.

Prophylactic ibuprofen
Prophylactic ibuprofen versus placebo: the review by Ohlsson 2020a showed that there was no evidence of a di erence between prophylactic ibuprofen and placebo or no intervention for all grades of IVH (RR 0.96, 95% CI 0.78 to 1.17; 6 RCTs, 901 infants).

Chest shielding during phototherapy
Chest shielding during phototherapy versus control: the review by Bhola 2015 showed that there was no evidence of a di erence between chest shielding during phototherapy and control for all grades of IVH (RR 0.53, 95% CI 0.10 to 2.71; 1 RCT, 74 infants).

Restriction of fluid intake
Restricted versus liberal fluid intake: the review by Bell 2014 showed that there was no evidence of a di erence between restriction of predominantly IV fluids and liberal fluid intake with respect to all grades of IVH (RR 0.74, 95% CI 0.48 to 1.14; 3 RCTs, 356 infants).

Pulmonary haemorrhage
One Cochrane Review reported on the outcome pulmonary haemorrhage. It included the following intervention.

Prophylactic indomethacin
Prophylactic indomethacin versus placebo: the review by Fowlie 2010 showed that there was no evidence of a di erence between prophylactic indomethacin and control for pulmonary haemorrhage (RR 0.84, 95% CI 0.66 to 1.07; 4 RCTs, 1591 infants).

Retinopathy of prematurity (ROP)
Five Cochrane Reviews reported on the outcome of ROP. They included the following interventions.

Library
Trusted evidence. Informed decisions. Better health.

Prophylactic indomethacin
Prophylactic indomethacin versus placebo: the review by Fowlie 2010 showed that there was no evidence of a di erence between prophylactic indomethacin and control for any stage of ROP (RR 1.02, 95% CI 0.92 to 1.12; 5 RCTs, 1571 infants), or for severe ROP (RR 1.75, 95% CI 0.92 to 3.34; 2 RCTs, 289 infants).

Prophylactic ibuprofen
Prophylactic ibuprofen versus placebo: the review by Ohlsson 2020a showed that there was no evidence of a di erence between prophylactic ibuprofen and placebo or no intervention for ROP (RR 1.01, 95% CI 0.73 to 1.38; 5 RCTs, 369 infants).

Prophylactic acetaminophen
Prophylactic acetaminophen versus placebo: the review by Jasani 2022 showed that there was no evidence of a di erence between prophylactic acetaminophen and placebo or no intervention for ROP (defined as any ROP that required treatment; RR 3.25, 95% CI 0.14 to 76.01; 1 RCT, 48 infants).

Chest shielding during phototherapy
Chest shielding during phototherapy versus control: the review by Bhola 2015 showed that there was no evidence of a di erence between chest shielding during phototherapy and control for any stage of ROP (RR 0.53, 95% CI 0.10 to 2.71; 1 RCT, 74 infants).

Duration of hospitalisation (days)
Two Cochrane Reviews reported on duration of hospitalisation. They included the following interventions.

Prophylactic ibuprofen
Prophylactic ibuprofen versus placebo: the review by Ohlsson 2020a showed that there was no evidence of a di erence between prophylactic ibuprofen and placebo or no intervention for duration of hospitalisation (mean di erence (MD) 1.30 days, 95% CI -3.07 to 5.67; 6 RCTs, 447 infants).

Chest shielding during phototherapy
Chest shielding during phototherapy versus control: the review by Bhola 2015 showed that there was no evidence of a di erence between chest shielding during phototherapy and control for duration of hospitalisation (MD -8.05 days, 95% CI -18.04 to 1.94; 2 RCTs, 128 infants).

Moderate/severe neurodevelopmental disability
One Cochrane Review reported on the outcome of moderate/ severe neurodevelopmental disability. It included the following intervention.

Prophylactic indomethacin
Prophylactic indomethacin versus placebo: the review by Fowlie 2010 showed that there was no evidence of a di erence between prophylactic indomethacin and control for moderate/ severe neurodevelopmental disability (RR 0.96, 95% CI 0.79 to 1.17; 3 RCTs, 1286 infants).

All-cause mortality
Six Cochrane Reviews reported on mortality (all time points included). They included the following interventions.

Restriction of fluid intake
Restricted versus liberal fluid intake: the review by Bell 2014 showed that there was no evidence of a di erence between restriction of predominantly IV fluids and liberal fluid intake for mortality before discharge (RR 0.81, 95% CI 0.54 to 1.23; 5 RCTs, 582 infants).

Library
Trusted evidence. Informed decisions. Better health.

Necrotising enterocolitis (NEC)
Five Cochrane Reviews reported on NEC. They included the following interventions.

Prophylactic indomethacin
Prophylactic indomethacin versus placebo: the review by Fowlie 2010 showed that there was no evidence of a di erence between prophylactic indomethacin and control for NEC (RR 1.09, 95% CI 0.82 to 1.46; 12 RCTs, 2401 infants).

Prophylactic ibuprofen
Prophylactic ibuprofen versus placebo: the review by Ohlsson 2020a showed that there was no evidence of a di erence between prophylactic ibuprofen and placebo or no intervention for NEC (RR 0.96, 95% CI 0.61 to 1.50; 9 RCTs, 1028 infants; moderate-certainty evidence).

Prophylactic acetaminophen
Prophylactic acetaminophen versus placebo: the review by Jasani 2022 showed that there was no evidence of a di erence between prophylactic acetaminophen and placebo or no intervention for NEC (RR 0.36, 95% CI 0.02 to 8.45; 1 RCT, 48 infants).

Restriction of fluid intake
Restricted versus liberal fluid intake: the review by Bell 2014 showed that compared to liberal fluid intake, restriction of predominantly IV fluids reduced NEC (RR 0.43, 95% CI 0.21 to 0.87; 4 RCTs, 526 infants).

Gastrointestinal perforation
Two Cochrane Reviews reported on gastrointestinal perforation. They included the following interventions.

Prophylactic indomethacin
Prophylactic indomethacin versus placebo: the review by Fowlie 2010 showed that there was no evidence of a di erence between prophylactic indomethacin and control for gastrointestinal perforation (RR 1.13, 95% CI 0.71 to 1.79; 1 RCT, 1202 infants).

Prophylactic ibuprofen
Prophylactic ibuprofen versus placebo: the review by Ohlsson 2020a showed that there was no evidence of a di erence between prophylactic ibuprofen and placebo or no intervention for gastrointestinal perforation (RR 4.88, 95% CI 0.87 to 27.36; 2 RCTs, 167 infants).

Gastrointestinal bleeding
One Cochrane Review reported on gastrointestinal bleeding. It included the following intervention.

Oliguria
Three Cochrane Reviews reported on the outcome of oliguria. They included the following interventions.

Prophylactic acetaminophen
Prophylactic acetaminophen versus placebo: the review by Jasani 2022 showed that there was no evidence of a di erence between prophylactic acetaminophen and placebo or no intervention on oliguria (RR 0.78, 95% CI 0.29 to 2.11; 1 RCT, 48 infants).

Serum/plasma levels of creatinine a er treatment
One Cochrane Review reported on this outcome. It included the following intervention.

Prophylactic ibuprofen
Prophylactic ibuprofen versus placebo: the review by Ohlsson 2020a showed that there was no evidence of a di erence between prophylactic ibuprofen and placebo or no intervention on serum creatinine levels post-treatment (weighted mean di erence (WMD)** 0.09 mg/dL, 95% CI 0.05 to 0.13; 6 RCTs, 800 infants; lowcertainty evidence).
** Please note that Cochrane now uses mean di erence.

Increase in serum/plasma levels of creatinine a er treatment
Two Cochrane Reviews reported on this outcome. They included the following interventions.

Prophylactic indomethacin
Prophylactic indomethacin versus placebo. The review by Fowlie 2010 showed that there was no evidence of a di erence between prophylactic indomethacin and control on serum/plasma levels of creatinine a er treatment (RR 1.09, 95% CI 0.47 to 1.79; 4 RCTs, 618 infants).

Prophylactic ibuprofen
Prophylactic ibuprofen versus placebo: the review by Ohlsson 2020a showed that compared to placebo or no intervention, prophylactic ibuprofen increased serum/plasma levels of Cochrane Database of Systematic Reviews creatinine a er treatment (RR 3.70, 95% CI 1.05 to 12.98; 2 RCTs, 285 infants).

Serum/plasma levels of bilirubin a er treatment
One Cochrane Review reported on this outcome. It included the following intervention.

Prophylactic acetaminophen
Prophylactic acetaminophen versus placebo: the review by Jasani 2022 showed that there was no evidence of a di erence between prophylactic acetaminophen and placebo or no intervention on serum/plasma levels of bilirubin a er treatment (MD 1 µmol/L, 95% CI -10.35 to 12.35; 1 RCT, 48 infants).

Interventions (pharmacological or surgical) for management of asymptomatic PDA in preterm infants
Only one Cochrane Review, which compared indomethacin for asymptomatic PDA versus placebo addressed this objective (Cooke 2003). It included the following outcomes (Table 7).

Symptomatic PDA
Compared to placebo, treatment of asymptomatic PDA with indomethacin reduced the development of symptomatic PDA posttreatment RR 0.36, 95% CI 0.19 to 0.68; (3 RCTs, 97 infants).

Proportion of infants requiring invasive PDA closure (surgical ligation or transcatheter occlusion)
There was no evidence of a di erence between the treatment of asymptomatic PDA with indomethacin and placebo on the need for invasive PDA closure (RR 0.45, 95% CI 0.17 to 1.21; 2 RCTs, 73 infants).

Chronic lung disease (CLD)
There was no evidence of a di erence between treatment of asymptomatic PDA with indomethacin and placebo for CLD (RR 0.91, 95% CI 0.62 to 1.35; 2 RCTs, 45 infants).

Retinopathy of prematurity (ROP)
There was no evidence of a di erence between treatment of asymptomatic PDA with indomethacin and placebo for any stage of ROP (RR 0.68, 95% CI 0.26 to 1.78; 2 RCTs, 55 infants).

Duration of hospitalisation (days)
There was no evidence of a di erence between treatment of asymptomatic PDA with indomethacin and placebo on the duration of hospitalisation (MD 11 days, 95% CI -45.21 to 23.21; 1 RCT, 26 infants).

Mortality
There was no evidence of a di erence between treatment of asymptomatic PDA with indomethacin and placebo for mortality (RR 1.32, 95% CI 0.45 to 3.86; 2 RCTs, 73 infants).

Necrotising enterocolitis (NEC)
There was no evidence of a di erence between treatment of asymptomatic PDA with indomethacin and placebo on NEC (RR 0.41, 95% CI 0.05 to 3.68; 1 RCT, 47 infants).

Failure of PDA closure a er completion of allocated treatment
Eight Cochrane Reviews reported on the outcome of failure of PDA closure. They included the following interventions (Table 8).

Indomethacin
Indomethacin versus placebo or no treatment: the review by Evans 2021 showed that compared to placebo or no treatment, indomethacin reduced failure of PDA closure post-treatment (RR 0.30, 95% CI 0.23 to 0.38; 10 RCTs, 654 infants; high-certainty evidence).
Prolonged versus short course of indomethacin: the review by Herrera 2007 showed that there was no evidence of a di erence between a prolonged and short course of indomethacin on failure of PDA closure post-treatment (RR 0.82, 95% CI 0.51 to 1.33; 4 RCTs, 361 infants).

Cochrane Database of Systematic Reviews
Echocardiogram-guided versus standard IV ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between echocardiogram-guided and standard IV ibuprofen on failure of PDA closure post-treatment (RR 1.31, 95% CI 0.44 to 3.91; 1 RCT, 49 infants).
Continuous infusion of ibuprofen versus intermittent boluses of ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between continuous infusion of ibuprofen and intermittent boluses of ibuprofen on failure of PDA closure post-treatment (RR 1.18, 95% CI 0.88 to 1.5; 1 RCT, 111 infants).
Rectal versus oral ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between rectal and oral ibuprofen on failure of PDA closure post-treatment (RR 0.83, 95% CI 0.28 to 2.4; 1 RCT, 72 infants).

Acetaminophen
Acetaminophen versus ibuprofen: the review by Jasani 2022 showed that there was no evidence of a di erence between acetaminophen and ibuprofen on failure of PDA closure posttreatment (RR 1.02, 95% CI 0.88 to 1.18; 18 RCTs, 1535 infants; moderate-certainty evidence).
Acetaminophen versus indomethacin: the review by Jasani 2022 showed that there was no evidence of a di erence between acetaminophen and indomethacin on failure of PDA closure posttreatment (RR 1.02, 95% CI 0.78 to 1.33; 4 RCTs, 380 infants; lowcertainty evidence).
Late acetaminophen versus placebo: the review by Jasani 2022 showed that there was no evidence of a di erence between late acetaminophen and placebo on failure of PDA closure posttreatment (RR 0.85, 95% CI 0.72 to 1.01; 1 RCT, 55 infants; lowcertainty evidence).
Acetaminophen and ibuprofen combination therapy versus ibuprofen alone: the review by Jasani 2022 showed that there was no evidence of a di erence between acetaminophen and ibuprofen combination therapy and ibuprofen alone on failure of PDA closure post-treatment (RR 0.77, 95% CI 0.43 to 1.36; 2 RCTs, 111 infants; low-certainty evidence).

Surgical ligation
Surgical PDA ligation versus medical treatment with indomethacin: the review by Malviya 2013 showed that compared to medical therapy, surgical ligation reduced failure of PDA closure posttreatment (RR 0.04, 95% CI 0.01 to 0.27; 1 RCT, 154 infants).

Death or moderate/severe neurodevelopmental disability
No reviews reported on the combined outcome of death or moderate/severe neurodevelopmental disability.

Proportion of infants receiving open-label medical treatment
Four Cochrane Reviews reported on the outcome of receipt of openlabel treatment. They included the following interventions (Table  9).

Indomethacin
Indomethacin versus

Proportion of infants requiring invasive PDA closure (surgical ligation or transcatheter occlusion)
Five reviews reported on the outcome of invasive PDA closure. They included the following interventions ( Oral ibuprofen versus indomethacin: the review by Ohlsson 2020b showed that there was no evidence of a di erence between oral ibuprofen and indomethacin on the need for invasive PDA closure (RR 0.93, 95% CI 0.50 to 1.74; 4 RCTs, 174 infants; lowcertainty evidence).
Oral ibuprofen versus IV ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between oral ibuprofen and IV ibuprofen on the need for invasive PDA closure (RR 0.41, 95% CI 0.14 to 1.21; 5 RCTs, 406 infants; moderatecertainty evidence).
High-dose ibuprofen versus standard-dose ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between high-dose ibuprofen and standard-dose ibuprofen on the need for invasive PDA closure (RR 1.00, 95% CI 0.15 to 6.71; 1 RCT, 70 infants).
Early treatment (initiated within seven days) versus expectant management: the review by Mitra 2020a showed that there was no evidence of a di erence between early treatment with ibuprofen and expectant management on the need for invasive PDA closure (RR 1.14, 95% CI 0.66 to 1.96; 3 RCTs, 305 infants).
Very early treatment (initiated within three days) versus expectant management: the review by Mitra 2020a showed that there was no evidence of a di erence between very early treatment with ibuprofen and expectant management on the need for invasive PDA closure (RR 1.00, 95% CI 0.36 to 2.75; 1 RCT, 60 infants).
Continuous infusion versus intermittent bolus of ibuprofen: the review by Ohlsson 2020b showed that compared to intermittent bolus of ibuprofen, continuous infusion of ibuprofen reduced the need for invasive PDA closure (RR 0.28, 95% CI 0.08 to 0.94; 1 RCT, 111 infants).
Rectal versus oral ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between rectal ibuprofen and oral ibuprofen on the need for invasive PDA closure (RR 1.00, 95% CI 0.15 to 6.72; 1 RCT, 72 infants).

Acetaminophen
Acetaminophen versus ibuprofen: the review by Jasani 2022 showed that there was no evidence of a di erence between acetaminophen and ibuprofen on the need for invasive PDA closure (RR 0.61, 95% CI 0.34 to 1.08; 6 RCTs, 603 infants).
Acetaminophen versus indomethacin: the review by Jasani 2022 showed that there was no evidence of a di erence between acetaminophen and indomethacin on the need for invasive PDA closure (RR 1.31, 95% CI 0.72 to 2.38; 2 RCTs, 237 infants).
Late acetaminophen (initiated on day 14 or later) versus placebo: the review by Jasani 2022 showed that there was no evidence of a di erence between late acetaminophen and placebo on the need for invasive PDA closure (RR 3.11, 95% CI 0.13 to 73.11; 1 RCT, 55 infants).

Proportion of infants receiving open-label medical or surgical treatment in the placebo or no treatment group
No reviews reported on this outcome.

Chronic lung disease
Six reviews reported on the outcome of CLD (all definitions included). They include the following interventions ( Acetaminophen and ibuprofen combination therapy versus ibuprofen alone: the review by Jasani 2022 showed that there was no evidence of a di erence between acetaminophen and ibuprofen combination therapy and ibuprofen alone for CLD (RR 0.80, 95% CI 0.28 to 2.27; 1 RCT, 24 infants).

Surgical ligation
Surgical PDA ligation versus medical treatment with indomethacin: the review by Malviya 2013 showed that there was no evidence of a di erence between surgical PDA ligation and medical therapy for CLD (RR 1.28, 95% CI 0.83 to 1.98; 1 RCT, 154 infants).

Pulmonary haemorrhage
Four Cochane Reviews reported on the outcome of pulmonary haemorrhage. They included the following interventions (Table 12).

Indomethacin
Indomethacin versus placebo or no treatment: the review by Evans 2021 showed that there was no evidence of a di erence between indomethacin and placebo or no treatment for pulmonary haemorrhage (RR 0.40, 95% CI 0.14 to 1.16; 1 RCT, 92 infants).

Severe intraventricular haemorrhage (grades III/IV)
Five Cochrane Reviews reported on the outcome of severe IVH. They include the following interventions (Table 13).

Indomethacin
Indomethacin versus placebo or no treatment: the review by Evans 2021 showed that there was no evidence of a di erence between indomethacin and placebo or no treatment for severe IVH (RR 0.33, 95% CI 0.01 to 7.45; 1 RCT, 24 infants). Acetaminophen and ibuprofen combination therapy versus ibuprofen alone: the review by Jasani 2022 showed that there was no evidence of a di erence between acetaminophen and ibuprofen combination therapy and ibuprofen alone for severe IVH (RR 1.50, 95% CI 0.30 to 7.43; 1 RCT, 24 infants).

Retinopathy of prematurity (ROP)
Six Cochrane Reviews reported on the outcome of ROP. They included the following interventions (Table 14).

Duration of hospitalisation (days)
Five Cochrane Reviews reported on duration of hospitalisation. They included the following interventions (Table 15).

Indomethacin
Indomethacin versus placebo or no treatment: the review by Evans 2021 showed that there was no evidence of a di erence between indomethacin and placebo or no treatment on the duration of hospitalisation (

All-cause mortality
Seven Cochrane Reviews reported on the outcome of mortality. They included the following interventions (Table 17).

Indomethacin
Indomethacin versus placebo or no treatment: the review by Evans 2021 showed that there was no evidence of a di erence between indomethacin and placebo or no treatment for all-cause mortality

Necrotising enterocolitis (NEC)
Seven Cochrane Reviews reported on the outcome of NEC. They included the following interventions ( Acetaminophen and ibuprofen combination therapy versus ibuprofen alone: the review by Jasani 2022 showed that there was no evidence of a di erence between acetaminophen and ibuprofen combination therapy and ibuprofen alone for NEC (by radiological diagnosis; RR 0.33, 95% CI 0.01 to 7.45; 1 RCT, 24 infants; lowcertainty evidence).

Surgical ligation
Surgical PDA ligation versus medical treatment with indomethacin: the review by Malviya 2013 showed that there was no evidence of a di erence between surgical PDA ligation and medical therapy for NEC (by radiological diagnosis; RR 0.95, 95% CI 0.29 to 3.15; 1 RCT, 154 infants).

Gastrointestinal bleeding
Three Cochrane Reviews reported on gastrointestinal bleeding. They included the following interventions (Table 19).

Indomethacin
Indomethacin versus placebo or no treatment: the review by Evans 2021 showed that there was no evidence of a di erence between indomethacin and placebo or no treatment for gastrointestinal bleeding (RR 0.33, 95% CI 0.01 to 7.58; 2 RCTs, 119 infants; lowcertainty evidence).

Gastrointestinal perforation
Four Cochrane Reviews reported on gastrointestinal perforation. They included the following interventions (

Oliguria
Five Cochrane Reviews reported on the outcome of oliguria. They included the following interventions (Table 21).

Indomethacin
Prolonged versus short course of indomethacin: the review by Herrera 2007 showed that compared to a short course of indomethacin, a prolonged course reduced oliguria (urine output < 1 mL/kg/hour; RR 0.27, 95% CI 0.13 to 0.60; 2 RCTs, 197 infants).
Very early treatment (initiated within three days) versus expectant management: the review by Mitra 2020a showed that there was no evidence of a di erence between very early treatment with indomethacin and expectant management for oliguria (urine output < 1 mL/kg/hour; RR 5.00, 95% CI 0.63 to 39.39; 1 RCT, 44 infants).

Ibuprofen
IV ibuprofen versus placebo or no treatment: the review by Ohlsson 2020b showed that compared to placebo or no treatment, IV ibuprofen increased oliguria (urine output < 1 mL/kg/hour; RR 39.00, 95% CI 2.40 to 633.01; 1 RCT, 134 infants).
Oral ibuprofen versus indomethacin: the review by Ohlsson 2020b showed that there was no evidence of a di erence between oral ibuprofen and indomethacin for oliguria (RD 0.00, 95% CI -0.10 to 0.10; 1 RCT, 36 infants).
Oral ibuprofen versus IV ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between oral ibuprofen and IV ibuprofen for oliguria (urine output < 1 mL/kg/ hour; RR 0.14, 95% CI 0.01 to 2.66; 4 RCTs, 304 infants; low-certainty evidence).
Echocardiogram-guided versus standard IV ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between echocardiogram-guided and standard IV ibuprofen for oliguria (urine output < 1 mL/kg/hour; RR 5.31, 95% CI 0.29 to 97.57; 1 RCT, 49 infants).
Continuous infusion versus intermittent bolus of ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between continuous infusion and intermittent bolus of ibuprofen for oliguria (urine output < 1 mL/kg/hour; RR 0.51, 95% CI 0.05 to 5.45; 1 RCT, 111 infants).
Acetaminophen and ibuprofen combination therapy versus ibuprofen alone: the review by Jasani 2022 showed that there was no evidence of a di erence between acetaminophen and ibuprofen combination therapy and ibuprofen alone for oliguria (RR 0.50, 95% CI 0.05 to 4.81; 1 RCT, 24 infants).

Adjunct therapies
Dopamine versus control: the review by Barrington 2002 showed that there was no evidence of a di erence between the combination of dopamine and indomethacin versus indomethacin alone for oliguria (RR 0.73, 95% CI 0.35 to 1.54; 1 RCT, 33 infants).

Serum/plasma levels of creatinine a er treatment
Two Cochrane Reviews reported on this outcome. They included the following interventions (Table 22).

Ibuprofen
IV ibuprofen versus placebo or no treatment: the review by Ohlsson 2020b showed that compared to placebo or no treatment, IV ibuprofen increased serum creatinine post-treatment (MD 29.17 µmol/L, 95% CI 12.60 to 45.741 RCT, 134 infants).
Oral ibuprofen versus indomethacin: the review by Ohlsson 2020b showed that there was no evidence of a di erence between oral ibuprofen and indomethacin for serum creatinine posttreatment (MD -0.51 µmol/L, 95% CI -6.04 to 5.01; 5 RCTs, 190 infants; very low-certainty evidence).

Interventions for patent ductus arteriosus (PDA) in preterm infants: an overview of Cochrane Systematic Reviews (Review)
High-dose ibuprofen versus standard-dose ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between high-dose ibuprofen and standard-dose ibuprofen for serum creatinine post-treatment (MD 8.84 µmol/L, 95% CI -4.41 to 22.09; 1 RCT, 60 infants).
Echocardiogram-guided versus standard IV ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between echocardiogram-guided and standard IV ibuprofen for serum creatinine post-treatment (MD -11.49 µmol/L, 95% CI -29.88 to 6.90; 1 RCT, 49 infants).
Continuous infusion versus intermittent bolus of ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between continuous infusion and intermittent bolus of ibuprofen for serum creatinine post-treatment (MD 2.10 µmol/L, 95% CI -4.92 to 9.12; 1 RCT, 111 infants).

Adjunct therapies
Dopamine versus control: the review by Barrington 2002 showed that there was no evidence of a di erence between the combination of dopamine and indomethacin versus indomethacin alone for serum creatinine post-treatment (MD 2.04 µmol/L, 95% CI -17.90 to 21.97; 2 RCTs, 59 infants).

Increase in serum/plasma levels of creatinine a er treatment
Three Cochrane Reviews reported on this outcome. They included the following interventions (Table 23).

Adjunct therapies
Furosemide versus control: the review by Brion 2001 showed that there was no evidence of a di erence between the combination of furosemide and indomethacin versus indomethacin alone for increase in serum creatinine post-treatment (MD -0.88 µmol/L, 95% CI -12.38 to 10.61; 3 RCTs, 70 infants).

Serum/plasma levels of bilirubin a er treatment
Two Cochrane Reviews reported on this outcome. They included the following interventions (Table 24).
Rectal ibuprofen versus oral ibuprofen: the review by Ohlsson 2020b showed that there was no evidence of a di erence between rectal ibuprofen and oral ibuprofen for serum bilirubin levels posttreatment (MD 7.01 µmol/L, 95% CI -11.23 to 25.25; 1 RCT, 72 infants).

Increase in serum/plasma levels of bilirubin a er treatment
No review reported on this outcome.

Subgroup Analyses
None of the reviews provided data on any of our pre-specified subgroups.

Summary of main results
We included 16 Cochrane Reviews (138 randomised controlled trials (RCTs), 11,856 preterm infants) on the management of patent ductus arteriosis (PDA) in preterm infants. The number of trials included in each review ranged from none to 39. Six reviews (N = 4976) reported on prophylactic interventions for the prevention of PDA, and included pharmacological prophylaxis with prostaglandin inhibitor drugs (indomethacin, ibuprofen, and acetaminophen) and prophylactic surgical PDA ligation and nonpharmacologic interventions (chest shielding during phototherapy and restriction of fluid intake). One review (N = 97) reported on the use of indomethacin for the management of asymptomatic PDA. Nine reviews (N = 6783) reported on interventions for the management of symptomatic PDA, and included pharmacotherapy with prostaglandin inhibitor drugs (indomethacin, ibuprofen and acetaminophen) in various routes and dosages; surgical PDA ligation; and adjunct therapies (use of furosemide and dopamine in conjunction with indomethacin). The certainty of the evidence, when reported by the respective reviews for the primary outcomes for prevention of PDA, ranged from moderate to low, while those for the primary outcomes for treatment of PDA, ranged from high to low.

Interventions for prevention of PDA and related complications in preterm infants
Prophylactic indomethacin probably reduces severe intraventricular haemorrhage (IVH), while it does not appear to a ect the composite outcome of death or moderate/severe neurodevelopmental disability. Prophylactic ibuprofen probably marginally reduces severe IVH (moderate-certainty evidence), while the evidence is very uncertain on the e ect of prophylactic acetaminophen on severe IVH. There is no evidence on the e ect of either prophylactic ibuprofen or acetaminophen on the composite outcome of death or moderate/severe neurodevelopmental disability. There is a paucity of evidence for any other prophylactic intervention on the primary outcomes of severe IVH and the composite of death or moderate/severe neurodevelopmental disability.
With respect to other patient-important outcomes, both prophylactic indomethacin and ibuprofen (moderate-certainty evidence) reduced the need for invasive PDA closure. Necrotising enterocolitis (NEC) appeared to be lower with both prophylactic surgical ligation and fluid restriction. There was no e ect of the other prophylactic interventions on any other clinically relevant outcomes, such as mortality or chronic lung disease (CLD).

Interventions for management of asymptomatic PDA in preterm infants
Overall evidence is limited (3 RCTs, 97 infants) for the management of asymptomatic PDA. Treatment of asymptomatic PDA with indomethacin appears to reduce the development of symptomatic PDA post-treatment. There is no evidence on the e ect of asymptomatic PDA treatment on the composite outcome of death or moderate/severe neurodevelopmental disability.

Interventions for management of symptomatic PDA in preterm infants
All available prostaglandin inhibitor drugs appear to be more e ective in PDA closure when compared to placebo or no treatment (high-certainty evidence for indomethacin; moderatecertainty evidence for ibuprofen; low-certainty evidence for early administration of acetaminophen). Oral ibuprofen appears to be more e ective in PDA closure compared to ibuprofen (moderate-certainty evidence); and high-dose ibuprofen appears to be more e ective in PDA closure compared to standard-dose ibuprofen (moderate-certainty evidence). There was no evidence of any di erence in PDA closure e ectiveness between the three available prostaglandin inhibitor drugs (low-to moderate-certainty evidence). There is no evidence on the e ect of treatment of symptomatic PDA on the composite outcome of death or moderate/ severe neurodevelopmental disability.
From a safety perspective, compared to indomethacin administration, NEC appears to be lower with ibuprofen (any route; moderate-certainty evidence), exclusive oral administration of ibuprofen (low-certainty evidence), and with acetaminophen (low-certainty evidence). On the contrary, NEC appears to be more common with a prolonged course of indomethacin versus a shorter course. Oliguria is also higher with use of indomethacin versus ibuprofen (moderate-certainty evidence), the use of ibuprofen versus either placebo or acetaminophen, and with early pharmacological treatment of PDA, initiated within the first seven days of life versus later treatment.

Overall completeness and applicability of evidence
We found reviews for all our prespecified objectives. However, there was substantial variation in the certainty of available evidence for the di erent interventions for patient-important outcomes. For prophylactic interventions, the precision of the estimate of e ects is best with indomethacin, while the evidence is limited for ibuprofen and sparse for acetaminophen. Evidence from RCTs does suggest a definite benefit with prophylactic indomethacin, and a probable benefit with prophylactic ibuprofen with a reduction in severe IVH. However, the results should be interpreted with caution, as several of the RCTs contributing to the reviews on prophylactic indomethacin and ibuprofen were conducted more than 20 years ago, when NICU practices were vastly di erent, including the use of antenatal corticosteroids, approaches to mechanical ventilation, and the use of surfactant. It is unclear whether the treatment e ects shown in these trials still apply today to extremely preterm infants at higher risk of severe IVH. From a safety perspective, neither indomethacin nor ibuprofen prophylaxis was shown to increase patient-important adverse outcomes, such as NEC or gastrointestinal perforation. However, the trials included in the respective reviews did not consider the e ect of co-administration of other drugs that might cause harm. This might be an important consideration for clinicians, especially with the emergence of newer prophylactic therapies, such as prophylactic hydrocortisone. A recent individual patient data (IPD) meta-analysis of RCTs showed that prophylactic lowdose hydrocortisone can improve survival without CLD (adjusted odds ratio (OR) 1.48, 95% CI 1.02 to 2.16), however, the concomitant use of prophylactic indomethacin and hydrocortisone increases the risk of gastrointestinal perforation (OR 2.50; 95% CI, 1.33 to 4.69; Sha er 2019). The largest trial contributing to the said IPD meta-analysis, the PREMILOC trial (N = 1072), failed to demonstrate similar harm in the subgroup of infants who were co-administered hydrocortisone and ibuprofen (47% of infants enroled in the intervention arm of the trial received ibuprofen (Baud 2016)). Therefore, clinicians should weigh the current applicability of existing evidence for benefit against the potential for harm with concomitant use of other medications, while considering the use of prophylactic non-steroidal anti-inflammatory drugs (NSAIDs) in preterm infants. Similarly, clinicians should exercise caution while considering non-pharmacologic interventions, such as prophylactic fluid restriction to prevent a symptomatic PDA, given the trials were conducted between 1980 and 2000 in moderately preterm infants, and therefore, may not be applicable to extremely preterm infants in the current context. Further, clinicians should refrain from extrapolating this evidence to using fluid restriction as a therapeutic option for treatment of symptomatic PDA, given there is no evidence to support the latter.
With respect to treatment of asymptomatic or symptomatic PDA, the availability of RCT evidence is substantially variable, depending on the intervention used. Overall, RCT evidence consistently demonstrates that the use of prostaglandin inhibitor drugs is e ective in closing a PDA. Despite e ective PDA closure, current evidence fails to demonstrate a benefit of prostaglandin inhibitor drugs for patient-important clinical outcomes, such as need for invasive PDA closure, CLD, or mortality. However, several study limitations prevent us from drawing firm conclusions on the lack of e icacy of the prostaglandin inhibitor drugs for clinical outcomes. First, there was wide variation in PDA definitions in the included trials, especially the trials for treatment of Cochrane Library Trusted evidence. Informed decisions. Better health.
Cochrane Database of Systematic Reviews symptomatic PDA. Symptomatic PDA was defined in most trials based on characteristic clinical signs, along with echocardiographic evidence of an increased PDA shunt volume. However, the trials did not have consistent eligibility criteria, from either a clinical or an echocardiographic standpoint. Further, the most used echocardiographic criteria, the PDA size, and the le atrium to aortic root ratio, have been shown to have poor inter-rater reliability, and therefore, may represent suboptimal inclusion criteria (de Freitas Martins 2018;Zonnenberg 2012). In addition, the trials did not attempt to di erentiate between PDAs with moderate versus high shunt volume, based on any clinical or echocardiographic criteria. These drawbacks of existing RCTs may have led us to include a highly heterogeneous population in the meta-analyses, especially, more mature infants with smaller PDA shunt volumes, in whom spontaneous PDA closure was highly likely to occur. Second, as evident from their wide confidence intervals, the e ect estimates for the most important clinical outcomes were imprecise, which failed to provide convincing evidence for an absence of e ect on such outcomes. Third, a substantial proportion of infants in the placebo or no treatment group ended up receiving open-label medical therapy, thereby, likely pulling the e ect estimate towards the null. The latter, especially, might be an important reason why e ective PDA closure did not necessarily translate into improved longer-term clinical benefit.
The need for subsequent open-label therapy, including definitive surgical PDA closure, also highlights the fact that medical therapy, though better than placebo, is by no means a highly e ective option for PDA closure. Therefore, most RCTs of medical treatment were essentially trials of drug therapy, rather than the elimination of the PDA shunt. Therefore, despite growing calls for accepting the null hypothesis and abandoning further clinical trials on PDA management, the current evidence underscores the need to clearly establish which PDA shunts, if any, are associated with worse clinical outcomes, and pursue further clinical trials that include only those infants at the highest risk of PDA-attributable morbidities, and explore highly e ective and safe shunt elimination strategies for such clinically important PDA shunts.

Quality of the evidence
The quality of reviews as assessed by the AMSTAR 2 criteria was variable. We only judged two reviews to be of high quality, while five were of low quality, and two of critically low quality ( Table  2). Reviews that we judged as critically low quality failed to use a satisfactory technique for assessing the risk of bias in individual studies, in addition to omissions in other critical domains of the AMSTAR 2 criteria. Of note, none of the included reviews provided a rationale for including only randomised controlled trials in their review. This may be associated with Cochrane Neonatal's approach of traditionally including only RCTs in reviews of interventions to obtain the most unbiased estimates of treatment e ects. However, in the absence of well-done RCTs, other study designs, such as large observational studies, may be an important source of evidence, especially related to the safety of the interventions. Further, the majority of the reviews did not explicitly include information on funding sources for the trials. This did have an impact on the quality of the reviews as per the AMSTAR 2 criteria, as full disclosure of any funding is important to ensure that no financial incentive introduced bias (Lundh 2017).
Only five of the Cochrane Reviews assessed the overall certainty of the evidence using GRADE methodology (Evans 2021; Jasani 2022; Mitra 2020a; Ohlsson 2020a; Ohlsson 2020b). We did not reassess the certainty of evidence, but summarised the certainty assessed by the respective review authors. With regard to the primary outcomes defined in this overview, the certainty of the evidence was not reported for all available interventions. For PDA prevention, the certainty of the evidence, which was available only for prophylactic ibuprofen for the primary outcome of severe IVH, was deemed to be moderate. The certainty of the evidence was not assessed for interventions for the management of asymptomatic PDA. While for symptomatic PDA treatment, the certainty of the evidence for the primary outcome of failure of PDA closure was available for all available prostaglandin inhibitor drugs. The overall certainty for symptomatic PDA closure was high for indomethacin, moderate for ibuprofen, and moderate-low for acetaminophen. The most common reason for downgrading the certainty of the evidence was serious risk of bias, followed by imprecision in e ect estimates.

Potential biases in the overview process
We are confident that this overview is a comprehensive summary of all currently available Cochrane Reviews on the management of the PDA in preterm infants. We did not apply any date restrictions to the search. Five of the 16 reviews were either first published or updated in the past two years, making this a comprehensive summary of the best available evidence. One potential source of bias is that two of the overview authors are first authors or co-authors on three of the included reviews. However, quality assessment of the reviews, using the AMSTAR 2 criteria, was carried out in duplicate to minimise any intellectual bias (Table 2).

Agreements and disagreements with other studies or reviews
With respect to prophylactic therapies, the results of this overview largely align with the recently published Cochrane network metaanalysis by Mitra 2022. Both the overview and the network metaanalysis showed that prophylactic indomethacin reduces the risk of severe IVH and the need for surgical PDA closure, increases the risk of oliguria, and likely does not increase the risk of NEC or gastrointestinal perforation. In addition, both the overview and the network meta-analysis demonstrated that prophylactic ibuprofen also reduces the need for surgical PDA closure, and likely does not increase the risk of NEC or gastrointestinal perforation. However, the network meta-analysis by Mitra 2022 failed to demonstrate a di erence for severe IVH and oliguria with prophylactic ibuprofen, unlike the Ohlsson 2020a review, which showed a marginal reduction in severe IVH, in addition to a definite increase in oliguria. These observed di erences in results may be related to corresponding di erences in the datasets analysed in the two reviews, as the search for the Ohlsson 2020a review was updated in October 2018, while the search for the Mitra 2022 review was updated in December 2021. However, given the considerable overlap of studies included in the network meta-analysis by Mitra 2022 and the Ohlsson 2020a review, the more likely rationale for the observed di erences in results could be related to di erences in analytical methods. While the Ohlsson 2020a review used the traditional Cochrane Neonatal approach of using fixed-e ect metaanalysis, thereby, generally obtaining more precise estimate of e ects, the network meta-analysis used a Bayesian random-e ects model, which was likely to produce more conservative estimates, especially in the absence of a substantial contribution from the indirect comparisons, thereby, failing to establish di erences for the said outcomes. With respect to prophylactic acetaminophen, Library Trusted evidence. Informed decisions. Better health.
Cochrane Database of Systematic Reviews both reviews failed to draw meaningful conclusions due to overall paucity of evidence. With regard to treatment of symptomatic PDA, the results of this overview align with two previous network metaanalyses, which both demonstrated that prostaglandin inhibitor drugs were e ective in closing a PDA, which however, failed to translate into a clinically meaningful benefit (Jones 2011;Mitra 2018).
Overall, our findings generally support the current recommendations from the Canadian Pediatric Society (CPS (Mitra 2022a)), and the American Academy of Pediatrics (AAP (Hamrick 2020)), that include: considering prophylactic indomethacin to prevent severe IVH in high risk extremely preterm infants, and refraining from pharmacotherapy for PDA closure in clinically stable preterm infants, given the lack of clear evidence for benefit, while judiciously weighing the benefits and harms of PDA treatment in clinically unstable, extremely preterm infants, given the overall lack of RCT evidence in this vulnerable population. However, it is important to note that both the CPS and AAP statements suggest considering invasive PDA closure (surgical ligation or percutaneous transcatheter closure) if the PDA remains persistently symptomatic, despite limited RCT evidence on the benefit of invasive PDA closure on clinically relevant outcomes.

Implications for practice
Prophylactic indomethacin probably reduces severe intraventricular haemorrhage (IVH), while it does not appear to a ect the composite outcome of death or moderate/severe neurodevelopmental disability. Prophylactic ibuprofen probably marginally reduces severe IVH (moderate-certainty evidence), while the evidence is very uncertain on the e ect of prophylactic acetaminophen on severe IVH.
All available prostaglandin inhibitor drugs appear to be e ective in closing a symptomatic patent ductus arteriosus (PDA) compared to no treatment (high-certainty evidence for indomethacin; moderate-certainty evidence for ibuprofen; low-certainty evidence for early administration of acetaminophen). Oral ibuprofen appears to be more e ective in PDA closure than intravenous ibuprofen (moderate-certainty evidence). High-dose ibuprofen appears to be more e ective in PDA closure than standard-dose ibuprofen (moderate-certainty evidence). There is no evidence of a di erence in PDA closure e ectiveness between the three available prostaglandin inhibitor drugs (low-to moderate-certainty evidence). There is limited evidence on the e ect of invasive PDA closure on clinical outcomes.

Implications for research
From a PDA prophylaxis perspective, any future clinical trial should only include extremely preterm infants at the highest risk of mortality and major morbidity. Given the low rate of adverse clinical outcomes in older preterm infants, lack of clear benefit, and potential for harm with routine use, there is no clinical equipoise for further clinical trials on prophylactic interventions for PDA in older preterm infants, especially those born a er 28 weeks of gestation.
Regarding PDA treatment, future research should focus on defining the infant population, including the PDA characteristics that would benefit most from the elimination of the PDA shunt. Future clinical trials should exclusively enrol this high-risk population to explore and describe the safest and most e ective shunt elimination strategy that leads to meaningful improvement in infant and family important clinical outcomes.

A D D I T I O N A L T A B L E S Review ID and title
Date of search and date assessed as up-to-date

Anabrees 2011
Fluid restriction and prophylactic indomethacin vs prophylactic indomethacin alone for prevention of morbidity and mortality in extremely low birth weight infants 1966 to December 2010

None
RCTs and quasi-RCTs Infants, < 1000 g at birth, who received prophylactic indomethacin in the first 24 hours of life.
Fluid restriction (to achieve at least 10% weight loss in the first week of life) plus indomethacin prophylaxis (starting within the first 24 hours for 3 doses) versus indomethacin prophylaxis alone                                  Cochrane Database of Systematic Reviews 4 RCTs, N = 400 Table 24. Interventions for symptomatic PDA: serum/plasma levels of bilirubin a er treatment (Continued) CI confidence interval;IV: intravenous; PO: per os; PDA: patent ductus arteriosus; PMA: post-menstrual age; PR: per rectum; RCT: randomised controlled trials; RD: risk di erence; ROP: retinopathy of prematurity; vs: versus Reference is the listed comparison therapy, unless otherwise indicated by * Di erences between intervention and comparison provided as mean di erence (MD), unless otherwise specified

H I S T O R Y
Protocol first published: Issue 4, 2020

C O N T R I B U T I O N S O F A U T H O R S
SM conceived the project. SM, WdB, DW, and PSS dra ed the overview, reviewed all dra s, and approved the final version of the overview.

D E C L A R A T I O N S O F I N T E R E S T
SM is an Associate Editor, Cochrane Neonatal Group. However, he had no involvement in the editorial processing of this overview. He has also published medical articles related to the management of PDA in preterm infants.
WdB has published medical articles related to PDA in preterm infants. He was the project leader of the BeNeDuctus trial (Hundscheid 2023), a international, multicentre, randomised non-inferiority trial of early treatment versus expectant management of patent ductus arteriosus in preterm infants (study protocol: Hundscheid 2018, and statistical analysis plan: Hundscheid 2021).
DW has no conflict of interest to declare.
PSS is an Associate Editor, Cochrane Neonatal Group. However, he had no involvement in the editorial processing of this overview.

Internal sources
• No sources of support provided

External sources
• Vermont Oxford Network, USA Cochrane Neonatal Reviews are produced with support from Vermont Oxford Network, a worldwide collaboration of health professionals dedicated to providing evidence-based care of the highest quality for newborn infants and their families.

D I F F E R E N C E S B E T W E E N P R O T O C O L A N D R E V I E W
We made the following changes to the protocol (Mitra 2020b).
1. In this overview, we only included reviews that specifically reported on interventions primarily intended to prevent or treat a PDA, and not all interventions that reported PDA as an outcome. We clarified this under the Types of Interventions section, by adding the following sentence: "In this overview, we specifically included reviews of therapies primarily intended to prevent or treat a PDA".