Types of studies

We will include in the review all randomised controlled trials (RCTs) comparing enteral feeding supplementation with prebiotics versus distilled water or no supplementation.

Types of participants

We will include all studies that enrolled neonates and will categorise neonates into three groups: term neonates (gestational age ≥ 37 weeks); late preterm neonates (35 to < 37 weeks' gestation); and premature neonates (< 35 weeks' gestation). We will exclude infants with the following conditions: asphyxia, major congenital anomalies, Rhesus isoimmunisation, G6PD deficiency, inborn errors, proven sepsis or other infection, cephalohaematoma, and subgaleal haemorrhage. We will also exclude infants with enteral feeding supplementation with specific formulas, probiotics, or zinc.

Types of interventions

Enteral feeding supplementation given with prebiotics during the first ten days of life is the intervention under study. Prebiotics can be provided in the form of galacto‐oligosaccharides (GOSs), fructo‐oligosaccharides (FOSs), acidic oligosaccharides (AOSs), lactulose, or inulin. We will include studies that initiated supplementation with prebiotics between days 2 and 10 of life at a dosage of 0.5 to 1.5 gr/kg/d, with concentrations of 1% and 0.8 gr/dL for oligosaccharides, lactulose, and inulin. To eliminate confounding effects of prebiotics in human breast milk and in formula, we will include studies that added these compounds to neonatal enteral milk. We will include studies of three types: trials that add prebiotics to breast milk in breast‐fed only infants; trials that add prebiotics to formula in bottle‐fed only infants; and trials providing mixed forms (i.e. adding prebiotics to both breast milk and formula). Investigators must compare interventions versus enteral supplementation of milk with distilled water, no supplementation, or both. We will include in this review only trials that are continued for at least seven days. For management of neonatal jaundice, all infants in both groups will be treated according to the protocol for phototherapy described in guidelines of the American Academy of Pediatrics 2004.

Types of outcome measures

Primary outcomes

• Neonatal hyperbilirubinaemia: incidence of hyperbilirubinaemia at any time during the first ten days of life, considered as the primary outcome. Hyperbilirubinemia is defined:

• • for term and late preterm neonates (gestational age ≥ 35 weeks): as total bilirubin (TB) level eligible for phototherapy, as described in guidelines of the American Academy of Pediatrics 2004 (Figure 1), or as absolute TB level ≥ 15 mg/dL; and

  • for premature neonates (< 35 weeks’ gestation): as TB level that is eligible for phototherapy, as described in guidelines of the American Academy of Pediatrics 2004 (Figure 2), or as absolute TB level > 1% of body weight.

Figure 1

---

---

Guidelines for phototherapy in infants at ≥ 35 weeks' gestation. 

(American Academy of Pediatrics Subcommittee on Hyperbilirubinemia. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004;114:297.)

Figure 2

---

---

Suggested guidelines for initiating phototherapy or exchange transfusion in premature infants.

(Maisels MJ, Watchko JF, Bhutani VK, et al. An approach to the management of hyperbilirubinemia in the preterm infant less than 35 weeks of gestation. Journal of Perinatology 2012;32:660‐664.)

• Acute bilirubin encephalopathy: defined as a clinical syndrome, in the presence of severe hyperbilirubinaemia, that ranges from the initial phase of lethargy, hypotonia, reduced movement, and poor sucking and suckling to the advanced phase of deep stupor, hypertonia, inability to feed, and shrill cry or seizures (Canadian Pediatric Society 2007).

Electronic searches

We will use criteria and standard methods of the Cochrane Collaboration and the Cochrane Neonatal Review Group and will undertake a comprehensive search including:

• the Cochrane Central Register of Controlled Trials (CENTRAL) in the Cochrane Library (April 2015) (Appendix 1);

• MEDLINE (January 1996 to current) (Appendix 2);

• Embase (January 1980 to current) (Appendix 3); and

• the Cumulative Index to Nursing and Allied Health Literature (CINAHL; 1982 to current).

We will check Proceedings of the Prenatal Society of Australia and New Zealand (PSANZ ‐ 2005 to current) and of the Pediatric Academic Societies (PAS) ‐ American Pediatric Society, Society for Pediatric Research, and European Society for Pediatric Research (2000 to current).

We will combine search terms with the highly sensitive search strategy presented in Boluyt 2008 to identify RCTs on neonates in the light of sensitive key terms in our search strategy (Appendix 4). The proposed search strategy will use a combination of controlled vocabulary; we will use subject terms and free text terms for MEDLINE and will adapt these for use in the other databases mentioned.

Searching other resources

We will apply no language and publication restrictions; therefore, we will search reference lists of identified clinical trials and review articles and will contact authors of published articles to ask about possible unpublished trials. We will search all ongoing and recently completed clinical trials available through www.ClinicalTrials.gov and www.Controlled‐trials.com.

Selection of studies

Two review authors will independently check the titles and abstracts of all studies retrieved through the literature search to identify those that meet the inclusion criteria, will independently confirm their eligibility by reviewing the full text of retrieved articles, and will resolve disagreements through discussion. We will consult the Cochrane Neonatal Review Group for advice when necessary.

Data extraction and management

We will design a data extraction form and will evaluate its validity and reliability before use. At least two review authors will independently extract data and will resolve differences in data interpretation with assistance provided by trial authors. We will resolve uncertainties regarding trial information and incomplete or missing data by contacting trial authors. We will enter data into Review Manager (RevMan 5.3) and will check them for accuracy.

Assessment of risk of bias in included studies

Two review authors will independently assess the risk of bias (low, high, or unclear) of all included trials using the Cochrane ‘Risk of bias’ tool (Higgins 2011) for the following domains.

• Selection bias.

• Performance bias.

• Attrition bias.

• Reporting bias.

• Any other bias.

We will resolve disagreements by discussion or by consultation with a third assessor. See Appendix 5 for a detailed description of risk of bias for each domain.

Measures of treatment effect

We will present pooled estimates of dichotomous outcomes as risk ratios (RRs) or risk differences (RDs) with corresponding 95% confidence intervals (CIs). For continuous outcomes, we will use mean differences (MDs) along with 95% CIs to compare intervention and control groups. When outcomes are measured by different methods or by different scales, we will use standardised mean differences (SMDs). If data are missing or incomplete and the RD statistic is statistically significant, we will present the number needed to treat for an additional beneficial or harmful outcome (NNTB/NNTH).

Unit of analysis issues

We will consider the participating infant as the unit of analysis in trials of cross‐over design and in parallel clinical trials, and the participating neonate unit as the unit of analysis in cluster‐randomised trials. We may consider appropriate corrections on reported estimates from included cluster‐randomised trials. We will deal with an approximate analysis as suggested by the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We will multiply the standard error of the reported effect estimate by the square root of the design effect and will calculate the design effect from the average cluster size and the intracluster correlation coefficient (ICC). If such a correction is not possible, we will include the proposed cluster trial in the review but not in the meta‐analysis.

Dealing with missing data

If clarification of information or missing data is needed, we will contact the relevant study authors. We will exclude from the meta‐analysis studies with a high drop‐out rate for which we cannot obtain appropriate details of missing data and will perform a sensitivity analysis to evaluate the impact of exclusion on results of analysis.

Assessment of heterogeneity

We will use the I² statistic and the following categories to assess for heterogeneity among studies.

• Less than 25%: no heterogeneity.

• 25% to 49%: low heterogeneity.

• 50% to 74%: moderate heterogeneity.

• 75% or greater: high heterogeneity.

We will use Cochran's Chi² Q test and the associated P value in evaluating heterogeneity. We will consider P < 0.1 as an indication of the presence of heterogeneity. We will present a Galbraith plot for heterogeneity and will perform visual assessment to identify obvious overlaps and outliers. In cases of moderate or high heterogeneity, we will explore potential sources of heterogeneity by performing sensitivity analysis and subgroup analyses.

Assessment of reporting biases

We plan to use funnel plots to assess possible reporting or publication bias. For funnel plots to be meaningful, ten or more trials must be included. We will conduct Begg and Egger linear tests in evaluating publication bias.

Data synthesis

We will apply standard methods of the Cochrane Neonatal Review Group and RevMan V.5.3 for meta‐analyses. We will use a fixed‐effect method in combining the effects of studies that are sufficiently similar. When heterogeneity between studies is evident, we will perform random‐effects subgroup analysis or meta‐regression, if appropriate. For estimates of combined risk ratio and risk difference, we will use the Mantel‐Haenszel method. We will use mean differences (MDs) or standardised mean difference (SMDs) to combine estimates of quantitative data as reported in clinical trials measuring the same outcome using the same scale and in those using different scales or measurement methods, respectively. We will report risk ratios (RRs) and MDs and SMDs along with 95% CIs and will base analysis on numbers needed to treat for additional beneficial and harmful outcomes.

Subgroup analysis and investigation of heterogeneity

We will run a subgroup analysis based on the following.

• Gestational age (term neonates (gestational age ≥ 37 weeks), late preterm neonates (35 to 37 weeks' gestation), premature neonates (< 35 weeks' gestation)).

• Birth weight (≥ 2500 grams, < 2500 grams, < 1500 grams).

• Type of prebiotic (oligosaccharide, lactulose, inulin).

• Type of feeding (only breast milk‐fed infants, only formula‐fed infants, infants given a mixed form of feeding).

• Duration of supplementation.

Sensitivity analysis

We will perform sensitivity analyses based on both outcomes and baseline characteristics if the chance of bias in study findings is high, to assess the impact of bias on results of the meta‐analysis.