Background/Purpose: Fracture Liaison Services systematically identify patients with fragility fracture and screen patients for osteoporosis and associated conditions and instigate evidence-based ‘secondary’ fracture prevention management. The success of such a service is thus demonstrated by its effect on reducing secondary fractures, particularly hip fractures, which are most important in terms of causing morbidity and their cost. There is a need for Fracture Liaison Services to demonstrate a direct and specific effect on fracture prevention.

Methods: A Fracture Liaison Service was implemented at The Ipswich Hospital in 2003. Expected hip fracture incidence 2004-13 for the hospital was estimated by applying age and sex specific hip fracture rates, derived from UK Hospital Episode Statistics (2007-8) to changes in the demographic profile of the local population (2001 and 2011 population census) and applied to the 2009 estimates of the Hospital catchment population (Public Health England). Hip fracture (codes ICD10 S72.0-3) rate ratios were then calculated comparing observed (local derived) versus expected hip fractures. SDs and CIs for the rate ratios were estimated using the Poisson function. Results were compared with data derived similarly from Norwich, an area with no Fracture Liaison Service and Cambridge, where a Hospital-based service has been in place since 2006.

Results: Over a 10y period 2004-13: annual hip fracture incidence varied 355-511 (Ipswich), 684-798 (Norwich) and 448-706 (Cambridge). There was no change in hip fracture incidence 2004-13 (versus expected) in Norwich, where there had been no service activity (rate ratio mean 0.97 [CIs 0.9-1.04]). For the hospitals where a service had been active, non-parametric trend analysis showed a significant reduction in hip fractures 2004-13 (all ages >55y) in Cambridge (p<0.05) and at Ipswich (service from 2003) mean hip fracture rate ratio decreased 15% (1.1[CI ±0.26] to 0.93[CI ±0.22]; NS) for subsequent 5y periods (2004-8 and 2009-13) at ages <75y (though no significant trend for hip fracture reduction for all ages >55y).

Conclusion: Overall, these analyses, based on low event numbers, fairly wide fracture rate ratio CIs (±0.1) and the need to derive accurate population demographic and expected hip fracture data, illustrate the challenges in linking changes in hip fracture incidence to local Fracture Liaison Service activity. There is, however, some evidence to suggest hip fracture reduction in areas where a Fracture Liaison Service had been active, but no evidence of a change in hip fractures (versus expected) even in the non-elderly, in an area with no historic service activity.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/incidence-of-fragility-hip-fractures-over-10-years-2004-13-in-three-uk-centres-with-reference-to-local-fracture-prevention-fracture-liaison-service-activity/