TRANSPORTABILITY-INFORMED CLINICAL AND ECONOMIC EVALUATION OF TIRZEPATIDE FOR LOCAL HTA DECISION-MAKING
Author(s)
Nathaniel Dyrkton, MSc1, Shomoita Alam, PhD2, Rasha Abdul-Karim, MD3, Jay J. Park, PhD4.
1Core Clinical Sciences Inc, Vancouver, BC, Canada, 2Core Clinical Sciences Inc, Montreal, QC, Canada, 3Department of Medicine, Western University, London, ON, Canada, 4Core Clinical Sciences, Vancouver, BC, Canada.
1Core Clinical Sciences Inc, Vancouver, BC, Canada, 2Core Clinical Sciences Inc, Montreal, QC, Canada, 3Department of Medicine, Western University, London, ON, Canada, 4Core Clinical Sciences, Vancouver, BC, Canada.
OBJECTIVES: Randomized clinical trials (RCTs) are central to health technology assessment (HTA), yet their applicability to local decision-making may be limited when trial populations do not represent the target jurisdiction. Pivotal tirzepatide trials demonstrated substantial weight-loss efficacy but lacked Canadian sites, raising uncertainty about external validity. This study aimed to estimate the clinical short-term cost-effectiveness of tirzepatide in Canada by transporting trial evidence to the population.
METHODS: We conducted Bayesian transportability analysis using patient-level data from the SURPASS-3 trial and the Canadian Community Health Survey (CCHS), a nationally representative survey. Trial eligibility criteria were emulated in the CCHS to define the target population. Transportability was assessed by adjusting trial-based treatment effects for differences in pre-specified effect modifiers (baseline bodyweight, age, sex, and race) between the trial and Canadian populations using a Bayesian g-computation with survey-weighted re-sampling. The transported estimand was the mean difference in bodyweight change (kg) at 52 weeks for tirzepatide (15 mg) versus insulin degludec. Transported treatment effects were estimated nationally and by province/territory. National transported estimates were incorporated into a short-term, payer-perspective cost-effectiveness analysis using a decision-tree model and compared with trial-based efficacy.
RESULTS: In the SURPASS-3 trial, tirzepatide (15 mg) was associated with a mean weight-loss difference of −13.1 kg (95% credible interval CrI: [−14.2, −12.0]) compared with insulin degludec. After the transport to the Canadian population, the effect was attenuated to −12.2 kg (95% CrI: [−13.3, −10.9]), with similar benefits across provinces/territories. While incremental costs were comparable, the reduced incremental benefit under the transported scenario resulted in a higher incremental cost-effectiveness ratio ($599.40 per kilogram lost) compared with the trial scenario ($577.40 per kilogram lost).
CONCLUSIONS: Our study demonstrates the feasibility of generating locally relevant evidence using a transportability analysis. Transported effect estimates can then be used throughout value assessment to generate cost-effectiveness estimates that better reflect local contexts.
METHODS: We conducted Bayesian transportability analysis using patient-level data from the SURPASS-3 trial and the Canadian Community Health Survey (CCHS), a nationally representative survey. Trial eligibility criteria were emulated in the CCHS to define the target population. Transportability was assessed by adjusting trial-based treatment effects for differences in pre-specified effect modifiers (baseline bodyweight, age, sex, and race) between the trial and Canadian populations using a Bayesian g-computation with survey-weighted re-sampling. The transported estimand was the mean difference in bodyweight change (kg) at 52 weeks for tirzepatide (15 mg) versus insulin degludec. Transported treatment effects were estimated nationally and by province/territory. National transported estimates were incorporated into a short-term, payer-perspective cost-effectiveness analysis using a decision-tree model and compared with trial-based efficacy.
RESULTS: In the SURPASS-3 trial, tirzepatide (15 mg) was associated with a mean weight-loss difference of −13.1 kg (95% credible interval CrI: [−14.2, −12.0]) compared with insulin degludec. After the transport to the Canadian population, the effect was attenuated to −12.2 kg (95% CrI: [−13.3, −10.9]), with similar benefits across provinces/territories. While incremental costs were comparable, the reduced incremental benefit under the transported scenario resulted in a higher incremental cost-effectiveness ratio ($599.40 per kilogram lost) compared with the trial scenario ($577.40 per kilogram lost).
CONCLUSIONS: Our study demonstrates the feasibility of generating locally relevant evidence using a transportability analysis. Transported effect estimates can then be used throughout value assessment to generate cost-effectiveness estimates that better reflect local contexts.
Conference/Value in Health Info
2026-05, ISPOR 2026, Philadelphia, PA, USA
Value in Health, Volume 29, Issue S6
Code
P21
Topic
Study Approaches
Topic Subcategory
Decision Modeling & Simulation
Disease
SDC: Diabetes/Endocrine/Metabolic Disorders (including obesity)