Country-year agreement between GBD 2023 ulcerative colitis and Crohn’s disease rates and GIVES population-based observations: an ecological agreement study

<title>Abstract</title> <p>Background GBD 2023 reports ulcerative colitis (UC) and Crohn’s disease (CD) separately, but the country-year interchangeability of modelled rates with population-based observations is uncertain. We quantified absolute agreement between GBD 2023 and matched observations from the Global Inflammatory Bowel Disease Visualization of Epidemiology Studies in the 21st Century (GIVES-21). Methods We conducted an ecological agreement study using national, all-age, both-sex crude incidence and prevalence rates. GIVES observations were exactly matched to GBD 2023 by disease, outcome, country and year, then collapsed within disease-country-year-outcome units. The primary estimand was log[GBD rate/GIVES rate]. Agreement was summarized with Bland-Altman differences, proportional-bias regression, Deming regression and Lin concordance correlation coefficients (CCC). Confidence intervals were estimated from 5,000 country-block bootstrap samples. Sensitivity analyses addressed stricter population-based eligibility, equal-country weighting, publication-level source overlap, multiyear observations and age-standardized rates. Results The matched dataset contained 1,255 disease-specific rows from 66 studies and 30 countries, yielding 784 collapsed units; 777 had positive finite rates. High-income settings contributed 69.0% of valid units, and the ten largest contributing countries accounted for 78.2%. Median GBD-to-GIVES rate ratios were 0.66 (95% CI 0.57–0.74) for UC incidence, 0.56 (0.45–0.66) for CD incidence, 0.62 (0.42–0.85) for UC prevalence and 0.53 (0.38–0.81) for CD prevalence. Corresponding CCCs were 0.854 (0.629–0.955), 0.852 (0.697–0.943), 0.795 (0.591–0.927) and 0.833 (0.688–0.930). Proportional-bias slopes were compatible with zero for incidence but negative for UC prevalence (-0.181, 95% CI -0.352 to -0.011) and CD prevalence (-0.246, -0.373 to -0.119). Key restrictions preserved the predominantly negative point-estimate differences, although precision and concordance varied. Conclusions GBD 2023 and GIVES rates showed positive cross-unit concordance but insufficient country-year agreement for interchangeability, particularly for prevalence. The findings compare two imperfect evidence sources; they do not identify the true rate or justify a uniform correction factor.</p>