Markov Model

Death was treated as the absorbing state in the Markov model. The model began at age 40 and followed people until age 100 or death. The study used a decision tree for initial assessment and DXA referral linked to a Markov cohort model for lifetime outcomes. The model included nine health states covering osteoporosis, osteopenia, normal bone density, fracture, post-fracture, and death. The model's first part used trial-arm, testing, test-result, de-labelling, and treatment-failure states to model 12-month costs and QALYs. The model extrapolated later costs using regression equations based on post-12-month participant data and label and treatment-failure status. The economic model extrapolated 12-month trial results to a five-year time horizon. The model was developed in R using the open-source DARTH framework and used the English NHS perspective. Unit costs were based on 2022 US Medicare national average payment rates inflated to 2025 values. The model started with 1000 individuals and used annual transition probabilities from literature and data sources.