Typology of Patients With Lung Cancer Based on Demographics and Exposure to Environmental Factors in France: A Real-World Study
Speaker(s)
Youinou M1, Rosé M2, Ricci JF3, Petrica N4
1Alira Health, Paris, France, 2Alira Health, Barcelona, B, Spain, 3Alira Health, Basel, BS, Switzerland, 4Alira Health, Paris, 75, France
Presentation Documents
OBJECTIVES: The study aims to identify profiles of patients with lung cancer (LC) and the association to the exposure to different environmental and socio-economic factors and to describe the impact on medical burden in patients with LC.
METHODS: A retrospective analysis of the French national hospital claims databases (PMSI) merged with environmental and socio-economic open-source data from 2017 to 2022 was conducted. Adult patients hospitalized for LC in metropolitan France were included. Merging of the different data sources was done at the patient level using postal codes of the patient or nearest neighbor methods when postal code was unavailable. French open data on environmental, socio-economic, and demographics factors were utilized to create patient typologies. Cluster-based analyses using the k-means method were employed to identify specific profiles of patients with LC. The project received approval from the French regulatory authorities.
RESULTS: 277,287 patients were identified with a mean age of 67 years (±10.5) and 34.6% female. Environmental factors used in this study are exposition to greenhouse gas emission, radon, atmospheric pollution (concentration of PM10, PM2.5 and NO2 in the air). The social deprivation index and the “Hospital Frailty Risk Score” were used as socio-economics factors. Altitude, urbanization, and agricultural/industry specialization were used as geographical factors, while age, gender, and patient comorbidities were selected for demographics. The estimated patient profiles will be mapped by classes of risks factors, patient care, treatment, and access to healthcare.
CONCLUSIONS: This is the first study to describe the socio-environmental factors and medical burden associated with LC using claims data and open-source real-world data in France. The research holds significant public health interest, particularly in assessing the impact of environmental factors for the outcomes of patients with LC.
Code
SA70
Topic
Epidemiology & Public Health, Methodological & Statistical Research, Study Approaches
Topic Subcategory
Artificial Intelligence, Machine Learning, Predictive Analytics, Public Health
Disease
Oncology, Respiratory-Related Disorders (Allergy, Asthma, Smoking, Other Respiratory)