study published in JAMA Internal Medicine.according to a
One of the screenings, the U.S. Fracture Risk Assessment Tool (FRAX), proved relatively ineffective at identifying women who developed osteoporosis. The other screening, the Osteoporosis Self-Assessment Tool (OST), excelled at identifying osteoporosis for women in every racial and ethnic group, but also failed at identifying who was most likely to experience a fracture. Osteoporosis experts say that primary care physicians should test for the condition in anyone with any risk factor for it, even if a screening tool suggests doing so is unnecessary.
The United States Preventive Services Task Force (USPSTF) recommends routine testing of bone mineral density in women age 65 years and older to detect risk of developing osteoporosis, which in turn leads to an increased risk for fractures of the hip, spine, shoulder, or forearm. For women aged 50-64, whether bone mineral density accurately reflects who will develop osteoporosis is less clear. In this age range, the USPSTF recommends using either FRAX or OST rather than routine bone mineral density tests.
“I have the utmost respect for the United States Preventive Services Task Force, which lists both of these as valid screening tools for younger postmenopausal women. What I hope this study does is to inform the next iteration of the screening guidelines,” by maintaining the recommendation to use the OST while not keeping FRAX, said Carolyn J. Crandall, MD, MS, an internal medicine physician and health services researcher at University of California, Los Angeles, who helped conduct the research.
The U.S. version of FRAX requires identifying someone’s race, height, and weight, then answering whether they have different risk factors for a fracture such as a previous fracture, rheumatoid arthritis, or smoking. The result was thought to indicate a cumulative risk for major fracture over the next 10 years. Patients at significant risk should then undergo a bone density test.
The tool can also incorporate information about bone mineral density, if available, but the FRAX analyses in Dr. Crandall’s study did not include those data because the study aimed to test the measure’s predictive ability in the absence of a bone scan.
The OST includes only two variables – weight and age – to calculate risk for osteoporosis, and generally takes seconds to complete. It does not include race. As with FRAX, anyone deemed at significant risk for developing osteoporosis should undergo a bone density test.
“OST is really simple; that makes it very appealing,” Dr. Crandall said. “OST could probably be automatically calculated in the electronic medical record.”
Using data from the Women’s Health Initiative, Dr. Crandall and colleagues tracked more than 67,000 women aged 50-64 years for 10 years following enrollment in the study to see who experienced a fracture or developed osteoporosis over that decade. The investigators found that neither FRAX nor OST was particularly good at predicting who went on to experience a fracture.
The accuracy of FRAX at fracture prediction peaked at 65% for Asian women (area under the receiver operating curve, 0.65; 95% confidence interval, 0.58-0.71), and was lowest for Black women (AUC 0.55; 95% CI, 0.52-0.59). OST also was most accurate for Asian women, but only up to 62% (AUC 0.62; 95% CI, 0.56-0.69), and was again lowest for Black women (AUC 0.53; 95% CI, 0.50 - 0.57)
“It is just very hard to predict fractures in this age group,” Dr. Crandall said, noting that more evidence exists about risk for fracture in people older than 65.
The story diverges with predicting risk of osteoporosis in the neck. The OST did this roughly 80% of the time, for all racial groups. That figure proved better than FRAX, without including race.