Better bone density reporting: T-score report versus fracture risk report with outcome analysis

To determine whether different classification systems for bone fracture risk result in different diagnoses and treatments.

Assessment of 2 diagnostic classification systems by physicians who routinely read bone density scans, using a survey administered by an audience response system.

At a professional meeting, 103 rheumatologists were asked to read bone density results from 9 case studies in 2 formats: a standard format using T-scores and a format using absolute fracture risk. As physicians were shown each of the 18 scans, they diagnosed each patient with (1) normal bone, (2) osteopenia, (3) osteoporosis, or (4) unknown, and provided a treatment recommendation. Test scans included hip dual-energy x-ray absorptiometry (DEXA) (60%), spine DEXA (10%), phalangeal DEXA (10%), and calcaneal ultrasound (20%).

Approximately 1850 anonymous responses were recorded. For women, diagnostic agreement between physicians was 17% for the T-score report and 89% for the absolute risk report (P < .001). For men, diagnostic agreement also was better with the absolute risk report (66% vs 10%; P < .001). With DEXA, overall diagnostic agreement improved from 30% to 68% with the absolute risk report (P < .001). With ultrasound, diagnostic agreement improved from 4% to 63% with the absolute risk report (P < .001). Bisphosphonates were more likely to be prescribed in cases of increasing age and risk if scans were reported in terms of absolute fracture risk.

A direct, risk-based bone density assessment may improve consistency and efficiency of diagnosis and treatment of osteoporosis.