From the question to its use, and back again
The Network starts from the questions decision makers actually face, follows them through to a decision, then keeps the evidence current.
The route from question to decision
-
Frame the question
We work with ministries, Parliament and agencies to turn a policy problem into a question research can answer: who it is about, what is being tried, compared with what, and which outcomes matter.
- Population
- Pregnant women in rural districts
- Intervention
- Home visits by community health workers
- Comparison
- Facility-based care only
- Outcome
- Births in a health facility
Example question, framed as population, intervention, comparison and outcome
-
Check what is known
Before anything new is commissioned, members search and appraise existing evidence, including unpublished local material. Often the answer, or most of it, already exists.
Records found1,240After duplicates removed860Full texts read74Studies included18Illustrative search and screening, the first step of a systematic review. Hover or tab to a stage for detail.
-
Produce the evidence
Where gaps remain, members carry out rapid reviews, systematic and living reviews, and primary research, often in teams across institutions. Meta-analysis pools comparable studies into a single, more precise answer.
Illustrative forest plot. Each square is a study, sized by its weight; the diamond is the pooled result. Select a study to leave it out and watch the pooled result change.
-
Bring it back to the decision
Findings return to the people who asked, in forms they can use, through briefings, dialogues and convenings, with a plain statement of how certain the evidence is.
Evidence brief
Community health worker home visits probably increase births in health facilities
Moderate certaintyExample brief, with certainty rated using the GRADE approach
-
Keep it current
Reviews are updated as new studies appear, so decisions do not rest on stale evidence. Each update narrows the uncertainty, or flags when the answer has changed.
Illustrative living review: the same estimate, growing more certain with each update. Hover or tab to a row for its numbers.
New questions come out of every decision, and the cycle starts again.
Comparing many options at once
Decision makers rarely choose between just two options. Network meta-analysis compares three or more interventions in a single analysis, combining direct and indirect evidence across a network of studies.
Direct comparison. Two options compared head to head in the same study. Here, studies have compared A with B, and A with C.
Indirect comparison. No study has compared B with C, but both have been compared with A. The network can estimate B against C through that common comparator. This is valid only if the studies are similar enough in their populations, settings, outcomes and methods: the transitivity assumption.
Consistency. Where a study comparing B with C directly does exist, its result is checked against the indirect estimate. If the two agree, the network is consistent; if not, the difference must be explained before anyone relies on the result.
Methods members bring
The Network brings together people trained across evidence synthesis and primary research.
Rapid reviews
Streamlined reviews for questions that cannot wait, with their shortcuts stated openly.
Systematic reviews and meta-analysis
Finding, appraising and pooling every relevant study on a question.
Network meta-analysis
Comparing and ranking several interventions at once, using direct and indirect evidence.
Living reviews
Reviews kept up to date as new studies are published, supported by Sabi Core.
Evidence gap maps
Visual maps of where evidence exists in a field and where it is missing.
Certainty of evidence
Rating how far a finding can be trusted, using GRADE and, for networks, CINeMA.
Alongside this, the Network strengthens the research community itself
Connecting researchers
Finding collaborators and assembling strong teams for joint research and bids.
Raising research quality
Shared standards, peer support, research integrity guidance and help with grant management.
Building the next generation
Training, mentoring and communities of practice, with deliberate routes for early-career researchers into leading roles.
Specialist centres
Producers and users brought together around a theme, under the Network's oversight.