Multisource Coherence Analysis of the First European Multicenter Cohort Study for Cancer Prevention in People...

Area of Science:

• Public Health
• Health Services Research
• Data Science

Background:

• Ensuring data coherence across multicenter datasets is crucial for reliable health data reuse and to prevent biases.
• Heterogeneity in data can lead to suboptimal generalization of research findings.
• The study addresses data quality for a novel European multicenter dataset on cancer prevention and early detection among the homeless population.

Purpose of the Study:

• To characterize and label data coherence across sites in a European multicenter dataset for homeless individuals.
• To enable research addressing health disparities and healthcare access barriers faced by people experiencing homelessness.
• To adapt and implement a health navigator model within this population.

Main Methods:

• Analysis of a dataset comprising 652 cases from Austria, Greece, Spain, and the United Kingdom.
• Longitudinal data collection via questionnaires at baseline, 4 weeks, and end of intervention.
• Assessment of data variability using information theory and geometric methods to analyze distribution and completeness.

Main Results:

• Substantial variability observed across the 4 pilot countries in overall analysis and specific domains.
• Greatest discrepancies noted in health care empowerment, quality of life, and interpersonal communication.
• Spain showed the most pronounced differences, with high missing values in interpersonal communication and healthcare service use.

Conclusions:

• Health data may be comparable, but significant differences in other questionnaire domains necessitate country-specific analyses.
• The study highlights the heterogeneity among people experiencing homelessness.
• Emphasizes the critical need for data quality assessments to guide future research and policy in this area.