Modeling Framework for Assessing Cost-Effectiveness of New Treatments in Chronic Kidney Disease
Author(s)
Junwen Zhou, PhD, Claire Williams, PhD, Borislava Mihaylova, DPhil.
Health Economics Research Centre, University of Oxford, Oxford, United Kingdom.
Health Economics Research Centre, University of Oxford, Oxford, United Kingdom.
OBJECTIVES: We propose a new model to assess the cost-effectiveness of new treatments to reduce kidney disease progression and cardiovascular risks in patients with chronic kidney disease (CKD) at different CKD stages, with and without diabetes, albuminuria or other comorbidities.
METHODS: We used a large UK primary healthcare population data, the Clinical Practice Research Datalink (CPRD), linked with hospital care and mortality data, to derive an open 2005-2020 cohort of adults with CKD. Data on individuals’ sociodemographic and clinical characteristics and outcomes (stroke, myocardial infarction [MI], hospitalisation for heart failure [HHF], acute kidney injury [AKI]; incident diabetes; kidney function; and vascular [VD] and nonvascular deaths [NVD]) during follow-up were extracted. We conceptualized a microsimulation model reflecting the relationship between outcomes and contributing socio-demographic and clinical characteristics through literature reviews and discussion with clinical experts. The cohort was split into estimation (east/central England 66.7%) and validation (west England: 33.3%) cohorts. Cox proportional hazards risk equations with time-updated outcomes were estimated and validated.
RESULTS: The cohort included 1.7 million people with CKD [at entry: mean age 71, 42% men, 30% with prior diabetes, 50% with prior CVD, CKD stage (CKD1: 8.6%; CKD2: 23.7%; CKD3a: 40.7%; CKD3b: 13.1%; CKD4: 2.6%; CKD5:0.3%; on kidney replacement therapy (KRT): 0.9%; unknown: 10.1%)) with a mean follow-up of 6 years. More advanced CKD was associated with higher risk of cardiovascular outcomes [e.g. hazard ratio (95%CI) of CKD4 compared with CKD3a: MI 1.58 (1.53-1.63); stroke 1.27 (1.27-1.30); HHF 2.15 (2.09-2.21); VD 1.79 (1.74-1.83)], kidney outcomes [AKI 4.64 (4.50-4.78); KRT 12.6 (12.0-13.3)], and NVD [1.76 (1.73-1.79)] but not incident diabetes [0.87 (0.84-0.91)]. The risk equations performed well in internal and external validations.
CONCLUSIONS: This new CKD model, further integrated with evidence of quality of life and healthcare costs, can inform assessments of cost-effectiveness of treatments modifying CKD progression and cardiovascular risks in CKD.
METHODS: We used a large UK primary healthcare population data, the Clinical Practice Research Datalink (CPRD), linked with hospital care and mortality data, to derive an open 2005-2020 cohort of adults with CKD. Data on individuals’ sociodemographic and clinical characteristics and outcomes (stroke, myocardial infarction [MI], hospitalisation for heart failure [HHF], acute kidney injury [AKI]; incident diabetes; kidney function; and vascular [VD] and nonvascular deaths [NVD]) during follow-up were extracted. We conceptualized a microsimulation model reflecting the relationship between outcomes and contributing socio-demographic and clinical characteristics through literature reviews and discussion with clinical experts. The cohort was split into estimation (east/central England 66.7%) and validation (west England: 33.3%) cohorts. Cox proportional hazards risk equations with time-updated outcomes were estimated and validated.
RESULTS: The cohort included 1.7 million people with CKD [at entry: mean age 71, 42% men, 30% with prior diabetes, 50% with prior CVD, CKD stage (CKD1: 8.6%; CKD2: 23.7%; CKD3a: 40.7%; CKD3b: 13.1%; CKD4: 2.6%; CKD5:0.3%; on kidney replacement therapy (KRT): 0.9%; unknown: 10.1%)) with a mean follow-up of 6 years. More advanced CKD was associated with higher risk of cardiovascular outcomes [e.g. hazard ratio (95%CI) of CKD4 compared with CKD3a: MI 1.58 (1.53-1.63); stroke 1.27 (1.27-1.30); HHF 2.15 (2.09-2.21); VD 1.79 (1.74-1.83)], kidney outcomes [AKI 4.64 (4.50-4.78); KRT 12.6 (12.0-13.3)], and NVD [1.76 (1.73-1.79)] but not incident diabetes [0.87 (0.84-0.91)]. The risk equations performed well in internal and external validations.
CONCLUSIONS: This new CKD model, further integrated with evidence of quality of life and healthcare costs, can inform assessments of cost-effectiveness of treatments modifying CKD progression and cardiovascular risks in CKD.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
EE588
Topic
Economic Evaluation, Methodological & Statistical Research, Study Approaches
Disease
Cardiovascular Disorders (including MI, Stroke, Circulatory), Diabetes/Endocrine/Metabolic Disorders (including obesity), Urinary/Kidney Disorders