|Network meta-analysis of indomethacin versus ibuprofen versus placebo for PDA in preterm infants
|Jones LJ, Craven PD, Attia J, Thakkinstian A, Wright I
This review found intravenous indomethacin or ibuprofen promoted ductal closure in pre-term infants with linked pulmonary artery and aorta, but other short-term benefits were not seen; ibuprofen may increase chronic lung disease rates. Some caution is needed in interpretation of the results given the small amount of data available, widespread use of indomethacin by control groups, and diverse study populations.
To compare the benefits and harms of indomethacin, ibuprofen and placebo in pre-term infants with patent ductus arteriosus (an open channel between the pulmonary artery and the aorta).
MEDLINE, EMBASE, CINAHL, The Cochrane Library, clinicaltrials.gov and controlled-trials.com were searched from inception to August 2008. Search terms were reported. Abstracts from meetings of professional bodies, including the American Pediatric Society, Pediatric Academic Societies and the European Society for Paediatric Research, and Effective Care of the Newborn Infant were handsearched. No search restrictions were applied.
Randomised controlled trials (RCTs) and quasi-randomised trials that compared intravenous indomethacin and intravenous ibuprofen versus each other or placebo in pre-term (under 37 weeks) or low-birth weight (under 2,500gms) babies with clinically important patent ductus arteriosus beyond 24 hours postnatal age were eligible for inclusion. Patent ductus arteriosus could be diagnosed using echocardiographic and/or clinical criteria.
The primary review outcome was patent ductus arteriosus closure. Secondary outcomes were death before hospital discharge, necrotising enterocolitis in the neonatal period, intraventricular haemorrhage, chronic lung disease (at a range of time points) and neurodevelopment (measured with the Bayley Scales of Infant Development).
Infants in the included trials varied widely in birth weight (from under 1,000gms to 2,800gms), gestational age, and age at treatment (from one to 14 days). Patent ductus arteriosus was identified by echocardiogram in most trials. In most trial groups, infants received non-randomised indomethacin as rescue treatment: this was the case in seven of nine placebo groups in trials of indomethacin versus placebo (range 0 to 85%, where stated).
The authors did not state how many reviewers performed the selection.
Assessment of study quality
Trials were assessed for quality and could score up to a maximum of 5 points for criteria of randomisation, allocation concealment, blinding, and losses to follow-up.
Two reviewers independently assessed trial quality; disagreements were resolved with a third reviewer.
Risk ratios (RRs) and 95% confidence intervals (CIs) were extracted or calculated for each trial outcome. Primary trial authors were contacted for more information if required.
Two reviewers extracted the data; disagreements were resolved with a third reviewer.
Methods of synthesis
Trials were combined using inverse variance methods to calculate pooled risk ratios and 95% confidence intervals, using fixed-effect and random-effects models. Heterogeneity was assessed using Χ2 and I2. As only one RCT directly compared ibuprofen with placebo, indirect pooled risk ratios for this comparison were calculated from trials with one common treatment arm (and similar participant characteristics). This was achieved by fitting natural logarithms derived from direct meta-analyses into a meta-regression model (network meta-analysis). Numbers needed to treat were calculated.
Sensitivity analyses were conducted to examine the impact of trial quality score (under 3 versus 3 and over), blinding, age at treatment (under 72 hours versus 72 hours and over) and diagnostic method for patent ductus arteriosus. Where significant heterogeneity was detected, post-hoc analyses were conducted to investigate the impact of gestational age, mean birth weight and number of drug doses.
Publication bias was assessed with funnel plots and the Egger test.
Results of the review
Nineteen RCTs were included in the review (range 16 to 410 infants). Four trials scored 5 for quality, two RCTs scored 4, one RCT scored 3, eight RCTs scored 2, and four RCTs scored 1.
There was no significant difference between indomethacin and ibuprofen in rates of patent ductus arteriosus closure (direct comparison, 10 RCTs, n=615 infants; I2=0%). Both drugs were twice as effective as placebo (indomethacin RR 2.41, 95% CI 1.72 to 3.36, direct comparison, nine RCTs, n=653, I2=61%; ibuprofen RR 2.40, 95% CI 2.03 to 2.84, one direct comparison RCT, n=28 plus indirect comparison, I2=0%). Sensitivity and subgroup analyses did not alter findings, except that indomethacin and placebo did not differ in effectiveness in trials of infants with mean gestational age of under 28 weeks (data not presented).
Ibuprofen was associated with significantly higher rates of chronic lung disease at any reported age than indomethacin (RR 1.28, 95% CI 1.03 to 1.60, direct comparison, six RCTs; I2=0%). The rate did not differ significantly between Ibuprofen and placebo (RR 1.34, 95% CI 0.99 to 1.82, one direct comparison RCT plus indirect comparison). There was no significant difference between indomethacin and placebo in rates of chronic lung disease (six RCTs; I2=25%). There were no significant findings in any other analyses of secondary outcomes.
That authors calculated that use of intravenous indomethacin or ibuprofen in 100 preterm infants would close an additional 33 patent ductus arteriosus and might prevent one case of necrotising enterocolitis, compared with placebo. Use of ibuprofen rather than placebo might cause an additional 14 cases of chronic lung disease, three extra cases of intravenous haemorrhage, and one death. Use of indomethacin rather than placebo might cause an additional two cases of chronic lung disease, two extra cases of intravenous haemorrhage and one death.
Egger test results showed possible publication bias in the analysis comparing death rates associated with indomethacin and ibuprofen (p=0.04). No other significant publication bias was detected.
Intravenous indomethacin or ibuprofen promoted patent ductus arteriosus closure in pre-term infants, but were not associated with other short-term benefits. Ibuprofen may increase the risk of chronic lung disease.
The objectives and inclusion criteria of the review were clear. Relevant sources were searched for studies, with no restrictions by language or publication status. Publication bias was assessed using appropriate methods. Steps were taken to minimise the risk of reviewer bias and error by more than one reviewer independently undertaking quality assessment and data extraction, but it was unclear whether this applied to study selection.
Appropriate statistical techniques were used to combine the data and explore statistical heterogeneity. As the authors noted, the review findings were difficult to interpret in view of widespread use of indomethacin in the placebo groups. They also noted that the review was limited by the small size and number of available trials, clinical heterogeneity between trials and possible confounding by the use of indirect comparisons. The authors' conclusions were limited to pre-term infants and did not refer to those with low-birth weight. Supplementary data were not available online.
The review was well conducted, but some caution may be needed in interpretation of results due to the small amount of data available, widespread use of indomethacin in the control groups, and diversity of trial populations.
Implications of the review for practice and research
Practice: The authors did not state any implications for practice.
Research: The authors stated that future placebo-controlled RCTs assessing the use of intravenous ibuprofen or indomethacin for closing patent ductus arteriosus in pre-term babies should report complete data on lower birth weight subgroups, length of exposure to patent ductus arteriosus and duration of follow-up. Longer term respiratory and neurological outcomes should also be reported. The authors calculated that a sample size of at least 130 infants per group would be required to detect a difference in chronic lung disease at 36 weeks’ corrected age, although smaller samples would be feasible if treatment in the placebo arm could be avoided.
University of Newcastle Research Scholarship Central.
Jones LJ, Craven PD, Attia J, Thakkinstian A, Wright I. Network meta-analysis of indomethacin versus ibuprofen versus placebo for PDA in preterm infants. Archives of Disease in Childhood. Fetal and Neonatal Edition 2011; 96(1): F45-F52
Subject indexing assigned by NLM
Cardiovascular Agents /therapeutic use; Cyclooxygenase Inhibitors /therapeutic use; Ductus Arteriosus, Patent /drug therapy; Female; Humans; Ibuprofen /therapeutic use; Indomethacin /therapeutic use; Infant, Newborn; Infant, Premature; Infant, Premature, Diseases /drug therapy; Male; Randomized Controlled Trials as Topic; Treatment Outcome
Date bibliographic record published
Date abstract record published
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.