Send to

Choose Destination
Osteoporos Int. 2005 Aug;16(8):898-906. Epub 2004 Nov 27.

A triage strategy based on clinical risk factors for selecting elderly women for treatment or bone densitometry: the EPIDOS prospective study.

Author information

INSERM (Institut National de la Santé et de la Recherche Médicale), Unité de recherche U149 (Recherches épidémiologiques en santé périnatale et santé des femmes), 94807 Villejuif Cedex, France.


A triage strategy, based on a clinical hip fracture risk score, may be used to classify elderly women into three groups: one at high risk and requiring treatment, another needing further assessment by bone densitometry, and a third at low risk. We used prospective data from the EPIDOS study (7512 women older than 75 years and followed for an average of 3.9 years) to assess the potential value of such a strategy for identifying elderly women with a hip fracture risk twice the cohort average (i.e. > or =20 per 1000 woman-years). An individual fracture risk score was calculated with the final risk function (Cox model). To compare this strategy with systematic BMD measurement and with current European recommendations, we examined the number of high-risk women identified, their average risk levels, sensitivity for hip fracture, and the number of high-risk women who need to be treated to prevent one hip fracture (hypotheses: all identified women are treated; sensitivity is equal to the point estimate; treatment reduces fracture risk by 35%). A triage strategy based on age, fracture history since the age of 40 years, body mass index, number of instrumental activities of daily living for which assistance is needed, grip strength, and visual acuity can identify 20% of the cohort as at high risk, 75% of them from clinical factors only, and the rest after BMD measurements (threshold: -2.5 T-score). The triage strategy would be significantly more sensitive than systematic BMD screening (51 versus 35%) and would require many fewer BMD examinations (10%). Compared with current recommendations, triage would identify fewer women (20 versus 28%) but at a significantly higher average risk of hip fracture (30 versus 20 per 1000 woman-years). Fewer high-risk women would be treated to prevent one hip fracture (29 versus 41) and fewer bone densitometry tests would be needed (10% versus 54%). The proposed triage strategy may be a useful clinical tool for selecting elderly women for treatment or bone densitometry.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Springer
Loading ...
Support Center