La co-creazione come strumento decisionale nel progetto PNC "Atlante Aria e Salute" per definire i contenuti di un atlante integrato a supporto delle decisioni e della ricerca.

BACKGROUND: the 'ATLAS Aria Outdoor e Salute' (Outdoor Air and Health Atlas) project aims to develop a national atlas to estimate the effects of air pollution on health, making it accessible to both professionals and citizens and available as open data. OBJECTIVES: to describe the co-creation decision-making process adopted to define the contents of the Atlas. DESIGN: two rounds of questionnaires with open-ended questions were conducted following the Delphi method, collecting proposals on possible options for data sources, territorial disaggregation, exposures, health outcomes, methods for calculating attributable fractions, and counterfactual scenarios on which to estimate the potential benefits of policies. The Delphi process started by selecting health outcomes and pollutants for which dose-response functions were already available in the scientific literature. Participants were then given the opportunity to propose additional outcomes or pollutants for inclusion in the Atlas. The responses collected in the questionnaire administered during the first round, and subsequently discussed in a plenary session, informed the development of the second-round questions. The results of the second round identified the health outcomes and pollutants eligible for inclusion in the Atlas, which were then discussed in a plenary session held with the participation of external experts. SETTING AND PARTICIPANTS: internal members of the working group participated in the first round; the second round included, in addition, other stakeholders and representatives from Italian Regions not involved in the Atlas. Results were discussed in online and in-person meetings with experts and stakeholders until an agreement was reached. MAIN OUTCOME MEASURES: for each proposed health outcome and pollutant, it is assessed: the level of certainty regarding the causal link between the pollutant and the health outcome; the relevance for public health; the accuracy of the information available in routine databases; the usefulness of including that outcome in the Atlas maps. For all other parameters, the level of agreement reached across the two Delphi rounds and the plenary meetings was evaluated. RESULTS: sixteen experts participated in the first round, 22 in the second. The process led to the definition of data sources to be used (cause-specific mortality data from the Italian National Institute of Statistics, hospital discharge records for hospitalisations), the spatial resolution of exposure estimates (1×1 km) and the model used to derive them (annual averages produced by a previously validated and published model from BIGEPI project), the level of data aggregation for the analyses (municipal level), and the pollutants to be considered as exposures for calculating the attributable fraction (PM2.5 and NO2). Both long-term and short-term effects were considered. Considering the available data sources, the outcomes that met the criteria of certainty, public health relevance, and exposure accuracy were, for long-term effects, mortality and hospital admissions due to natural, cardiovascular, cerebrovascular, and respiratory causes, as well as lung cancer; for short-term effects, the same outcomes were identified except for lung cancer. The outcomes to be included in the Atlas are mortality and incidence or, where relevant, prevalence. It was further decided to show the attributable fraction for each pollutant, explicitly illustrating the steps involved in its derivation: starting from pollutant concentrations and the resident population, the population-weighted exposure (PWE) is calculated. By combining the PWE with the disease burden and the dose-response function provided in the literature, it is then possible to estimate the attributable fraction by comparing the current scenario with a counterfactual scenario in which pollution levels correspond to WHO guideline values (5 µg/m³ for PM2.5 and 10 µg/m³ for NO2). In addition, it was decided to allow for the estimation of the number of attributable deaths under alternative, potentially achievable future pollution levels. Regarding the choice of outcomes to be included, a comparison with the most recent systematic review was conducted during the final plenary meeting, which confirmed the decisions made. CONCLUSIONS: the co-creative process adopted allowed for broad participation and sharing of content while maintaining consistency with the best available scientific evidence.