Variation in Osteoporosis Treatment Gaps Across Fracture-Related and Elective Orthopedic Surgeries: A Single-Center Retrospective Cohort Study in Taiwan

Background/Objectives: Fracture liaison services (FLSs) typically use fragility fractures as triggers to identify patients for osteoporosis care. Patients undergoing elective orthopedic surgery who already have osteoporosis documented on a preoperative densitometry report fall outside this trigger. This study compared recorded initiation of osteoporosis treatment after elective operations with that after operations performed for a fracture. A secondary, exploratory objective was to describe how preoperative T-score relates to subsequent fractures in patients without recorded treatment. Methods: A retrospective cohort study (2015–2025) was performed at a tertiary center in Taiwan. The source population was the institutional densitometry cohort (all patients with at least one DXA examination at the hospital), within which five index operations were identified: vertebroplasty/kyphoplasty, hip fracture internal fixation, hip arthroplasty (hemi- or total), lumbar spinal fusion, and total knee arthroplasty. Operations in patients already receiving osteoporosis therapy were excluded. The primary outcome was recorded initiation of anti-osteoporosis therapy, defined as the first dispensing at the study hospital within 365 days of surgery. Odds ratios were estimated by logistic regression with patient-clustered standard errors. Results: Among 6171 operations in 5739 patients, recorded one-year initiation was 28.4% overall: 46.9% after vertebroplasty, 43.1% after hip fracture fixation, 33.0% after hip arthroplasty, 17.4% after lumbar fusion, and 7.2% after total knee arthroplasty. Adjusted odds ratios versus fusion ranged from 3.16 (95% CI: 2.69–3.72) for vertebroplasty to 0.28 (0.21–0.37) for knee arthroplasty. Within hip arthroplasty, initiation was 43.9% after hemiarthroplasty, which is performed for femoral neck fracture, and 6.7% after total hip arthroplasty. Across all operations, an accompanying fracture diagnosis was associated with initiation (adjusted OR: 6.19, 5.26–7.29). The contrast persisted among the 871 operations with osteoporosis documented before surgery (48.2% after vertebroplasty, 7.5% after knee arthroplasty). In the exploratory analysis of 1971 operations without recorded treatment (99 fractures), each one-unit lower T-score was associated with a higher hazard of major osteoporotic fracture (adjusted HR: 1.25, 1.04–1.49). The estimate for the T ≤ −2.5 dichotomy was less precise (HR: 1.33, 0.83–2.12). Conclusions: In this selected single-center densitometry cohort, preliminary findings indicate that recorded initiation of osteoporosis treatment after orthopedic surgery followed whether the operation was performed for a fracture rather than the patient’s documented bone density, including among patients whose osteoporosis was already documented. A national reimbursement rule that requires a fracture in addition to low bone density probably contributes. Treatment outside the study hospital was not captured, so initiation proportions are lower bounds. Whether an FLS entry point keyed to existing densitometry results would close this gap remains to be tested.