Guidebook on Best Practices in Public Health

🧭 Choose an intervention from elsewhere, check that it can work at home, implement it and then evaluate it: the OECD details a 3-step approach, with grids and examples mainly drawn from European countries.

. This guide from the OECD presents itself as a tool to make the selection, implementation and evaluation of public health interventions more rigorous (p. 3). It offers grids to check if a successful action elsewhere can be transferred at home (p. 28 to 36). It also provides a list of elements of a implementation protocol (p. 56) and guidelines to plan for evaluation from the outset (p. 39, 63). Examples drawn from member countries make the approach concrete (p. 7). It was designed for decision-makers, but the guide also claims to be useful for those who plan, carry out or support an intervention (p. 10).

1️⃣ SUMMARY

In the face of the rise in non-communicable diseases, mental health issues, air pollution and antibiotic resistance, aggravated for some by COVID-19, particularly mental health (p. 3, 7), the OECD proposes a three-step guide inspired by the Plan-Do-Study-Act cycle (p. 10). Step 1 (“select”) identifies the best practices using a multi-criteria analysis based on 5 criteria. It then assesses their transferability across 4 contexts: population, sectoral context, policy, economy (p. 12-39). Step 2 (“implement”) organises the preparation and execution according to the logic of “who does what, when” (p. 40-60). Step 3 (“evaluate”) describes the logical model, the choice of indicators, study and data collection methods, the analysis and use of results (p. 61-79). The guide targets public decision-makers and is applicable, according to it, to all types of public health interventions (p. 3, 10). Real-life experiences from member countries illustrate it (p. 7).

2️⃣ WHAT TO REMEMBER

1️⃣ Five criteria to define a best practice. Effectiveness in real-world conditions (Effectiveness), efficiency (Efficiency), equity (Equity), strength and validity of evidence (Evidence-base) and coverage of the target population (Extent of coverage): the “5 E” in English (table 1.1, p. 14). . Equity is described as “a fundamental requirement of any public health intervention” (p. 17). An intervention not yet fully evaluated but with a solid theoretical foundation can be considered “promising” (p. 15).

2️⃣ A best practice here is not necessarily one elsewhere. The guide provides 4 Yes/No grids comparing the place of origin and the target place (tables 1.6 to 1.9, p. 28-36). Three possible outcomes: transfer as is, transfer an adapted version, or not transfer and return to step 1a (p. 38).

3️⃣ Fidelity and adaptation do not necessarily oppose each other.It is necessary to define the "essential components" (core components) of the intervention, preferably from the transferability analysis and at the latest during the preparation. They must be respected, while others can be modified to fit the context, keeping track of the changes (box 2.1, p. 43).

3️⃣ TO ACT

1️⃣ Check the transferability of an action identified elsewhere with the grids from tables 1.6 to 1.9 (p. 28-36) and the example of questions from box 1.16 (p. 37-38). Case : existing tool cited by the document. The guide specifies that these indicators are a simple point of reference, to be completed or simplified according to the intervention (p. 25).

2️⃣ Follow up the implementation with a short reporting template filled regularly by the participants. Excerpts: what have you done this month? what successes? have there been any obstacles? was any support necessary? what do you plan for next month? to what extent are you sure you will achieve it? (box 2.12, p. 58; pilots CHRODIS-PLUS in Iceland, Lithuania and Spain). Case: existing tool cited by the document.

3️⃣ Reread a guide for an interview before an evaluation with the checklist from box 3.9: structure, one question at a time, open and non-judgmental questions, no jargon (p. 73). Case : existing tool cited by the document.

Suggestion Health Practices: the grids being in English, a working translation of tables 1.6 to 1.9 and box 3.9 can facilitate their use in teams.

4️⃣ 3 PRACTICAL QUESTIONS

1️⃣ Can we retain an action that has not yet been fully evaluated? Yes : the guide encourages not to dismiss these interventions outright, as results can take years to appear. With a solid theoretical foundation, they can be classified as “ promising”. The same applies in cases of limited efficiency data (p. 15, 16).

2️⃣ Is a randomised trial essential for evaluation? No. The randomised trial remains the reference, but it is often not applicable in public health (ethics, cost, contamination of the control group). The guide presents other methods: non- randomised trial, cluster randomised trial, before-after study, noting their limitations (p. 18, 69-70). A well- designed cohort study may be better than a poorly designed randomised trial (p. 19).

3️⃣ When should data be collected? At the beginning, before any activity; during the intervention; at the end and, if resources allow, some time later to observe a lasting effect (p. 75, 79) . The guide notes that this follow-up after the end is often absent due to lack of resources (box 3.11, p. 75).

5️⃣ TO GO FURTHER

Blog Health Practices: Using evidence (for decision-making): what are the effective practices?. Link identified by search engine indexing; automatic access blocked by the site. To be checked manually before publication.

Recent external resource: High Health Authority, "Evidence-based data to accelerate the preventive shift", prospective analysis and press release of 12 November 2025. It offers an evaluation toolkit (reference framework and practical guide) and a national space for dialogue between science and decision-making. has-sante.fr/jcms/p_3737647 (page viewed and verified, October 2026).

6️⃣ #️⃣ HASHTAGS

#healthpractices #HealthPromotion #Transferability #EvaluationOfInterventions #ProvenInterventions #Implementation #HealthEquity #OECD

To go further, order a research or documentary watch 

This article: An editorial production based on experience enhanced by artificial intelligence - The English version of the site is an automatic translation: the French version is authoritative - Click on the image