This article is more than 5 years old. Information may no longer be current.
Simple fracture risk prediction tests as useful as FRAX
The Fracture Risk Assessment Tool failed to predict risk for fracture any better than simpler tools, results from a 3-year prospective study show.
“In this study, we found that Fracture Risk Assessment Tool (FRAX) and simpler screening tools, such as OST, ORAI, OSIRIS, SCORE and even age alone, performed similarly in predicting fractures in a screening scenario without [bone mineral density] assessment,” Katrine Hass Rubin, PhD, of the Institute of Clinical Research at the University of Southern Denmark and department of medical endocrinology at Odense University Hospital, and colleagues wrote.
The population-based study included 5,000 randomly selected women aged 40 to 90 years (BMI, 26 kg/m2; mean age, 64 years) living in southern Denmark. Questionnaires regarding risk factors for osteoporosis were administered to the patients; self-reported baseline data were used to measure 10-year likelihood of fracture by FRAX and to estimate the risk based on other tools. Age was also considered as a continuous variable, according to researchers.
Katrine Hass Rubin
Rubin and colleagues identified incident fracture outcomes at follow-up, including “major osteoporotic fractures” (ie, FRAX-defined major osteoporotic fracture and hip, clinical vertebral, wrist or humerus fracture), in addition to all other fractures, excluding those of the fingers, toes, skull or face.
After 3 years, 3,860 questionnaires were completed (10,385 person-years), with 225 (6%) of women reporting an osteoporotic fracture. According to data, patients with incident fractures tended to be older (73 years vs. 63 years; P=.0001), had more frequent history of fractures (22% vs. 9%; P<.001) and falls during the past 12 months (14% vs. 6%; P<.001), had diseases related to secondary osteoporosis (26% vs. 18%, P=.011), and with less frequent estrogen use (3% vs. 11%; P=.001).
Furthermore, age restrictions in the analysis did not yield significant differences in area-under-the-curve values between the various screening tools, according to researchers.
“These findings suggest that simpler tools based on fewer risk factors, which would be easier to use in clinical practice by the general practitioner or the patient herself, could just as well as FRAX be used to identify women with increased risk of fracture and therefore should be referred to a DXA scan,” researchers wrote.
Disclosure: The researchers report no relevant financial disclosures.
Perspective
Back to Top
It is important to have tools to assess the risk of fracture in individual patients so that pharmacologic therapy can be targeted appropriately. As fracture risk increases, the benefit of treatment is enhanced and the risk-benefit profile is optimized. The most important determinants of fracture risk are age, bone mineral density and prior fracture history. Because of the very strong correlation between body weight and BMD, weight or BMI can be used as a surrogate when BMD is not available. Other risk factors, including smoking, excessive alcohol intake, parental hip history of fracture and glucocorticoid use modify the risk of fracture, in complex ways, in individuals who have these risk factors. The FRAX tool is a very sophisticated, scientifically sound computer-based risk assessment tool incorporating all of these risk fractures, adjusting for the complex interplay among the various risk factors. The output of FRAX is the 10-year likelihood of either major osteoporotic fractures or hip fracture. In the USA, race is also considered. Age-specific mortality is taken into account since, especially among elderly patients; the risk of dying reduces the risk of having a fracture.
The study by Rubin and colleagues compared the effectiveness of the FRAX model without BMD with age alone and several more simple tools to identify patients who developed a fragility fracture over a 3-year follow-up interval in a community-based cohort of women in southern Denmark. The non-FRAX tools had been developed and validated to identify patients with low bone density and had not been previously validated as predictors of fracture risk. The authors correctly conclude that the simple models were as good or nearly as good as the more complicated FRAX calculation in predicting fracture risk in their patient population. At first glance, these results suggest that the FRAX tool is needlessly complex. However, the women evaluated in this study were quite healthy and had few risk factors for fracture. Their average age was 64 years. Only 9% had a history of prior fracture, 4% had a history of glucocorticoid use, and 2% reported excessive use of alcohol, explaining why their average 10 year probability of a major osteoporotic fracture by FRAX was only 6%. Very few were elderly, and the short follow-up time of 3 years blunted the impact of including mortality in FRAX. For these relatively healthy women, the risk factors that drive the FRAX score are predominantly and body size, the same risk factors that form the basis of the “more simple” tools. The “complexities” of FRAX are most useful in elderly patients who have other risk factors for fracture. In addition, FRAX quantifies the risk whereas the “simple” tools segregate patients into categories of risk.
In my opinion, the results described by Rubin are not expected and do not undermine the utility or importance of FRAX. The FRAX algorithm is the most scientifically validated tool to predict fracture risk and is the only tool whose results have been incorporated into treatment guidelines by the National Osteoporosis Foundation, The Endocrine Society, AACE and other professional organizations. The importance of FRAX, or of any tool that accurately predicts fracture risk, is to allow us to make treatment decisions based on knowledge of fracture risk and to move away from the incorrect strategy of simply treating bone density values.