Table 2.

A comparison of systematic review and systematic evidence mapping methodology and their respective roles in risk management decision-making (adapted from James et al., 2016).

Step	Conduct of step in SRs related to assessing chemical health risks	Conduct of step in SEMs related to assessing chemical health risks	SR vs SEM for responding to risk management needs
Pre-published protocol	Define all methods in advance of conduct of review	Same	Provides transparency; reduces bias; opportunity for peer review and stakeholder engagement. Applies to both SRs and SEMs.
Statement of objectives	Question concerns the effect of an exposure on health; or the effect of intervening to reduce exposure in terms of health benefit. Usually targets a single or few exposures and outcomes.	Question concerns the state of the evidence base for a topic. Usually open-ended and encompassing a range of multiple related exposures and outcomes.	SR: Focused, closed questions of SRs best service specific RM decisions such as characterising specific health risks/TDIs. SEM: Open questions of SEMs best service scenarios in which evidence should be surveyed and scoped, such as problem identification and priority-setting.
Comprehensive search	Search terms highly resolved and specified for most key elements of the objective statement, returning a moderate volume of evidence.	Wide ranging search strings of lower specificity based on topic rather than defining all key elements of the objective in the search.	SR: Narrow searches efficiently identify evidence related to exposure-outcome pairs. Maximum feasible number of sources searched to ensure collation of all relevant evidence for synthesis. SEM: Broader, topic-based SEM search allows evidence supportive of multiple decision scenarios to be identified. Flexible number of sources searched, or sources searched in a step-wise manner as appropriate to broader research objectives.
Screening against eligibility criteria (study inclusion)	Inclusion criteria specified in detail for all key elements of the objective.	Inclusion criteria defined in terms of topic rather than key elements of the objective.	SR: As for search, specific inclusion criteria ensure SRs efficiently service a specific research question. SEM: Broad objectives ensure inclusion of evidence relating to multiple decision scenarios.
Data extraction using tested extraction sheets	Complete extraction of meta-data and study findings.	Extraction of meta-data; optional extraction of study findings and other study characteristics depending on SEM objectives.	SR: Data extraction determined by objectives. SEM: Data extraction more flexible and can respond to needs of risk management process to develop fit-for-purpose maps of varying degrees of comprehensiveness.
Coding of extracted data using controlled vocabularies	Coding facilitates grouping of included studies for synthesis/integration according to review objectives. Coding is closely related to review objectives and data extraction process, whereby narrow research question and PECO statement inherently define specific code applicable to raw extracted data.	Coding facilitates broad comparison of heterogeneous data across an evidence base. Broad map objectives necessitate extensive coding process, whereby specific code must be defined in a step distinct from the formulation of end-users’ specific research questions.	SR: Tight review objectives pre-specify applied code (e.g. considering ages 0–18 as ‘Child’ for reviews focusing on a population of ‘Children’). Narrower range, or greater specificity of controlled vocabulary terms applicable per item of extracted data. SEM: Code pre-specified where possible, but addition of new terms (which could not be accounted for a priori) considered flexible. Any one item of extracted data may be coded by multiple and variably resolved terms. Openly accessible ontologies may be used for coding to promote consistency and interoperability.
Critical appraisal of included studies	Assessment of internal validity (risk of bias) conducted for all included studies.	Study validity assessment is optional and to some extent restricted if outcome is not a defined aspect of the SEM; study characteristics relevant to risk of bias assessment can be extracted.	SR: Describe the internal validity of the evidence base, which is an essential step of characterising confidence in the evidence. SEM: Flexible, critical appraisal step can be omitted; study methods are mapped or methodological quality assessed to goals, can be part of stepwise approach where quality only assessed for studies addressing key outcomes etc.
Synthesis of included studies	Quantitative synthesis where possible to produce characterisation of hazard from exposure; qualitative synthesis where pooling studies is not possible.	Reports of systematic maps can provide narrative synthesis of characteristics of the evidence key to a given decision-making context.	SR: Synthesis supports a specific type of decision context. SEM: Primary output is a more context-agnostic database which can be used by risk managers to support multiple decisions in the RM workflow; or to aid in a stepwise approach.
Characterisation of confidence in the evidence	Assessment of confidence or certainty in the results of the synthesis, according to characteristics of the evidence base taken as a whole.	SEMs do not synthesise included studies. SEMs help identify regions of evidence with characteristics indicative of being worth further, detailed analysis in support of a prospective decision.	SR: Provide detailed conclusions on certainty of evidence in hazard characterisation or to support risk assessments. SEM: Support a range of decisions, particularly decisions to focus research and review, e.g. indicating clusters where evidence may be strong enough to warrant SR (e.g. have a reasonable likelihood of changing a TDI), fill in gaps to reduce uncertainty and for surveillance.
Drawing conclusions/key review outputs	SRs primarily provide a summary effect estimate and surrounding uncertainty based on strength of the evidence and review methods.	SEMs primarily provide a searchable database of the characteristics of the evidence base, making the knowledge base locked away in manuscripts accessible to decision-makers.	SR: provide a qualitative and/or quantitative summary effect estimate in answer to a narrow and specific decision-making question. SEM: identify evidence gluts for synthesis. When combined with an understanding of RM needs, transparent criteria for prioritization of gluts for synthesis and gaps for commissioning primary research can be presented.

SR = systematic review, SEM= systematic evidence map, RM =risk management, TDI= tolerable daily intake.