Timely, AQL-Driven Clinical Cohort Identification in openEHR Infrastructures

Timely cohort identification remains a challenge in hospitals due to fragmented systems, delayed coding, and limited IT capacity. We present CohortMailer, a lightweight tool that automates extraction of patient cohorts from openEHR repositories using Archetype Query Language (AQL) and delivers filtered case lists via scheduled email. Deployed across two sites of the University Hospital Schleswig-Holstein (UKSH), the system executed structured queries, applied rule-based filters, and distributed tabular summaries to study teams. During a 30-day evaluation, CohortMailer achieved 83.3% availability, identifying 274 unique patients. In the respiratory cohort, it reached a positive predictive value (PPV) of 62.9% and sensitivity of 50.3%. The results demonstrate how modular, standards-based automation can reduce manual screening effort and support real-time clinical workflows.