THU0470 Hospital Utilisation and Mortality of Traumatic Hip and Non-Hip Fractures in Patients with Osteoporosis in Western Australia: A Ten-Year Study Using Linked Administrative Data

Background Osteoporosis is common. Traumatic fracture is more likely in those with osteoporosis at significant cost. Given the burden of the osteoporotic fracture and need to prioritise fracture prevention and management initiatives in WA, we sought to further investigate the 10-year linked data from the previous study to examine hospital utilisation and mortality associated with traumatic fractures in those with osteoporosis. 1 Objectives To compare hospital utilisation and mortality outcomes of “traumatic” hip with non-hip fractures in patients with osteoporosis in Western Australia (WA) over 10 years. Methods Design: Retrospective cohort study (2002 to 2011). Setting: Population-based data linkage of the WA Hospital Morbidity Data System and Mortality Registry. Participants: WA residents aged ≥50 years hospitalised with an index “traumatic” fracture and osteoporosis. Index fractures sites included hip, spine, pelvic, rib or “other” fracture sites. Main outcome measures: Index hospitalisations, hospital readmission due to any subsequent fractures, length of stay (LOS), admission costs, characteristics of hospital stay and mortality. Results Specific fracture sites for index cases were identified in 974 patients (18% of all index cases) due to coding inconsistencies. Index hip and spine fractures were the most common fractures requiring hospitalisations, with index hip being the most expensive to treat ($40,570 per person on average) and the cost accounted for 65% of total cost of index cases. Non-hip fractures accounted for 68% of admissions and have similar hospital utilisation as hip fracture in terms of LOS (30 days on average), requirement for hospital transportation (over 60%) and rate of readmission. The risk of readmission following index spine fracture was 1.5-fold higher (95%CI=1.14–1.98) than index hip fractures. No significant mortality difference was found between hospitalised hip and non-hip cases. The 1 year, 5 years and 10 years cumulative probability of death for hip fracture was 21%, 61% and 84%, respectively. Conclusions Fracture prevention strategies need to target all with osteoporosis (marker of frailty) to reduce incidence of first fracture (including traumatic). Although mortality and readmission were similar for hospitalised hip and non- hip fracture patients, hip fracture accounted for 32% of index fractures and 65% of costs. Better coding system of osteoporotic fracture in the hospital data in Australia is required to monitor the disease progression in the aging population. References Briggs AM, Sun W, Miller LJ, Geelhoed E, Huska A, Inderjeeth CA. Hospitalisations, admission costs and re-fracture risk related to osteoporosis in Western Australia are substantial: a 10-year review. Aust N Z J Public Health 2015. Disclosure of Interest None declared