Abstract
BACKGROUND: A substantial portion of women diagnosed with osteoporosis (OP) do not initiate pharmacotherapy to reduce fracture risk. In clinical practice, gastrointestinal (GI) events have been linked with OP therapy discontinuation. However, there is limited research examining GI events as barrier to treatment initiation following an OP diagnosis. The objective of this study was to examine the association between gastrointestinal (GI) events and osteoporosis (OP) treatment initiation among post-menopausal women diagnosed with osteoporosis in France. METHODS: A retrospective claims analysis of the Mediplus France database during 1997 to 2010 identified women aged ≥ 55 with an OP diagnosis and without prior OP treatment (first diagnosis date was defined as the index date). GI events were identified during the 1 year pre-index and up to 1 year post-index. OP treatment initiation post-index was identified based on the presence of claims for any bisphosphonate (BIS) or non-BIS OP medication within 1 year post-index. Multivariate models (logistic regression, Cox proportional hazards regression and discrete choice) adjusted for pre-index patient characteristics were used to assess the association of pre- and post-index GI events with the likelihood of initiating OP treatment, and the type of treatment initiated (BIS vs. non-BIS). RESULTS: A total of 10,292 women (mean age 70.3 years) were identified; only 25% initiated OP treatment. Post-index GI events occurred in 11.5% of patients, and were associated with a 75.7% lower likelihood of initiating OP treatment. Among treated patients, a discrete choice model estimated that patients with post-index GI events were 34.6% less likely to receive BIS vs non-BIS as compared to patients without post-index GI events. CONCLUSION: Among women aged ≥ 55 years with an OP diagnosis, post-index GI events were associated with a lower likelihood of OP treatment initiation.