Living evidence syntheses are less a new statistical method than a new operating rhythm for established methods. They rest on the same rigorous machinery as conventional systematic reviews — comprehensive literature searches, explicit inclusion criteria, risk-of-bias appraisal and structured synthesis — but instead of being frozen at the moment of publication, they are maintained as a rolling surveillance process. When new randomized trials or observational studies appear, the synthesis team screens them, re-runs analyses where necessary and refreshes the graded certainty of the evidence, often on a weekly or monthly cycle. The approach proved itself during the pandemic, powering Australia’s weekly updated national living guidelines for COVID-19 care and the World Health Organization’s living guideline on the clinical management of COVID-19. The same logic now extends into living health technology assessment, in which the value of a drug or device is recalculated as real-world data accumulate. In principle, this gives decision-makers something a traditional review cannot: a permanent, current and critically appraised answer to a standing policy question.

Yet whether that promise survives contact with real policy processes has been largely unknown. To find out, a team led by Samantha Chakraborty of Cochrane Australia at Monash University’s School of Public Health and Preventive Medicine — working with colleagues at the World Health Organization, McMaster University in Canada and the National Institute for Health and Care Excellence in England — conducted a two-stage qualitative study. Between February 2024 and June 2025, the researchers carried out semi-structured, in-depth interviews with 18 people who had used, or tried to use, living evidence syntheses in health policy, either directly as decision-makers and public servants in government, or indirectly as knowledge intermediaries outside government and researchers producing syntheses for policy use. Participants were recruited through purposive sampling supplemented by snowball sampling and spanned every world region, although none worked primarily in a low-income country. In July 2025, nine participants — five of whom had also been interviewed — joined facilitated roundtables in which they validated the emerging findings and proposed practical strategies, bringing the total to 22.

The analysis was deliberately mechanical in the scientific sense: the team wanted mechanisms, not just opinions. First, they coded every account of successful or attempted evidence use against Proctor’s Implementation Outcomes Framework, a taxonomy from implementation science that separates eight dimensions of uptake: acceptability, appropriateness, feasibility, adoption, fidelity, cost, penetration and sustainability. These clustered into two tiers — proximal outcomes such as acceptability, appropriateness and feasibility, which must be satisfied before anything else can happen, and distal outcomes such as adoption, fidelity, cost, penetration and sustainability, which follow only once the foundations hold. Barriers and facilitators were then mapped onto a socio-ecological framework spanning four levels of influence: the individual policy-maker, the policy-maker’s inner network of advisers and public servants, the local policy setting, and the wider global research and policy ecosystem. Finally, using a causal pathway mapping approach, the investigators traced how specific strategies could plausibly travel through those barriers to deliver the intended outcomes, refining draft mechanisms iteratively until the whole team agreed.

Across the interviews, one determinant towered above all others: trust. Living evidence was judged acceptable only when it was demonstrably trustworthy, useful and usable — and that trust depended heavily on candid communication between the people producing the syntheses and the people making decisions. Because a living synthesis can change its conclusions from one month to the next, unmanaged expectations can become corrosive. “People sometimes expect you to give them answers on something where the evidence is not completely settled,” one public servant in a high-income country told the researchers. “And if the communication is not strong and clear, people would actually lose trust. Like, a month ago, you told us to do this and now you’ve changed your opinion.” When evidence shifted from supporting a “yes” decision to a “no” after a policy had already been made, participants reported, it created “big political problems” that degraded relationships and suppressed the willingness to rely on living evidence again. Relevance mattered just as much: syntheses rarely existed for the precise questions policymakers were asking, and evidence that could not be adapted to the local context was frequently unusable.

Appropriateness turned out to be a negotiation between two clocks. Researchers can now refresh evidence on demand, while policy cycles move at the speed of legislatures, constituents and ministries — often far more slowly. One knowledge intermediary in an upper-middle-income country worried openly about the mismatch: “There is just one tiny worry that we start to get very quick, very timely, and then we turn the relationship around being very pushy and bossy and not respecting the policy cycle moment.” A policy decision-maker in a high-income country was blunter about syntheses that arrive without relationship or context: “That’s great, it’s totally ignoring the realities of my day to day… It’s really hard to get there without that trust.” Participants also flagged a downstream risk unique to living evidence: if a synthesis keeps updating after a policy is approved, the policy itself may need repeated revision, stalling ministry plans. In low- and middle-income settings, imported evidence could even backfire. One public servant reported that when data came predominantly from high-income countries, the living synthesis “actually increased the ‘distrust’ of that critical data” among local users.

Feasibility emerged as the hard constraint. Producing and maintaining living syntheses demands sustained funding, skilled staff and technical infrastructure — and those are scarce even in wealthy systems. “We don’t have enough people to keep up the work, we don’t have the funding, and then we don’t have the tools,” one researcher in a low- and middle-income country said plainly. Participants also identified a structural bottleneck upstream: academic incentives. Because journal metrics reward publication, primary trial results are frequently withheld from synthesis teams until peer review is complete. One researcher recounted a “flagrant example” in which a major trial appeared first as a press release, but the investigators refused to share the underlying data for synthesis because “they wanted this to be in the paper to be published, to get the citations, to get the attention.” Even where a global synthesis existed, local adaptation imposed its own costs — updated local data, local cost figures, translation, and the convening of expert panels, which one upper-middle-income participant described as “a big amount of resources.”

The distal outcomes painted an equally nuanced picture. Adoption of living evidence rose where political will and internal champions fostered a culture of evidence use, and where established communication channels guaranteed that updates would actually reach decision-makers — something most systems currently lack. As one high-income public servant admitted, “Right now, we don’t have a plan for when those things get updated, so we’re not directly connected to any of those guidelines or policies.” Fidelity — keeping a synthesis’s conclusions intact as they are translated into different contexts — required producers to supply finer-grained detail and, in the words of one participant, to “hold the judgement” so that end-users could judge for themselves, since good decisions sometimes diverge from the evidence for legitimate local reasons. Cost loomed throughout the discussions. “I think living evidence is a little bit expensive probably at the moment,” one participant said, warning that commissioners must effectively “take a big leap [with] public money” without demonstrated return on investment. Staff turnover compounded the fragility: “Every time we have a staff change, priorities change, and then you are likely to start again.”

From this tangle of barriers, the team distilled six mechanisms through which strategies can genuinely make living evidence work in policy: understanding the needs of the people who will use the syntheses; increasing equitable access to useful syntheses; increasing reliable access to them; enhancing users’ capacity to apply what is available; enabling health systems to implement the policy decisions that living evidence informs; and building users’ confidence in deciding when and how to use living evidence at all. Sixteen concrete strategies mapped onto these mechanisms, and notably, none were aimed at a single group. They were framed as shared system goals, with joint responsibility distributed across producers, intermediaries, funders and policymakers — from exemplar demonstrations that show a skeptical ministry what a living synthesis can deliver, to regional step-by-step guides that support country-level implementation, to trusted bi-directional relationships that keep synthesis production aligned with real policy questions. Strikingly, participants from different continents and income levels proposed largely consistent solutions, a uniformity the authors attribute to the field’s novelty, though experiences may diverge as living evidence becomes embedded differently across contexts.

The implications reach well beyond the study’s participants. The authors argue that as living evidence infrastructure matures globally, the needs of low- and middle-income countries must be engineered in from the start; otherwise, disparities in data availability, language, technology and capacity could widen rather than narrow. They also see direct relevance for the World Health Organization, which is increasingly adopting living evidence processes and now faces the downstream challenge of translating continuously updated evidence into products that country-level decision-makers can actually deploy. The study carries honest limits: most participants were knowledge intermediaries or public servants rather than officials holding final decision-making authority, and no participant with direct professional experience in a low-income country joined, meaning some context-specific insights are necessarily missing. The authors call for systematic testing of the proposed strategies across diverse policy settings, drawing on political science and implementation science. If those tests succeed, the payoff could be substantial: policy that is not merely informed by evidence at a single frozen moment, but continuously nourished by it — and health systems that improve because their decisions always rest on the most current truth science can offer.