Healthcare implementation studies still struggle to establish value for money

A systematic review of healthcare implementation research found that most studies document what implementation strategies cost but do not test whether those costs produce better health outcomes. Among 22 eligible studies published from 2016 onward, only six included full economic evaluations linking resource use to effectiveness, health gains or monetized benefits. The other 16 were cost-only analyses, limiting their ability to show whether strategies represent good value for money or can be sustained when scaled. The findings suggest that implementation research has documented resource requirements more often than it has answered basic questions about affordability, efficiency and transferability.

The review was registered prospectively in PROSPERO as CRD420251018265 and reported under Preferred Reporting Items for Systematic Reviews guidelines. Investigators searched PubMed, Scopus, PsycInfo, PsycARTICLES, the International Clinical Trials Registry Platform, the International Standard Randomized Controlled Trial Number registry and ClinicalTrials.gov. The initial database search was conducted on April 6, 2025, and the search was updated on June 15, 2026. They screened 5,729 records, removed 691 duplicates and assessed 5,038 titles or abstracts, ultimately retaining 22 peer-reviewed, English-language studies. Eligible studies evaluated strategies intended to promote uptake of an evidence-based practice by healthcare providers in hospitals, primary care, community clinics and other healthcare settings. Studies in schools, patients’ homes and other non-healthcare environments were excluded, as were implementation efforts without economic outcomes.

Training and education were the most frequently evaluated implementation strategies, either alone or combined with supervision, coaching, technical assistance, booster sessions, electronic health record support, peer navigation, performance feedback or redesigned workflows. These approaches differed greatly in resource intensity. Brief educational sessions and modest modifications to clinical records generally required fewer personnel and material inputs, whereas coaching, facilitation, specialist supervision and multicomponent health-system strengthening required sustained staff time, coordination and support. Studies that described an approach simply as training often failed to distinguish its duration, intensity, delivery mode and cost. That imprecision limited comparisons among strategies, even when the studies concerned similar clinical practices. Many evaluations also combined implementation costs with the costs of delivering the clinical intervention, making it difficult to determine the additional resources required for uptake.

Of the six full economic evaluations, three used cost-effectiveness analysis, one used cost-utility analysis, one used cost-benefit analysis and one used cost-minimization analysis. Four of the studies considered incremental cost-effectiveness ratios, but their outcomes were not standardized: one used disability-adjusted life years, one used life-years gained and two used implementation-specific endpoints, such as cost per additional positive screening or per additional person initiating treatment. In two cases, the intervention model was both less costly and more effective than its comparator, precluding calculation of an incremental cost-effectiveness ratio. No study used quality-adjusted life years, and the scarce use of standardized health metrics limits comparison across interventions and settings. The authors cautioned that one outcome metric cannot fit every implementation problem; the endpoint must instead match the decision being made.

The cost-benefit study monetized benefits over five years from market, private, societal and stakeholder perspectives and found that benefits exceeded costs. The cost-minimization study was undertaken because clinical outcomes and quality of life were equivalent between groups; it found the physician-led medication-review intervention was cost-neutral over an average of seven months while reducing prescribed medications, potentially inappropriate prescriptions and telephone contacts. One cost-effectiveness analysis combined trial-based costs with microsimulation models to estimate lifetime disease incidence, life-years gained and healthcare costs. It was the only full evaluation in the review to extrapolate beyond the study period. Its modeled results suggested that a theory-based implementation approach was more effective and less costly than an intuition-based approach, but this was a model-based estimate rather than a directly observed long-term comparison.

Cost-only studies nevertheless offered useful descriptions of resource requirements. One activity-based costing analysis estimated US$374,863 for peer coaching across seven Veterans Affairs medical centres, with travel and personnel time accounting for most of the expenditure. Another estimated first-year implementation costs of approximately US$580 to US$1,482 for a paediatric oral-health intervention, depending on practice size, followed by lower subsequent annual maintenance costs. Across the review, personnel costs were consistently a major category, reflecting time spent on training, supervision, programme delivery, workflow development, coordination and coaching. Researchers estimated these costs through interviews, questionnaires, staff diaries, activity logs, observations, time-and-motion studies, project records, staffing allocations, programme expenditure data and assumed time inputs. The studies therefore captured several dimensions of workforce effort, but not always with the same level of precision.

Reporting weaknesses restricted the policy usefulness of much of the evidence. Only a minority of studies explicitly stated their analytical perspective, such as the healthcare system, society or a stakeholder group, while time horizons ranged from short start-up periods to several years and were sometimes not reported. Capital costs and patient costs were rarely detailed. Implementation costs, including training time and technical assistance, were also not consistently separated from the costs of delivering the clinical intervention itself. Research-related expenses such as evaluation and data collection were frequently excluded or left unspecified. These ambiguities made it difficult to determine the total resource requirement of implementation and to transfer findings to other settings. A larger expenditure could not be classified as either inefficient or justified without corresponding evidence about effects.

The six full economic evaluations were assessed with the 35-item Drummond checklist. Three were classified as high quality because more than 75% of applicable items were rated yes: the Lynch syndrome evaluation at 86%, the multicountry tuberculosis study at 80% and the generalist screening model at 77%. The other three were rated moderate quality, at 74%, 71% and 60% for the iron and folic acid supplementation, reablement and medication-review analyses, respectively. None was classified as low quality. Because several checklist domains do not apply to cost-only studies, the reviewers did not assign those 16 studies equivalent overall quality or risk-of-bias ratings. Instead, they assessed them descriptively for reporting of analytical perspective, time horizon, cost categories, cost boundaries and the identification, measurement and valuation of resource use.

The evidence was geographically uneven. Twelve studies came from the United States, three from Australia and two from Uganda, with additional studies from Bangladesh, Malawi, Zambia and Denmark. One multicountry study covered sites in Benin, Canada, Ghana, Indonesia and Vietnam. The multinational tuberculosis project was especially notable because it prospectively collected resource-use and cost data alongside implementation and effectiveness outcomes. It estimated a combined intervention and latent tuberculosis infection clinical-care cost of CA$1,348 per additional household contact initiating preventive treatment. The presence of studies in lower- and middle-income settings demonstrated that economic analysis is feasible in diverse health systems, but their underrepresentation limits broader conclusions. The reviewers also emphasized that institutional priorities, political considerations, workforce availability and patient preferences can shape real-world decisions alongside economic evidence, even when rigorous cost and outcome data are available.

The review has important limitations. One author screened titles and abstracts because the project lacked dedicated funding for a second reviewer at that stage, although both authors reviewed full texts and resolved discrepancies through discussion. Restricting inclusion to English-language studies in healthcare settings may also have omitted relevant work from social care, community settings and other languages, and inconsistent terminology may have caused additional studies to be missed. Incomplete reporting prevented quantitative synthesis. The findings therefore describe the present state of economic reporting and methodological practice rather than establish that any particular implementation strategy is more cost-effective than another. The authors conclude that rigorous economic evaluations should be designed alongside implementation studies, with prospectively collected costs and outcomes, explicit perspectives and time horizons, transparent cost boundaries and outcome metrics suited to the decision at hand. Without that integration, implementation research may establish that practices can work without demonstrating that organizations can afford to put them into practice.

Subject of Research: Economic evaluation methods and reporting in healthcare implementation science

Article Title: Economic evaluation in healthcare implementation: a systematic review

Article References: Ulfsdotter Gunnarsson, K., & Nilsen, P. (2026). Economic evaluation in healthcare implementation: a systematic review. Frontiers in Health Services, 6, Article 1838074. https://doi.org/10.3389/frhs.2026.1838074

Image Credits: AI Generated

DOI: 10.3389/frhs.2026.1838074

Keywords: implementation science, healthcare economics, cost-effectiveness, cost-utility analysis, economic evaluation, implementation strategies, health outcomes, systematic review, health policy, resource use, Economic, evaluation