Short birth interval prevalence, determinants and effects on maternal and child health outcomes in Asia-Pacific region: a systematic review and meta-analysis protocol

Protocol design

This study will be guided by the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols guidelines proposed by Moher et al.35 As this study will consist of five separate papers arising from an overarching search strategy, five separate protocols focusing on definition, measurement, prevalence and predictors; child mortality; child undernutrition; adverse perinatal outcomes; adverse maternal outcomes were registered with International Prospective Register of Systematic Reviews (PROSPERO).

Eligibility criteria

Eligible studies during the title and abstract screening will be guided based on the ‘population, concept and contexts (PCC)’ framework, proposed by Peters et al36 (table 1). This framework was applied due to the multiple outcomes contained in this search strategy. The detailed search terms are provided in online supplemental material I.

Supplemental material

Table 1

Inclusion criteria using the PCC framework

Information sources

A comprehensive literature search will be conducted using the following electronic databases: MEDLINE, Scopus, Cumulative Index to Nursing and Allied Health Literature (CINAHL), Maternity and Infant Care, and Web of Science (WoS). The reference lists of published review articles will be scrutinised for additional relevant publications. Peer-reviewed literature published from September 2000 (since the inception of the Millennium Development Goals) up to May 2023 will be retrieved.

Search strategy

The authors, in collaboration with an experienced librarian, developed a Boolean string from key search terms (online supplemental material I). Included will be terms related to SBI including its prevalence, neonatal health, infant health, under-five health, maternal health and countries in the Asia-Pacific region. Comprehensive search strategies will be used to identify the relevant studies and will be customised when reporting individual outcomes. Eligible studies will also be selected from the reference lists of retrieved articles.

Since the Asia-Pacific region varies in areas depending on the context, this study context will include countries from South Asia and East Asia and the Pacific region as described by the World Bank in 2022.37 38 The search strategy will be piloted to ensure the appropriateness of the proposed keywords and databases.

Study records

Data management

All eligible articles will be uploaded into Endnote X9 reference management software,39 to screen duplicates. Articles will then be exported to Covidence, which allows multiple reviewers to screen studies simultaneously. In Covidence, the first author will develop the eligibility criteria to determine inclusion/exclusion.

Selection process

Three authors will independently perform both title and abstract screening (DMS, CC and MH) and full-text review (TAH, CC and MH) using Covidence.40 Any disagreement will be resolved through a discussion between the three authors at both stages. Reference lists of potentially relevant publications will be manually searched. When the full text of potentially relevant publications cannot be located, an attempt will be made to get a copy either through the University library or by directly contacting the author(s) via email.

Data collection process

A spreadsheet of key factors will be developed to extract relevant data from each included study. Three authors (DMS, CC and MH) will test and refine the data extraction tool using 10 eligible studies before its use. Two members of the research team will then independently perform the data extraction for each of the specified outcomes (ie, definition, measurement, prevalence and predictors; child mortality; child undernutrition; adverse perinatal outcomes; adverse maternal outcomes). Any disagreements in data extraction will be resolved by consensus, and if necessary, through discussion with a third reviewer.

If any selected article has insufficient information, its corresponding author will be contacted by email or phone to receive the missing data. If the attempt to get the missing data will not be possible, the data will be deleted and will be commented on in the Discussion section.

Data items

The extracted data will include key variables such as (1) name of the first author and publication year, (2) country, (3) study setting, (4) study design, (5) participants’ characteristics (sample, inclusion criteria), (6) data collection methods, (7) key findings on SBI including its definition, measurement, reported prevalence, its association with adverse maternal and child health outcomes, and the confounders controlled in each study.

Outcomes and prioritisation and synthesis

Findings of the systematic review will be presented using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.41 A PRISMA flow diagram will be used to demonstrate the literature study selection process and search results. A descriptive numerical summary will be used to present the characteristics of the included studies. Information on the definition or classification, measurement, prevalence and determinants of SBI will be summarised using narrative synthesis. Determinants of SBI will include several factors such as sociodemographic and economic factors, reproductive factors and child related factors. Subjective to the scope of the literature that will be retrieved, the effects of SBI on adverse maternal and child health outcomes will be separately summarised and individually presented for the following outcomes: adverse maternal outcomes (ie, maternal anaemia, gestational diabetes, hypertensive disorder of pregnancy and depression), adverse perinatal outcomes (ie, LBW, preterm birth, small for gestational age, stillbirth, perinatal mortality), neonatal, infant and under-five mortalities, and under-five undernutrition (ie, underweight, wasting, stunting and anaemia). First, a narrative synthesis will be used to summarise the findings. Under each category, we will further provide data on the article’s characteristics, including, but not limited, to the total number of studies, types of study design, source of data, year of publications and key findings. Meta-analysis will be conducted, subject to the heterogeneity of the studies included. Studies that use different definitions and measurements for SBI will be analysed separately to ensure comparability among studies.

The evaluation of the heterogeneity will be realised according to the Cochrane Handbook criteria through the I2 statistic.42 A value of 0% demonstrates a lack of heterogeneity in studies and ≥50% values indicate considerable heterogeneity. If the I2 value is less than 50%, the heterogeneity is considered as low, and a fixed-effect model will be used. If the I2 value is 50% or more, a random effects model will be used. Also, once moderate or higher heterogeneity will be identified, we will explore the source of heterogeneity using subgroup analysis and meta-regression. Based on the availability of data, study and participant characteristics such as study design, publication year, study setting and sample size, participants’ age will be considered in the subgroup and meta-regression. We will perform a sensitivity analysis based on level of risk of bias, by comparing random and fixed-effect model, and by excluding possible outlying studies, if the visual inspection of the forest plot shows poorly overlapping confidence interval (CI). Publication bias will be examined by visual inspection of Funnel Plot asymmetry and Egger’s regression test.43 The trim-and-fill method will be used when evidence of publication bias is found, and to estimate and adjust for potentially missing studies, the effect size will be recalculated accordingly.44 Forest plots will be constructed to show the study-specific relative risk (RR)/odds ratio (OR) estimates and pooled RR/OR estimates.

If a study is eligible for inclusion in the systematic review but does not provide adequate data for inclusion in the meta-analysis, other study characteristics and results will be summarised narratively to synthesise and tabulate the results.

Risk of bias in individual studies

Joanna Briggs Institute quality appraisal tools45–49 will be used to assess the methodological quality of research publications by evaluating the extent to which they addressed the possibility of bias in areas of study design, conduct and analysis (online supplemental material II). The magnitude of studies with a total score indicating poor quality will be discussed. Two members of the research team will independently assess each included paper and any uncertainty regarding the quality of publications will be resolved through discussion. Although a formal assessment of the methodological quality of the included studies will be performed; articles with poor quality will not be excluded.

Supplemental material

Confidence in cumulative evidence

In addition to risk of bias assessment for individual studies, the quality of cumulative evidence will also be rated using the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) framework.50 GRADE tool classifies the studies as very low, low, moderate and high quality. Two authors will independently perform this evaluation, and any disagreements will be decided through discussion (third author).

This post was originally published on