Physician Differences in Managing Postmenopausal Osteoporosis: Results from the POSSIBLE US™ Treatment Registry Study

Background: Osteoporosis is a disease that often goes undetected until a fracture occurs. Previous reports indicate that disease diagnosis and care of patients with osteoporosis may vary within the medical community. Objective: Using data from the POSSIBLE US™ registry (October 2004–December 2009), we evaluated patterns of care for a group of primary care (i.e. first-contact) physicians who frequently prescribe osteoporosis medications to determine whether variations existed in the characteristics of their postmenopausal patients; physician approaches to diagnosis; treatment choices and monitoring; and patient-reported medication use. Methods: POSSIBLE US™ was a large prospective registry of postmenopausal women receiving osteoporosis treatment. We analysed data from 42 family practice physicians (FPPs), 50 internal medicine specialists (IMs) [internists, physicians], 41 gynaecologists (GYNs) and the 4917 patients they enrolled in the POSSIBLE US™ registry between October 2004 and January 2007. Women who had been postmenopausal for at least 1 year and who were newly initiating osteoporosis therapy, switching or augmenting therapy or continuing on a stable therapy regimen were investigated. Therapies included bisphosphonates, full-length or peptide derivative of parathyroid hormone, calcitonin, oral or transdermal postmenopausal estrogen, selective estrogen receptor modulators (SERMs), calcium and/or vitamin D supplements (alone or in combination with other therapies), or any combination of these agents. Data on physician characteristics were collected on an initial qualification questionnaire. Physicians reported data for enrolled patients at study entry and were also asked to provide relevant data obtained at clinic visits throughout the follow-up period. Patient-reported data were collected using questionnaires mailed out semi-annually throughout the follow-up period. Patient-reported and physician-reported data were assessed using ANOVA models and chi-squared (χ 2 ) or Cochran-Mantel-Haenszel tests to evaluate differences across physician types. Multivariate logistic regression models examined the odds of patients having an osteoporosis diagnosis, being prescribed specific agents and receiving an additional dual energy x-ray absorptiometry (DXA) scan after the initial diagnostic scan. Cox proportional hazards regression models were used to determine whether the risk of patient-reported treatment discontinuation during 12 months of follow-up differed by