The standard methods of the Cochrane Neonatal Review Group were employed in creating this update.
Selection of studies
Retrieved articles were independently assessed for eligibility by two review authors. Discrepancies were resolved by discussion and consensus.
Data extraction and management
Data were abstracted independently by two review authors. Discrepancies were resolved by discussion and consensus. Where data were incomplete, the primary investigator was contacted for further information and clarification.
Assessment of risk of bias in included studies
Standard methods of The Cochrane Collaboration and the Neonatal Review Group were used to assess the methodological quality (to meet the validity criteria) of the trials. For each trial, information was sought regarding the method of randomization, and the blinding and reporting of all outcomes of all the infants enrolled in the trial. Each criterion was assessed as yes, no, can't tell. Two review authors separately assessed each study. Any disagreement was resolved by discussion. This information was added to the table Characteristics of included studies. In addition, for the updates in 2010 and 2013, the following issues were evaluated and entered into the risk of bias table.
1) Sequence generation (checking for possible selection bias). Was the allocation sequence adequately generated?
For each included study, we categorized the method used to generate the allocation sequence as:
(2) Allocation concealment (checking for possible selection bias). Was allocation adequately concealed?
For each included study, we categorized the method used to conceal the allocation sequence as:
adequate (e.g. telephone or central randomization; consecutively numbered sealed opaque envelopes);
inadequate (open random allocation; unsealed or non-opaque envelopes, alternation; date of birth);
(3) Blinding (checking for possible performance bias). Was knowledge of the allocated intervention adequately prevented during the study? At study entry? At the time of outcome assessment?
For each included study, we categorized the methods used to blind study participants and personnel from knowledge of which intervention a participant received. Blinding was assessed separately for different outcomes or classes of outcomes. We categorized the methods as:
adequate, inadequate or unclear for participants;
adequate, inadequate or unclear for personnel;
adequate, inadequate or unclear for outcome assessors.
(4) Incomplete outcome data (checking for possible attrition bias through withdrawals, dropouts, protocol deviations). Were incomplete outcome data adequately addressed?
For each included study and for each outcome, we described the completeness of data including attrition and exclusions from the analysis. We noted whether attrition and exclusions were reported, the numbers included in the analysis at each stage (compared with the total randomized participants), reasons for attrition or exclusion where reported, and whether missing data were balanced across groups or were related to outcomes. Where sufficient information was reported or supplied by the trial authors, we re-included missing data in the analyses. We categorized the methods as:
(5) Selective reporting bias. Are reports of the study free of suggestion of selective outcome reporting?
For each included study, we described how we investigated the possibility of selective outcome reporting bias and what we found. We assessed the methods as:
adequate (where it is clear that all of the study’s pre-specified outcomes and all expected outcomes of interest to the review have been reported);
inadequate (where not all the study’s pre-specified outcomes have been reported; one or more reported primary outcomes were not pre-specified outcomes of interest are reported incompletely and so cannot be used; study fails to include results of a key outcome that would have been expected to have been reported);
(6) Other sources of bias. Was the study apparently free of other problems that could put it at a high risk of bias?
For each included study, we described any important concerns we had about other possible sources of bias (for example, whether there was a potential source of bias related to the specific study design or whether the trial was stopped early due to some data-dependent process). We assessed whether each study was free of other problems that could put it at risk of bias as:
If needed, we planned to explore the impact of the level of bias through undertaking sensitivity analyses.
Measures of treatment effect
For dichotomous outcomes, relative risk (RR), risk difference (RD), and the number needed to treat to benefit (NNTB) and the associated confidence intervals (CIs) were calculated. For continuous outcomes, treatment effect was expressed as mean difference (MD) and its calculated standard deviation (SD). When median, range, and sample size were reported, the mean and SD were estimated using established methods (Hozo 2005).
Assessment of heterogeneity
Heterogeneity was defined as a significant test of heterogeneity (P < 0.1) and differences in the treatment effects across studies. Tests for between-study heterogeneity (including the I2 statistic) were applied. If noticed, possible sources of heterogeneity were examined, including differences in the type or dose of probiotics used, the population under study (VLBW versus ELBW infants), and the quality of the study.
Review Manager 5.2 software was used for statistical analysis. For estimates of typical RR and RD we used the Mantel-Haenszel method. For measured quantities we used the inverse variance method. All meta-analyses were done using the fixed-effect model.
Subgroup analysis and investigation of heterogeneity
The secondary objective was to conduct a subgroup analysis to investigate the effect of the probiotics in and for the following.
Different species of probiotics.
Different times of initiation of probiotics.
Different durations of probiotics administration.
A sensitivity analysis was carried out to assess the effect of trials methodological quality on the results of the meta-analysis. Studies were considered to be of high quality if allocation was concealed and adequately described.