Gary S. Rudolph

OUTCOME OF PATIENTS WHO PRESENT TO THE EMERGENCY DEPARTMENT WITH AN ELEVATED INTERNATIONAL NORMALIZED RATIO

To the Editor: We have recently established an Osteoporosis Clinic at our federally qualified health center and have now obtained bone mineral density (BMD) measurements using dual-energy x-ray absorptiometry (DXA) in 114 postmenopausal women and others at risk. For these patients, we have measurements of the lumbar spine and, where obtainable, both femurs. From these initial readings, we have noted that many of our patients have a discrepancy between left and right total hip BMD that is greater than precision error of the system can explain. Of the 76 patients who had both hips measured, 18 had T-score discrepancies of at least 0.5, a difference large enough in some cases to determine whether we would initiate pharmacological therapy. Given that weight-bearing exercise is known to influence bone density, we wondered whether one explanation for the large discrepancies might be gait asymmetry. In all five patients with a T-score discrepancy of 0.7 or greater, we have confirmed that the patient has an easily identifiable gait asymmetry (Table 1). Discrepancies between the right and left total hip T-scores in the same patient have been noted in some studies. Recommendations for how to use this information differ. Some feel that the percentage of patients with changes in diagnostic categories is so small that routine monitoring of bilateral femur BMD does not yield clinically significant benefit. Nevertheless, there has been the suggestion that from a public health perspective bilateral measurements may be warranted, because the absolute number of patients affected is large. Current recommendations from the International Society for Clinical Densitometry are that treatment decisions be based on the BMD measurement from one hip or, if both hips are measured, that the mean of the two measurements be used. One factor worth considering at this point is that the additional time needed to obtain a second femur reading is minimal, in contrast to the amount of time that was needed with first-generation scanners. Given that there is a potential physiological explanation for a difference in T-scores in the hips of a single individual, we suggest that, where possible, both hips be measured and that treatment decisions be based on the lower of the T-scores rather than on the mean T-score. If bilateral measurements are not feasible in all patients, we recommend that, at least in patients with obvious gait asymmetry, both hips be measured.