03 · What You Need to Know
The real question is not how much local evidence exists, but what information is missing
Start with what the international evidence actually establishes
“Strong international evidence” should mean more than a large number of papers from other countries. The relevant considerations include the methodological credibility of the studies, consistency and precision of findings, directness to the question, and, where available, the quality of an appropriate evidence synthesis.
A large literature can still provide uncertain evidence if the studies share serious limitations. Conversely, a well-conducted body of research across several settings may provide considerable information even when none of the studies comes from your locality.
Evidence-informed policymaking guidance has long treated the best available evidence from around the world as an appropriate starting point for judgments about the effects of policies and programmes. Restricting attention to a geographically local subset can itself be misleading when the broader evidence is relevant.
Do not ask whether international evidence “replaces” local evidence
Global and local evidence frequently answer different questions.
International studies may tell you whether an intervention has produced an effect across several settings, whether a relationship appears repeatedly, which mechanisms have been proposed, or which implementation approaches have been tried. Local evidence may tell you how common the problem is in your setting, what resources are available, how institutions operate, what an intervention would cost, whether the target population differs meaningfully, or what stakeholders consider acceptable.
Evidence that may transfer
Evidence about effects, mechanisms, relationships, or approaches may sometimes remain informative across settings, especially when findings are consistent under varied conditions.
Evidence about the local setting
Information about local prevalence, baseline conditions, resources, costs, institutional capacity, implementation barriers, values, and other contextual conditions generally cannot be supplied simply by accumulating more studies from elsewhere.
Consistency across different settings can strengthen the case for applicability
If similar findings have been observed across countries, populations, institutions, and time periods that differ in meaningful ways, there may be less reason to expect the result to disappear solely because it is applied in another setting.
Guidance on assessing the applicability of systematic-review findings specifically asks whether studies were conducted in the same or similar settings and whether findings were consistent across settings and periods. Consistency under varied conditions can support a judgment that findings are broadly applicable.
That judgment still requires care. The absence of observed variation does not prove that no important contextual modifier exists, particularly when the available studies represent only a narrow range of settings.
Look for effect modifiers rather than demanding a local replication automatically
The most useful question is often: What could plausibly make the finding different here?
Potential effect modifiers depend on the topic. They might include age or population composition, baseline risk, socioeconomic conditions, culture, institutional arrangements, infrastructure, implementation fidelity, professional expertise, technology access, regulatory environments, environmental exposure, or the precise form of an intervention.
If the international evidence already includes substantial variation across the factors most likely to matter and the findings remain reasonably consistent, the case for transferability becomes stronger. If your local setting differs on a factor known or plausibly expected to modify the result, local evidence becomes more important.
Absolute effects may change even when relative effects transfer
This distinction is particularly important for interventions and risk-related questions.
An intervention can have a similar relative effect across settings while producing different absolute benefits because the baseline level of risk or need differs. Guidance on applicability explicitly identifies baseline conditions as a factor that can change expected absolute effects even when relative effectiveness is similar.
This means that international evidence about effectiveness may remain informative while local baseline data are still needed to estimate what that effectiveness could mean in practice.
Implementation can be more context-sensitive than efficacy
An intervention that has worked elsewhere may depend on staffing, infrastructure, supply chains, institutional authority, training, financing, or user acceptance that cannot be assumed locally.
The underlying intervention may be effective, yet the local system may be unable to deliver it in the same way. Conversely, adaptations needed for local implementation may change the intervention enough that the original evidence becomes less directly applicable.
World Health Organization guidance on contextualizing evidence-based recommendations similarly emphasizes fitting evidence to available resources and organizational contexts rather than simply transferring recommendations unchanged.
Some local questions cannot be answered by international research at all
Suppose your question is, “What proportion of teachers in this province currently use generative AI?” A hundred international studies cannot provide that prevalence estimate.
They can help you choose definitions, instruments, predictors, or analytic approaches. They may show how AI use has been measured elsewhere. But the value of the local parameter remains unknown until appropriate local data exist.
The same applies to many questions about local costs, service availability, institutional capacity, stakeholder preferences, environmental conditions, and current implementation.
Watch Out
More international evidence does not gradually turn into local data. If the information you need is inherently specific to the setting, the limitation is not solved by finding another study from somewhere else.
Little local research does not mean no local evidence
Local evidence is broader than locally published journal studies. Depending on the question, useful local information may come from routine administrative data, surveys, government reports, institutional records, programme monitoring, qualitative research, stakeholder consultation, costs, implementation data, or other sources.
This distinction matters when a conventional literature search finds few local papers. Before concluding that local evidence is absent, consider whether the information needed for an applicability judgment exists in other forms.
A careful search designed around the local evidence environment may uncover material that a journal database does not.
Weak local studies should not automatically override stronger international evidence
A small local study may appear more directly relevant because it was conducted nearby. That does not make it methodologically stronger.
Guidance on using local conditions cautions against relying on local evaluations alone to estimate impacts when they are small or methodologically limited. Judgments are better informed in the context of all relevant evidence.
If a weak local study conflicts with a strong international evidence base, investigate the discrepancy. The local result might reveal a genuine contextual modifier, but it might also reflect bias, imprecision, measurement differences, or chance.
Where the local evidence is incompletely reported, preserve that uncertainty rather than giving it extra evidential weight simply because it is local.
07 · A Quick Checklist
Before relying heavily on international evidence in a local setting
Before applying the broader evidence, check:
Assess the strength of the international evidence rather than equating a large publication count with strong evidence.
Examine the populations, settings, interventions, exposures, and time periods represented in the broader evidence.
Check whether findings are reasonably consistent across meaningfully different settings.
Identify local factors that could plausibly modify the finding or alter its implementation.
Determine whether local baseline conditions could change the absolute magnitude of the expected outcome.
Identify which unanswered questions require genuinely local data rather than another international study.
Search beyond journal articles for local administrative, implementation, cost, resource, or stakeholder evidence where relevant.
State clearly when the application of international evidence remains uncertain because direct local evidence is unavailable.