When Is Local Evidence Necessary Before Applying an International Finding?

When should you refuse to assume that an international finding applies locally?

Researchers frequently encounter evidence produced somewhere else. A systematic review may contain no studies from their country. A validated intervention may have been tested in wealthier institutions. A relationship may be well documented internationally but scarcely examined in the local population.

One response is to insist that every finding must be replicated locally before it can be used. Another is to assume that strong international evidence applies unless someone proves otherwise. Both positions are too simple.

The more useful question is whether there is a plausible and consequential local difference that the existing evidence does not resolve.

Local evidence becomes necessary when context could materially change the conclusion

The need for local evidence depends on what could change across settings

Do not use national borders as the test for applicability

A finding does not become inapplicable merely because it crossed a border. Nor does evidence become automatically applicable because it was produced within the same country.

What matters are the characteristics capable of changing the finding. Two cities in different countries may share the relevant population, infrastructure, institutions, and baseline conditions. Two communities in the same country may differ substantially on those same dimensions.

Assess substantive differences rather than using geography as a proxy for them.

Local evidence is necessary when you need a local quantity

Some research questions are inherently about the setting. If you need to know the prevalence of a condition in one province, the proportion of students using a technology in one university, the local cost of implementing a programme, or the number of qualified staff available, evidence from elsewhere cannot provide the required value.

International studies may help define the variable or suggest how to measure it. They cannot establish the local quantity itself.

Local evidence is important when the population differs in a way that could modify the finding

Population differences matter when they are connected plausibly to the outcome, not merely because demographic profiles differ.

Depending on the question, relevant factors might include age, baseline risk, prior exposure, socioeconomic conditions, language, comorbidities, educational background, digital access, occupational characteristics, or other features capable of modifying the relationship or intervention effect.

If the international evidence already represents similar populations or demonstrates reasonably consistent findings across the relevant variation, less local verification may be necessary. If the local population falls substantially outside the populations studied, uncertainty increases.

Local evidence is necessary when baseline conditions determine the practical effect

A relative effect can remain similar while its practical consequences differ substantially.

Guidance on applicability highlights baseline conditions because they influence absolute effects. An intervention that reduces a risk proportionally may prevent many events in a high-risk population but few in a low-risk population. Likewise, an educational intervention may produce different practical gains depending on baseline performance or existing provision.

If your decision depends on the expected absolute benefit, local baseline information may therefore be necessary even when the relative effect is well established internationally.

Local evidence is necessary when implementation depends on local systems

Many interventions work through systems rather than in isolation. Their delivery may depend on staffing, infrastructure, financing, governance, supply chains, institutional authority, professional practice, technology, training, or coordination among organizations.

Applicability guidance explicitly asks whether differences in on-the-ground constraints or system arrangements could alter feasibility, acceptability, or how an option works.

If those conditions are central to the intervention, you need evidence about whether they exist locally. This may involve administrative records, resource assessments, implementation research, interviews, observations, or pilot work rather than a full effectiveness study.

Local evidence is necessary when acceptability and behavior are part of the causal pathway

Some interventions depend heavily on how people perceive, adopt, trust, or use them. An intervention that is technically effective when used as intended may have little practical impact if local users reject it, professionals resist it, or cultural and institutional norms alter participation.

Local qualitative evidence can be especially useful for understanding values, social and cultural norms, stakeholder preferences, barriers, and implementation risks. Recent WHO guidance on evidence-informed theories of change similarly distinguishes the contribution of local qualitative evidence for understanding social support, norms, values, preferences, and implementation barriers.

When these factors are plausibly central to whether the finding will translate into practice, assuming acceptability from another setting is difficult to justify.

Local evidence is necessary when the intervention itself will be meaningfully adapted

Sometimes international evidence concerns an intervention that cannot be implemented locally in its original form. Components may need to be removed, translated, delivered by different personnel, shortened, digitized, rescheduled, or integrated into a different institutional system.

Minor adaptations may preserve the intervention's relevant mechanism. Major adaptations may create something meaningfully different from what was evaluated.

WHO guidance on guideline contextualization emphasizes that applying evidence-based guidance in another setting requires consideration of available resources and organizational context rather than simple transfer. If adaptation changes components believed to produce the effect, additional local evaluation may become particularly important.

Local evidence is necessary when the measurement itself may not travel

International findings can depend on instruments, classifications, cut-off values, survey items, diagnostic procedures, or constructs that do not function identically in every language and population.

Translation alone does not establish measurement equivalence. If a construct is culturally sensitive or the instrument's interpretation depends on language, education, norms, or local practice, evidence that the measure performs adequately in the target population may be necessary before treating scores as directly comparable.

The relevant local evidence in this case may be validation or measurement-equivalence evidence rather than a replication of the substantive association you ultimately want to study.

Local evidence is especially important when the consequences of being wrong are substantial

Applicability is not only a statistical question. The acceptable level of uncertainty can depend on the consequences of the decision.

If an intervention is inexpensive, reversible, low risk, and easily monitored, decision makers may reasonably tolerate greater uncertainty about local applicability. If implementation is costly, difficult to reverse, potentially harmful, or likely to affect vulnerable populations, unresolved contextual uncertainty becomes more consequential.

This does not create a mechanical threshold for when a local trial is required. It means that the amount and type of local evidence should be proportionate to the uncertainty and consequences involved.

Local evidence is less necessary when the relevant conditions are already well represented

Requiring a local replication of every established finding would waste research resources and can produce studies with little additional informational value.

If strong evidence includes populations and settings similar to yours, findings are consistent across meaningful contextual variation, the mechanism is unlikely to depend on a distinctive local factor, and implementation conditions are sufficiently comparable, there may be little reason to demand another local study merely for geographic reassurance.

WHO's discussion of implementation research similarly notes that local implementation research requires reasonable doubt about contextual differences rather than assuming that every change in location creates a new unanswered question.

The local evidence you need may be much smaller than a full replication

This is one of the most useful distinctions in practice.

Suppose an intervention has a well-established effect, but you are uncertain whether your institutions have enough trained personnel to deliver it. The missing evidence concerns workforce capacity, not intervention efficacy. Conducting another effectiveness study before determining whether the intervention can even be delivered would answer the wrong question.

Likewise, if uncertainty concerns local cost, collect cost data. If it concerns acceptability, study acceptability. If it concerns baseline prevalence, estimate prevalence. If it concerns instrument validity, validate the instrument.

Local research should target the uncertainty that actually blocks application.

How to identify the local evidence that is actually necessary

The need for local evidence arose from specific contextual differences, not from the mere fact that the original studies were conducted abroad.

Common mistakes when deciding whether local evidence is necessary

Ask what local fact could make you change your conclusion

A practical applicability assessment can begin with one counterfactual question: What could I learn about this setting that would make me interpret or apply the international finding differently?

This framework also clarifies why strong international evidence can sometimes remain highly useful despite very limited local research. The missing local literature matters only insofar as it leaves an uncertainty that is relevant to the claim or decision.

If such an uncertainty exists, collect evidence targeted to it. If it does not, another local replication may produce a familiar finding with a new postal address, which is not quite the same thing as producing new knowledge.

Before applying an international finding locally

Questions about when local evidence is necessary

Require local evidence for a reason, not merely for a location

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.