Rehabilitation-need and population-projection dataset for Türkiye (GBD 2021 × TÜİK 2023–2100)

A compiled, openly reusable dataset describing the population that would benefit from rehabilitation in Türkiye by age, and its demographic projection to 2100. It brings together primary extracts from the Global Burden of Disease Study 2021 (via the WHO Rehabilitation Need Estimator) and official Turkish Statistical Institute (TÜİK) population projections 2023–2100, plus two derived data products and a self-contained derivation script. It is provided independently of any particular analysis. Contents. raw/: age-specific rehabilitation-need prevalence and counts by condition group (GBD 2021, with 95% uncertainty intervals); the de-duplicated all-ages total, 1990–2021; and TÜİK population projections (age band × year × scenario), as raw workbooks and a tidy CSV; plus provenance notes for two context variables (OECD physiotherapist density; older-adult internet use). derived/: age-specific de-duplicated rehabilitation-need prevalence (2021); and the projected number of people who would benefit from rehabilitation — total and among adults 65+ — by year (2023–2100) and TÜİK scenario. A data dictionary and a Python derivation script (derive_dataset.py) that rebuilds derived/ from raw/ are included. Sources. Global Burden of Disease Study 2021 and WHO Rehabilitation Need Estimator (IHME); method Cieza et al., Lancet 2020 (doi:10.1016/S0140-6736(20)32340-0); data GBD 2021 Collaborators, Lancet 2024 (doi:10.1016/S0140-6736(24)00757-8). Türkiye İstatistik Kurumu (TurkStat/TÜİK), Population Projections 2023–2100 (Bulletin 53699). Context variables: OECD Health Statistics (physiotherapists); Eurostat/TÜİK ICT-usage survey (older-adult internet use). Scope note. "Rehabilitation need" is the broad epidemiological prevalence of people living with a health condition that would benefit from rehabilitation (the WHO/GBD construct), not a clinical caseload or a demand estimate. Projections hold age-specific prevalence constant and apply demographic change only; within-age morbidity trends are not modelled, so far-horizon (2075–2100) figures are more uncertain than near-term ones. The de-duplication uses a conditional-independence approximation anchored to the GBD all-ages de-duplicated total.