Calcaneal ultrasound bone densitometry is not a useful tool to screen patients with inflammatory bowel disease at high risk for metabolic bone disease

Background: Up to 42% of patients with inflammatory bowel disease (IBD) have significant metabolic bone disease. The current method of screening for osteopenia or osteoporosis involves dual-energy x-ray absorptiometry (DXA). This is relatively costly and involves radiation exposure. What is needed is a safe, inexpensive, and quick screening tool to identify patients who would benefit from DXA testing. This would reduce the number of patients undergoing DXA testing unnecessarily. We tried to determine if calcaneal ultrasound bone densitometry is a useful tool in screening high-risk patients with IBD for metabolic bone disease. Methods: Patients with IBD who presented to the clinic between August 29, 2003 and December 22, 2003 were enrolled in this prospective study. All patients underwent calcaneal ultrasound bone densitometry screening using a GE Lunar Achilles Insight quantitative ultrasound densitometry machine (QUS). Patients who were at high risk for significant metabolic bone disease (i.e., significant previous prednisone use or a long history of severe IBD) or who had a T-score on QUS less than or equal to -0.7 had DXA testing performed. The DXA results and QUS results were compared. The radiologist was blinded to the results of QUS. Results: One hundred twenty-four patients with IBD were enrolled. Fifty (40%) were considered high risk for metabolic bone disease. This cohort was comprised of 29 men (58%), of which 21 (73%) had Crohn's disease (CD). Eighty percent of this high-risk group had CD, and in both groups, the majority had used corticosteroids. The overall risk of significant metabolic bone disease in this high-risk group was 62% (DXA ≤ -1.0). Heel density (T-score) correlated poorly with DXA (T-score) at either hip or spine at 0.40 even when 2 outlier patients (QUS = -2.9, DXA spine = 0.7, DXA hip = 0.8 and QUS = -3.6, DXA spine = -3, DXA hip = -4) were excluded. Likewise, no association in osteopenia or osteoporosis was seen between multiple variables. These included sex, disease type (ulcerative colitis or CD), smoking, and prior intestinal resection. The sensitivity of QUS to identify patients with significant metabolic bone disease was 74%, and specificity was 63%. A positive predictive value of 81% and negative predictive value of 53% were also less than ideal. The Altman-Bland analysis showed that the agreement between QUS and DXA was poor (-2.0, 2.1). Based on this analysis, QUS cannot replace DXA in the individual patient with IBD. Conclusions: Calcaneal ultrasound bone densitometry is not a useful tool to screen high-risk patients with IBD for metabolic bone disease.