Care-pathway discontinuity after WHO ICOPE screening: a three-year implementation evaluation of 11,827 assessments in Taiwan

Abstract Background The World Health Organization Integrated Care for Older People (ICOPE) approach is intended to connect screening for declines in intrinsic capacity with person-centred assessment and care planning. Evidence on whether older people who screen positive reach the next pathway step remains limited. We evaluated downstream-assessment completion, the usefulness of baseline screening profiles for identifying non-completion, and the stability of programme measurements over time. Methods We conducted a repeated cross-sectional evaluation of 11,827 routinely collected initial ICOPE assessments among adults aged 65 years and older in a hospital-based programme in Taiwan during 2023–2025. Initial and downstream-assessment records were linked within each calendar year. Among records with at least one abnormal core domain, modified Poisson regression with robust variance estimated adjusted risk ratios (aRRs) for non-completion. Logistic models assessed discrimination and year-by-domain interaction. Domain classifications were independently reconstructed from raw screening items to evaluate measurement stability. Results Among 9,517 records with an abnormal screen, 7,174 (75.4%) had no linked downstream assessment; annual non-completion was 76.7%, 72.8%, and 76.6%. Baseline factors showed only small associations with non-completion. Cognitive abnormality was associated with an aRR of 1.04 (95% CI 1.02–1.07; absolute risks 76.3% versus 73.5%), but this association varied by year (interaction P = 0.020). Discrimination was poor (c-statistic 0.550 for the full model versus 0.523 for year alone), providing little support for profile-targeted navigation. Reconstruction reproduced all six domain flags exactly but identified a change in the vision algorithm in 2024 and a marked shift in recorded chair-rise times despite an unchanged threshold. Conclusions In routine ICOPE delivery, screening coverage alone did not indicate completion of the intended care pathway: approximately three quarters of positive-screen records lacked a linked downstream assessment. Because baseline intrinsic-capacity profiles did not identify non-completion adequately, programmes should evaluate systematic follow-up for all positive screens and report downstream-assessment completion as an implementation-quality indicator. Longitudinal comparisons also require versioned algorithms and stable measurement procedures.