Date: Monday, November 6, 2017
Session Type: ACR Poster Session B
Session Time: 9:00AM-11:00AM
Background/Purpose: Osteoporosis-related fractures causes increased disability and morbidity and mortality. The FRAX algorithm quantifies the 10 year probability of a hip or major osteoporotic fracture. This study was initiated after observation of significant discrepancies between FRAX scores generated by the facility machine and those calculated using the online FRAX tool. We aimed to review a series of bone density scans to determine the reason for the discrepancy. Our hypothesis was that providers may not be using the correct measure to calculate the FRAX score.
Methods: This was a retrospective quality improvement study. 700 of 1200 DEXAs from 2013 to 2015 were randomly selected, 168 met the inclusion criteria of osteopenia. Data to calculate the score was obtained from the patient chart. The correct way to calculate the FRAX is to use the femoral neck bone mineral density (BMD) and to select the DEXA machine used. Selecting the femoral neck t-score gives the wrong DEXA result because the normative t-score values used by hologic differ from the FRAX. Selecting BMD without specifying the machine does not give an accurate score as it defaults to no bone mineral density result. Three separate FRAX scores were calculated: (1) FRAX by femoral neck T-score (T score FRAX), (2) FRAX by femoral neck BMD (No BMD FRAX), and (3) FRAX by BMD with hologic machine selected (BMD FRAX). The latter score was the gold standard. It was determined whether the patient was ultimately appropriately treated by chart review.
Results: Patients were primarily female (94%), age ranged from 40-89 years old, and race was primarily hispanic (54.7%) and black (31%). There was a difference when comparing BMD FRAX with the machine generated FRAX (p< .05), and between the BMD FRAX with T-Score BMD (p=0.001). However, these differences were not clinically significant. There were clinically and statistically significant differences when comparing treatment of BMD FRAX with no BMD FRAX (P < .001), all of whom would be over treated if using the latter. Four patients (2.4%) were over treated by FRAX guidelines, however, justification was given for each case. Four patients (2.4%) were undertreated. In 3 of these patients, no FRAX score was mentioned in patient notes and in 1 patient FRAX score was interpreted incorrectly.
Conclusion: Not selecting DEXA machine type can lead to significant overtreatment of patients. Many patients were undertreated due to lack of understanding of FRAX score by clinicians. Further education of clinicians regarding FRAX score use is needed. Modification of the FRAX score calculation may be required to ensure patients are not being over or under treated.
To cite this abstract in AMA style:Vig A, Johnson B, Francis T, Mendez-Agrusa B. Factors Affecting FRAX Score Calculation & Treatment in Practice [abstract]. Arthritis Rheumatol. 2017; 69 (suppl 10). https://acrabstracts.org/abstract/factors-affecting-frax-score-calculation-treatment-in-practice/. Accessed September 24, 2021.
« Back to 2017 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/factors-affecting-frax-score-calculation-treatment-in-practice/