VALIDATION OF THE CORE OBESITY AND DIABETES MODEL AGAINST A CHRONIC KIDNEY DISEASE MODEL AND OBSERVATIONAL EVIDENCE IN CHRONIC KIDNEY DISEASE
Author(s)
Simone Parisotto, PhD1, Runguo Wu, PhD2, Suramya Shukla, MSc3, Chiara Fattore, MSc1, Guillem Laborda, BSc4, Anamaria-Vera Olivieri, MSc5, Francesca Fiorentino, PhD1, Anke van Engen, MSc6.
1IQVIA, Milan, Italy, 2IQVIA, London, United Kingdom, 3IQVIA, Gurugram, India, 4IQVIA, Barcelona, Spain, 5IQVIA, Basel, Switzerland, 6IQVIA, Amsterdam, Netherlands.
1IQVIA, Milan, Italy, 2IQVIA, London, United Kingdom, 3IQVIA, Gurugram, India, 4IQVIA, Barcelona, Spain, 5IQVIA, Basel, Switzerland, 6IQVIA, Amsterdam, Netherlands.
OBJECTIVES: The IQVIA Core Obesity and Diabetes Model (CODM) is a microsimulation model of interconnected cardiometabolic diseases including chronic kidney disease (CKD), cardiovascular disease (CVD) and diabetes. We assessed external validity of CODM against a CKD model (CKD-PM) and observational studies.
METHODS: We informed CODM with Chronic Renal Insufficiency Cohort (CRIC) Phase-I baseline characteristics to simulate 5‑year cumulative incidences (CIs) of all-cause mortality (ACM), end-stage kidney disease/kidney replacement therapy (ESKD/KRT), CVD and heart failure. Predictions were compared with CKD-PM (Ramos et al. 2024) and observed CRIC outcomes (Grams et al. 2021) across baseline eGFR categories (G1-2, G3a, G3b and G4-5), with concordance assessed using R², Ordinary Least Squares - Linear Regression Line (OLS-LRL) and Root Mean Squared Error (RMSE). Using CRIC and EMPA-KIDNEY characteristics, CODM-predicted undiscounted life expectancy (LE) was compared with CKD-PM and life table-based estimates (Turin et al. 2012) adjusted for over-time survival improvement using USRDS longitudinal data.
RESULTS: Compared with CKD-PM, CODM more closely matched observed CRIC 5-year CIs, with higher R² (0.955 vs 0.917), OLS-LRL slope closer to unity (1.05 vs 1.08) and lower RMSE (5.34 vs 7.64) across outcomes. For patients starting in eGFR G1-2 to G4-5, ACM was 6.2%, 13.1%, 21.4% and 35.9% respectively (CODM), 4%, 11%, 14% and 21% (CRIC), and 4%, 8%, 21% and 48% (CKD-PM); ESKD/KRT was 0.0%, 2.6%, 15.8% and 71.1% (CODM), 0%, 4%, 13% and 70% (CRIC), and 0%, 0%, 16% and 72% (CKD-PM). Predicted LE in CRIC was 26.9, 18.7, 13.4 and 9.7 years (CODM), 27.2, 18.1, 12.3 and 9.1 years (adjusted life-table based estimates), and 20, 15, 11 and 9 (CKD-PM). Similarly, for EMPA-KIDNEY cohort the predicted LE were: 10.9 (CODM) vs 11.4 (adjusted life-table based estimates) vs 11 (CKD-PM) years.
CONCLUSIONS: CODM predictions showed strong agreement with external evidence from CRIC and EMPA-KIDNEY cohorts and with CKD-PM predictions.
METHODS: We informed CODM with Chronic Renal Insufficiency Cohort (CRIC) Phase-I baseline characteristics to simulate 5‑year cumulative incidences (CIs) of all-cause mortality (ACM), end-stage kidney disease/kidney replacement therapy (ESKD/KRT), CVD and heart failure. Predictions were compared with CKD-PM (Ramos et al. 2024) and observed CRIC outcomes (Grams et al. 2021) across baseline eGFR categories (G1-2, G3a, G3b and G4-5), with concordance assessed using R², Ordinary Least Squares - Linear Regression Line (OLS-LRL) and Root Mean Squared Error (RMSE). Using CRIC and EMPA-KIDNEY characteristics, CODM-predicted undiscounted life expectancy (LE) was compared with CKD-PM and life table-based estimates (Turin et al. 2012) adjusted for over-time survival improvement using USRDS longitudinal data.
RESULTS: Compared with CKD-PM, CODM more closely matched observed CRIC 5-year CIs, with higher R² (0.955 vs 0.917), OLS-LRL slope closer to unity (1.05 vs 1.08) and lower RMSE (5.34 vs 7.64) across outcomes. For patients starting in eGFR G1-2 to G4-5, ACM was 6.2%, 13.1%, 21.4% and 35.9% respectively (CODM), 4%, 11%, 14% and 21% (CRIC), and 4%, 8%, 21% and 48% (CKD-PM); ESKD/KRT was 0.0%, 2.6%, 15.8% and 71.1% (CODM), 0%, 4%, 13% and 70% (CRIC), and 0%, 0%, 16% and 72% (CKD-PM). Predicted LE in CRIC was 26.9, 18.7, 13.4 and 9.7 years (CODM), 27.2, 18.1, 12.3 and 9.1 years (adjusted life-table based estimates), and 20, 15, 11 and 9 (CKD-PM). Similarly, for EMPA-KIDNEY cohort the predicted LE were: 10.9 (CODM) vs 11.4 (adjusted life-table based estimates) vs 11 (CKD-PM) years.
CONCLUSIONS: CODM predictions showed strong agreement with external evidence from CRIC and EMPA-KIDNEY cohorts and with CKD-PM predictions.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
MSR272
Topic
Clinical Outcomes, Methodological & Statistical Research, Study Approaches
Disease
Cardiovascular Disorders (including MI, Stroke, Circulatory), Diabetes/Endocrine/Metabolic Disorders (including obesity), Urinary/Kidney Disorders