Protocol
The method for this systematic review was developed in accordance with the recommendations of the Selected Reporting Items for Systematic Review and the 2015 Meta-Analysis Protocol Statement (PRISMA-P) [25].
Eligibility criteria
We determine the eligibility criteria according to PECO statement as follows:
Population : the population of interest will include pregnant women and their children.
Exposure : - exposure to Hg before or during gestation,
- exposure to environmental Hg exposures measures through biological samples of the mothers during pregnancy
- Exposures measured through environmental monitoring or occupational exposures, such as surveys)
Comparators: mother who were exposed to low level Hg before or during pregnancy and their children
Outcomes : Any adverse child outcome such as premature birth, low birth weight, failure of hearing screening or audiology test, adverse score on intellectual ability or adaptive skills tests, diagnoses (Diagnostic and Statistical Manual [DSM] or ICD codes) of intellectual disability, autism spectrum disorder, epilepsy, cerebral palsy, attention deficit hyperactivity disorder (ADHD), (inattention, inattentiveness, sustained-attention, disruptive-behavior, habituation, hyperactivity, overactivity, hyperactive, overactive, impulse, impulsiveness, impulse-control, impulsivity, delay-of-reinforcement, delayed-reinforcement, motor-control, motor-activity, disinhibition, inhibition) mental illness, immune or autoimmune conditions and measured levels of cytokines, and evidence of structural or cell damage in live-born children.
Study Design
Studies will be limited on design and observational studies, cross-sectional studies, cohort studies, case control studies, and case series. Non-research letters and editorials, abstracts, seminar reviews, systematic review, non-human studies, and clinical trial will be excluded. Randomized controlled trials will be excluded due to unsuitability the type of questions to be answered. The searches will be focused to peer-reviewed full-text articles in English. However, unpublish information sources, such as thesis or dissertation will be included.
Information Sources
We will search the publication systematically from the international journal databases, include Google Schoolar, PUBMED, Sciencedirect, Proquest, Web of Science, Springerlink, DOAJ (Directory of Open Access Journals) and EBSCO MEDLINE. Addition publication also will be found from references of the articles. We will limit the publication years from 2015 to 2021. Contact with the authors for further information will be made when needed.
Search Strategy
We will construct structured electronic search strategies using the thesaurus terms of each database (e.g., MeSH for MEDLINE and PUBMED) and using keywords relevant to the population, exposure, and the outcomes. The strategy will then be applied to all of the databases.
Selection Procedure
First, duplicate publications will be removed. Then, the screening will be applied to the title and abstract of the papers to assess the eligibility. We will review the full texts of publications to find the papers meet the inclusion criteria using a checklist based on PRISMA checklist. We will include study with score of minimum 3. Two reviewers will work independently then compare their results and discuss the findings in the presence of differences. A third reviewer will be involved when a consensus is not feasible. Studies will be selected for retrieval following the procedure illustrated in Figure 1.
Data extraction and management
We will extract data on (1) general characteristics of the research include investigator names, study and publication year, objectives, search strategy, number of included studies, number of participants and funding source, (2) characteristics of the research population of that may have age, gestation age, location and ethnicity, (3) the exposure, which can include assessment method, distribution in the study population, the exposure pathway and Hg level, (4) outcomes, include outcome measurement methods, various type of outcomes, (5) comparators (6) methods applied which may consist of statistical analysis and factors adjusted for, and (7) results such as incidence and prevalence, the mean or median level of Hg exposure, measures of association, or stratified analyses, (8) review limitation.
For reference screening, we will use Covidence (https://www.covidence.org). We will apply SRDR (Systematic Review Data Repository; https://srdr.ahrq.gov/home/index) for data extraction and management. The data or results will be presented in summary tables and figures.
Any disputes occurring during the process of data extraction will be resolved by dialogue and consensus between the two reviewers and, if necessary, will include a third reviewer. We will try to contact the review authors if any data is missing or lacking. When it is difficult to obtain any data, then the study will be reported as an "included review" without data.
Quality and strength of Study Evidence
We will evaluate the evidence quality and strength of the included studies by measuring the risk of bias, rank the quality of the studies, and rank the power of the evidence. Risk of bias will be assessed by means of intrument of The Navigation Guide Systematic Review Methodology [27], [28]. According to the methodology, the risk of bias potentially assigned from, inter alia, recruitment strategy, blinding, exposure assessment, confounding, incomplete outcome data, selective reporting, funding source and conflict of interest. Refer to the instrument, We will classify the risk of bias into "low risk," "probably low risk," "probably high risk," "high risk," or "not applicable". Evidence quality will be assessed with some indicators start from moderate which will be downgrade or upgrade to high and low quality in conformity with some factors. The quality of evidence will be assessed with taking into consideration some factors, include the direction of effect estimates, confidence in effect estimates, other compelling attributes of the data that may influence certainty. It will classify into four groups, include sufficient evidence, limited evidence, inadequate evidence, and evidence of lack of toxicity [27].
Data synthesis
If possible, we will perform a meta-analysis. Meta-analysis of outcome variables will be carried out when the study are sufficiently meet the inclusion criteria and uniform (or clarity) reporting of outcome estimates. Otherwise, we will undertake a statistical analysis based on types of variables, as recommended by Grimshaw et al. [38], i.e., dichotomous or continuous data of Hg exposure and neurodevelopmental disorder outcomes. We will report risk ratios and their corresponding 95 % confidence intervals and p values regarding the dichotomous outcomes, and for continuous outcomes, mean differences will be reported [39]. We will use the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram [40] to show the overall process of studies selection.