A scoring index based on six readily documented parameters can identify postmenopausal women with osteoporosis and back pain who are likely to have a vertebral fracture and therefore should undergo spine radiography, investigators in France report in the
Annals of the Rheumatic Diseases for January.
Dr. Christian Roux and colleagues note that vertebral fractures occur in roughly 20% of postmenopausal women, and that such fractures put women at risk of future fractures and early mortality. Nevertheless, "vertebral fractures are underdiagnosed," they write.
Dr. Roux, a rheumatologist at Hopital Cochin in Paris, and associates analyzed clinical data from 397 women 65 to 85 years of age who had gone to a doctor because of back pain, defined as thoracic or lumbar pain with scores of 40 mm or higher on a 100-mm visual analog scale. They had all been diagnosed with osteoporosis on the basis of bone mineral density testing.
All the women underwent spinal radiography of vertebrae from T4 to L5, which identified 540 vertebral fractures in 215 (52.4%) of the patients.
Using univariate logistic regression model, the authors determined which predictors had a 20% level of significance, all of which were entered into a multivariate logistic model used to develop a quantitative index of risk.
Their final index included six parameters - age, intensity of back pain, height loss since age 25, history of low-trauma non-vertebral fracture, sudden occurrence of pain, and thoracic localization of pain.
Scoring was as follows: age (65-69 years, 1 point; 70-74 years, 2 points; 75-79, 3 points; 80 or higher, 4 points); pain intensity 65 mm or more (2 points); height loss (3 to 6 cm, 2 points, > 6 cm, 4 points); history of low-trauma non-vertebral fracture (1 point); sudden occurrence of pain (4 points); thoracic localization of pain (1 point).
The maximal score was 16. Dr. Roux and colleagues report that for a score of 7 or higher, the probability of an existing vertebral fracture was at least 43%. The positive predictive value was 70.9% and negative predictive value was 68.6%.
This scoring index "can be useful in helping doctors make decisions about the need for spinal radiography, in their search for treatment," the investigators conclude.
Ann Rheum Dis 2007;66:81-85.
