DEXA overestimates bone mineral density in adults with knee replacements

We examined bone mineral density (BMD) measurements made by dual-energy-xray-absorptiometry (DEXA) taken from 100 patients (♂46/♀54, 66±6yr) who previously underwent single total-knee arthroplasty (TKA) to determine if automated software-based artifact detection (ASAD) adequately removes implant artifact from the DXA image before analysis and if potential inaccuracies could be overcome through manual artifact correction (MAC). We also sought to determine if software-based inaccuracies would result in fracture risk misclassification (Low-BMD/Osteopenia = Young-Adult T-Score < -1). Select Results: When using ASAD, limbs with implants had higher BMD (+12.0 ​± ​1.7%, p ​< ​0.001) compared to control limbs resulting in a 2.5 ​± ​0.2% overestimation of total-body BMD (single implant). Consequently, the prevalence of osteopenia in 95% of patients who would have been observed to have low leg BMD (18/19 patients) and 80% of those found to have low total-body BMD (4/5 patients) would have gone un-diagnosed. This overestimation was eliminated when using MAC. These results reveal a potential issue with ASAD for total-body DEXA scans in TKA patients and highlight the importance of careful review and MAC in those with joint replacements before making diagnostic decisions.