Age‑dependent assessment thresholds to optimize patient care in a resource‑limited setting: an analysis based on the Sri Lankan FRAX model

Abstract

Summary Age-dependent upper and lower assessment thresholds help categorizing women aged 40 years or more according to their fracture risk, independent of BMD information. Introduction Age-dependent assessment thresholds of the FRAX algorithm help stratifying men and women aged 40 years or more according to their fracture risk. This allows clinicians to decide on those who require interventions without BMD assessment and those who require BMD input for further assessment. Methods Intervention thresholds were defned by 10-year probabilities of a major osteoporotic fracture (MOF) and hip fracture (HF) considering a woman with a BMI of 25.0 kg/m2 having a prior fragility fracture but no other clinical risk fac tors. The lower assessment thresholds (LAT) were set at 0.8 times the 10-year probabilities of a MOF and HF in a woman with a BMI of 25.0 kg/m2 , without previous fracture or other clinical risk factors. The upper assessment thresholds (UAT) were set at 1.2 times the intervention thresholds of MOF and HF. Fracture probabilities were estimated for the age range of 40–80 years, without BMD input. These values were applied to a group of women who underwent DXA for clinical reasons in a single center. Results The LATs of MOF and HF varied from 0.7 to 8.8% and 0.1 to 3.7%, from 40 to 80 years, respectively. The cor responding values for UATs were 2.5 to 21.6% and 0.3 to 8.4%. ITs of MOF and HF varied from 2.1 to 18% and 0.2 to 7%, respectively. When applied to a group of 315 postmenopausal women who underwent DXA for clinical indications, 22.9% of women were above the UATs (high-risk category) while 8.6% were below the LATs (low-risk category). The proportion of women in the intermediate category who require BMD for further assessment was 68.6% (95% CI 59.7 to 77.5%). Conclusions In nearly one-third of women aged 40 years or more, the decision to treat or not to treat can be achieved without BMD estimation. The remaining two-thirds will require a BMD assessment for further evaluation.