HERALD (health economics using routine anonymised linked data)

Area of Science:

• Health Economics
• Data Science
• Clinical Informatics

Background:

• Traditional health economic analysis uses isolated datasets, limiting comprehensive patient journey insights.
• Integrating diverse data sources offers a more holistic approach to understanding patient care pathways.

Purpose of the Study:

• To outline the implications of linking patient-reported data with routine healthcare datasets for health economic analysis.
• To demonstrate the utility of a unified data resource for longitudinal studies.

Main Methods:

• Linked individual-level data from patient questionnaires with routinely collected General Practitioner (GP), inpatient (IP), and outpatient (OP) data.
• Utilized data from an Ankylosing Spondylitis (AS) cohort in Wales to illustrate retrospective and prospective patient pathway mapping.
• Employed data linkage to track patient journeys, validate self-reported data, and assess healthcare resource utilization.

Main Results:

• Successfully linked and analyzed data for Ankylosing Spondylitis (AS) patients, tracking GP events, diagnostic codes, outpatient visits, and inpatient admissions.
• Presented patient stratification based on disease severity, function, age, sex, and symptom onset.
• Demonstrated the ability to obtain objective cost measures over extended time horizons, including pre-diagnosis periods.

Conclusions:

• Linked data systems enable enhanced longitudinal health economic analysis beyond the capabilities of isolated datasets.
• Data linkage offers valuable insights for optimizing diagnostic and referral pathways, improving clinical and resource efficiency.