Bone Abstracts

Objectives: To examine the association of gastrointestinal (GI) events and osteoporosis (OP) therapy initiation patterns among postmenopausal women following a diagnosis of OP from a large health plan in Israel.

Methods: Women aged ≥55 years were included in the analysis if they had ≥1 OP diagnosis (ICD-9 733.0X) (date for first OP diagnosis was index date), no estrogen use, no diagnosis of Paget’s disease or malignant neoplasm. OP treatment initiation was defined as use of OP therapy: bisphosphonates (BIS) (alendronate, ibandronate, risedronate, and zolendronic acid) and non-BIS (raloxefene, calcitonin, and teriparatide), during 12 months post-index. GI events (diagnosis of GI conditions) were reported for 12-month pre-index and post-index (from index to treatment initiation or end of 1-year post-index whichever occurred first). The association of post-index GI events (yes/no) and initiation of OP treatment (yes/no), and the type of therapy initiated (i.e. BIS vs non-BIS) were examined with logistic regression and also Cox proportional hazard regression as sensitivity analysis.

Results: Among 30 788 eligible patients, aged 65.0±7.6 (mean±S.D.) years, 17.5% had pre-index GI event, and 13.0% had post-index GI event. 70.6% of patients didn’t receive OP therapy in the year following OP diagnosis, 25.1% received BIS and 4.2% received non-BIS. The logistic regression showed that post-index GI events were associated with a lower odds of OP medication initiation by ~85% (P<0.001), and upon treatment initiation, post-index GI was not significantly associated with type of therapy initiated (BIS vs. non-BIS), controlling for baseline GI and patient characteristics.

Conclusions: Among newly diagnosed osteoporotic women from a large health plan in Israel, 70.6% did not receive pharmacological OP treatment within 1 year of OP diagnosis. Patients with post-index GI events were about 85% less likely to initiate OP treatment.