Comparison between real-world practice and application of the FRAX algorithm in the treatment of osteoporosis

Abstract

Background and aims: The most recent guidelines suggest treating patients whose FRAX 10-year fracture risk scores are ≥ 20%. However, this method of evaluation does not take into account parameters that are nonetheless relevant to the therapeutic choice. Our aim was to compare the therapeutic choices for treatment based on a wider assessment (real-world practice) with those based on FRAX scores, taking 20% as the cut-off score.

Methods: We obtained the medical history, bone mineral density (BMD) values, and the presence of major fragility fractures in a sample of 856 postmenopausal women. The 10-year FRAX risk of major osteoporotic fracture was calculated, and patients were grouped into risk classes ("FRAX < 20%" = low, "FRAX ≥ 20%" = high); we then compared the treated and untreated patients in each class. After an average interval of 2.5 years, changes in lumbar and femoral BMD and appearances of new fragility fractures were recorded.

Results: 83% of high-risk patients and 57% of low-risk patients were treated. The therapeutic decision was based mainly on densitometric values and the presence of vertebral fractures. At the 2.5 year follow-up, lumbar spine and femur BMD had decreased in the untreated group; 9.9% of the treated patients developed new vertebral fragility fractures, compared with 5.3% of the untreated patients.

Discussion and conclusions: Our wider assessment designated as at high fracture risk a group of patients who had not been identified by the FRAX assessment. FRAX could underestimate the risk of fracture in older people, for which the therapeutic choice should consider a broader approach, also based on individual patient's needs.

Keywords: FRAX; Multidimensional approach; Older persons; Osteoporosis.

Fig. 1

CHAID decision tree describing the decision-making process to determine outpatient therapy; UT untreated, T treated