Climate change and health indicators. Definitions and selection criteria

1. Analytical summary

A framework to make visible the health impacts of climate

The report is part of the 3rd National Adaptation Plan (PNACC-3), which prepares France for a trajectory of +4 °C by 2100 and makes health protection a primary objective (p. 6). There are high expectations for territorialised indicators linking climate and health. But an indicator cannot meet all uses: each must be defined for a specific objective, context, and use (p. 6). The stakeholders involved range from public health agencies and epidemiologists to elected officials, local authorities, and the general public, the final recipients of the indicators (p. 14). The central issue is to reduce the subjectivity of the choice of indicators to increase their relevance, transparency, and acceptability (p. 7).

Three families of criteria, transcribed into grids

The document provides a conceptual framework (adapted from McMichael and the WHO DPSEEA model, incorporating vulnerability and adaptation, p. 7-9), a typology of five types of indicators (hazard, exposure, vulnerability, impact, adaptation/intervention, p. 11), and above all rating grids in three classes covering three blocks: climatic relevance (eliminatory if the link is weak), metrological quality (9 criteria) and pedagogical quality (3 criteria) (p. 12-14). It also sets out strong usage precautions: these indicators show impacts that have already been observed, do not predict the future, and do not measure the effectiveness of an adaptation measure (p. 15).

2. Key points of the document

1️⃣ Expanded definition — an indicator "climate-health" provides quantitative and synthetic information on a health issue influenced, directly or indirectly, by climate or by adaptation/mitigation policies; the novelty in 2026 is the explicit integration of the co-benefits of climate policies (p. 10).

2️⃣ Five types of indicators — danger, exposure, vulnerability, impact, adaptation/intervention, each illustrated (e.g. deaths attributable to heat; deaths avoided by modal shift car → bike) (Table 1, p. 11).

3️⃣ Relevance with double lock — three criteria of climate relevance, two of which are eliminatory: the proven link to the climate and the strength of that link. An issue with a link to the climate that is too weak is dismissed outright (Table 2, p. 12).

4️⃣ Nine metrological criteria — transparency, validity, reproducibility, temporal / geographical / socio-economic precision, frequency of updates, cost, international comparability. No criterion is individually eliminatory: it is the accumulation of low scores (more than 5) that disqualifies (Table 3, p. 13).

5️⃣ Explicit usage precautions — an indicator informs for a given period and territory, does not prejudge future trends (non-linear developments) and never captures the entirety of effects — an indicator of heat-related mortality ignores delayed, mental and systemic effects (p. 15).

3. Action tracks for local actors

Note : the document being methodological, these tracks concern the reading, selection and communication of indicators — not on a direct intervention with the public.

1️⃣ Reflex of critical reading — before sharing a “climate-health” figure, check its type (danger, exposure, impact…) and its intended use, to avoid making it say what it does not (Table 1 p. 11; §3.1 p. 15).

2️⃣ Relevance filter — use the three relevance criteria (Table 2, p. 12) as a quick test against a piece of data: is the link to the climate proven and strong? If not, calibrate caution in communication.

3️⃣ Choice of indicators — rely on the criteria of clarity, narrative framing and reusability (Table 4, p. 14) to decide which indicators to present to a non-expert audience (elected officials, residents) without misinterpretation.

4️⃣ Structure a local diagnosis — mobilise the vulnerability framework (Figure 2, p. 9): cross exposure, individual susceptibility and capacity to act, with explicit attention to social health inequalities.

5️⃣ Frame the expectations — set from the outset the precautions for use — an indicator does not predict the future and does not measure the effectiveness of an action (p. 15) — to maintain realistic expectations with partners and elected officials.

6️⃣ Fill the unmet needs — the document provides neither ready-to-use indicators, nor local data, nor facilitation tools. Complement it with already published indicators and the operational resources listed in the References section.

4. Additional references

External references (post-2024, verified URLs)

• Committee 21 (2024). Adaptation to climate change in the health sector — Practical guide, 52 p. Intended for managers and professionals of health and social care establishments: direct operational complement. Access to the sheet (ACC Resource Centre)

• Ministry of Ecological Transition (2025). 3rd National Adaptation Plan for Climate Change (PNACC-3), 52 measures, trajectory +4 °C. Public policy framework in which the analysed report is situated. Access the PNACC-3 page

• UN DESA (2024). Global set of climate change statistics and indicators — Implementation guidelines, 94 p. International methodological supplement on the indicators (harmonisation). Also appears in the bibliography of the report [4] — noted for transparency. Access the PDF (unstats.un.org)

Internal References Health Practices

• "The 2025 report of the Lancet Countdown on health and climate change" (Oct. 2025). - https://pratiquesensante.odoo.com/en/blog/studies-5/le-rapport-2025-du-compte-a-rebours-de-lancet-sur-la-sante-et-le-changement-climatique-5309 

• "Climate action and health: a shared urgency, shared benefits" (Jan. 2026). https://pratiquesensante.odoo.com/en/blog/studies-5/action-climatique-et-sante-une-meme-urgence-des-benefices-partages-5569 

5. Frequently Asked Questions (FAQ)

1️⃣ What is the practical use of a "climate-health" indicator?

To provide quantitative and synthetic information on a health issue influenced by climate, to encourage stakeholders to integrate health into adaptation and mitigation policies (p. 2 and p. 10).

2️⃣ Can one indicator measure everything?

No. Each indicator is defined for a specific objective, context, and use; none meets all needs (p. 6).

3️⃣ Is it necessary to prove that a given extreme event is due to climate change?

No, for state and exposure indicators: science establishes that these events are becoming more frequent and intense; it is not necessary to attribute each individual event (p. 10).

4️⃣ Can we calculate the share of deaths "due" to climate change?

The definition does not impose it. The calculation is very complex for the direct effects (lack of robust studies, difficulty in separating the trends) and currently impossible for the indirect effects (p. 10).

5️⃣ How do we decide that an indicator is retained?

It must first be relevant (link with the climate proven and strong — eliminatory otherwise), then present sufficient metrological and pedagogical qualities; an imperfect but improvable indicator can be retained (p. 12-14).

6️⃣ Does an indicator allow us to predict future trends?

No. It informs for a given period and territory; climate changes are not linear (p. 15).

7️⃣ Do these indicators measure the effectiveness of an adaptation measure?

No, it is not their use. They illustrate the impacts of climate change on health (p. 2 and p. 15).

6. Rewriting in Easy to Read and Understand (FALC)

What does this document talk about?

• The climate is changing. It affects our health.

• This document comes from Public Health France.

• It explains how to choose good "indicators".

• An indicator is a number that shows a situation.

The main ideas

• An indicator shows a link between climate and health.

• There are 5 types of indicators. For example: the number of deaths during a heatwave.

• A good indicator has a clear and strong link with the climate.

• A good indicator is reliable. It can be repeated and verified.

• A good indicator is easy to understand.

What to pay attention to

• A indicator shows the past. It does not predict the future.

• A indicator does not say everything. It forgets certain effects.

• A indicator does not say if an action is working well.

What is it useful for you?

• It helps you to read the figures on climate and health well.

• It helps you to choose the right figures to show to the public.

7. Cross-analysis — values of Health Practices

• Literacy : yes : the "pedagogical quality" (clarity, storytelling) aims for understanding by non-experts, elected officials and the general public (p. 14).

• Empowerment : low and indirect: the target is the production of indicators by experts; the end beneficiaries do not participate in the design, even if autonomous appropriation by users is aimed for (p. 14).

• Participation : the scoring benefits from being collective for consensual choices (p. 11), but no co-construction with the concerned publics is described.

• Community health: collective dimension present through social health inequalities and the vulnerability of groups (elderly, disadvantaged, p. 11 and p. 15), without an explicit community approach.

• Ethics : biases are addressed methodologically (subjectivity of the scoring, transparency, p. 7 and p. 13); cultural biases of the publics are not specifically addressed.

• Human rights / equity: attention to social health inequalities is recurrent and integrated as a criterion (socio-economic precision, p. 13 and p. 15).

• Intersectorality : strong : cross-cutting approach mobilising health-environment, work, infectious diseases, statistics, and the logic of "One Health" (p. 7).

• Partnership : formalised mainly at the international level (IANPHI, Lancet Countdown, european observatory) rather than local (p. 15).

• Fight against discrimination: addressed indirectly through social health inequalities; no explicit mention of discrimination or non-judgment (p. 13 and p. 15).

8. Assessment of the reliability of the resource

Scientific relevance — high. Institutional report (Public Health France), specialist author, multidisciplinary working group, external review. Framework supported by strong international references (McMichael, WHO/DPSEEA, IPCC, UN, Australian and British agencies). Bibliography dated and coherent. Caveat: short report (16 p.), deliberately conceptual; publication from August 2026 not cross-checked online at the time of the analysis.

Operational relevance — indirect. Usable as a reading and selection grid for indicators, and as training support for data literacy. Does not provide either ready-to-use indicators, local data, or action protocols. Real added value upstream of action: diagnosis, communication, advocacy.

9. Strategic hashtags

#healthpractices #HealthEnvironment #ClimateChange #DataLiteracy #HealthIndicators #OneHealth #SocialInequalities #ClimateAdaptation