OBJECTIVES: To calculate key epidemiological indicators like prevalence and incidence using Real World Data (RWD), researchers should carefully determine the lookback period (LP) to ensure well-characterized patient health status at inclusion. This study aims to establish appropriate LPs for chronic pathologies (Asthma, Diabetes, HTA, COPD, Stroke, Parkinson, Depression), neoplasia: (Breast, lung, prostate, colorectal cancers) and acute pathologies (Urinary tract infection), across 7 European countries (Belgium, France, Germany, Italy, Romania, Spain and UK).

METHODS: Observational retrospective multinational study using RWD. Patients included are registered in THIN®, have at least one visit to a general practitioner during 2023 and at least 24 months of medical data between the first record in the database and the relevant year of study. Adequate LPs identifies locally stable incidence rates as the potential relationship between registration and record of the diagnosis is correctly accounted for. To assess the best LP for each pathology in each country, a visual inspection of the graph of the incidence curve as a function of the LP is carried out to identify relevant plateau signaling stabilization.

RESULTS: 7,73 M active patients were included for all studied countries. For France, Italy and UK the best LP interval with stable incidence rate (in weeks) are different according pathologies and countries as: Diabetes (FR: 12-24, IT: 12-24, UK: 24-36), Breast cancer (FR: 24-36, IT:24-36, UK: 60-72), UTI (FR: 12-24, IT: 12-24, UK: 36-48).

CONCLUSIONS: These results show that to develop a standardized methodology across countries for epidemiological indicators using RWD requires a deep knowledge of LPs. These depend on the national health system - registration based systems (UK) or GP-based (France)- and differs across pathologies.