Claim: A Bayesian decision rule that jointly models imminent fracture benefit and cardiovascular harm will outperform simple boxed-warning style exclusion when selecting romosozumab for very-high-risk osteoporosis.

Rationale: Current practice often collapses a multidimensional decision into yes-no screening. That ignores fracture urgency, antiresorptive failure, and heterogeneity of cardiovascular baseline risk. A net-benefit framework should better identify patients in whom anabolic benefit still dominates and those in whom even short-term exposure is unjustified.

Prediction: Compared with a binary rule based mainly on recent myocardial infarction or stroke plus clinician gestalt, a Bayesian competing-risk model using recent fracture count, T-score, antiresorptive failure, ASCVD, heart failure, CKD, hypertension, smoking, and age will yield higher clinical net benefit on decision-curve analysis and better calibration across external sites.

Suggested study: target-trial emulation with transportable hierarchical priors across health systems; primary endpoint is 12-month composite net benefit defined from prevented major osteoporotic fractures minus weighted MACE harm; validation by calibration slope, Brier score, and decision-curve dominance.

Why it matters: This would turn osteoporosis prescribing from rule-based caution into quantitatively auditable shared decision-making.

References: Cosman F et al. N Engl J Med. 2016. DOI: 10.1056/NEJMoa1607948; Saag KG et al. N Engl J Med. 2017. DOI: 10.1056/NEJMoa1708322; Vickers AJ, Elkin EB. Med Decis Making. 2006. DOI: 10.1177/0272989X06295361