Discrepancy of left and right hip bone mineral density (BMD) in Thai women: diagnostic agreement and misclassification.

OBJECTIVE: To determine the diagnostic agreement and the degree of misclassification when using data from the left and right hips. METHODS: The cross-sectional study of 1,943 perimenopausal and postmenopausal Thai women, who had bone mineral density (BMD) measurements at the left (non-dominant) and right hips for the screening of low bone mass (LBM) or osteoporosis (OP) in the Department of Radiology, Faculty of Medicine, Chiang Mai University from September 2008 to August 2010 was performed. The kappa statistic was used to assess diagnostic agreement. The prevalence of LBM and OP and the percentage of misclassification were reported. RESULTS: There was a significant correlation between the left and right BMD values for the femoral neck (FN) (r (2) = 0.83; p < 0.001) and the total hip (TH) (r (2) = 0.89; p < 0.001). The diagnostic agreement of the FN and TH regions was significant in all study groups ranging from 0.69 to 0.76 (p < 0.001). For the final diagnosis, which is based on the least T-score of the FN or TH regions, the diagnostic agreement was 0.73 for all women, 0.77 for perimenopausal women, 0.73 for postmenopausal women, 0.70 for postmenopausal women age less than 65 years and 0.71 for postmenopausal women age greater than or equal to 65 years. The percentage of misclassification for all women was 16.9 %, with 3.3 % being downgraded from normal to LBM and 3.4 % from LBM to OP. CONCLUSION: Despite the fact that good diagnostic agreement was demonstrated in this study, a significant number of diagnostic discordance between left and right hips (16.9 %) was also observed. BMD measurements of both hips are recommended for diagnosing LBM and OP in clinical practice.

Importance of ethnic base standard references for the diagnosis of osteoporosis in Thai women.

Many studies demonstrated the importance of using ethnic-specific normal database in the diagnosis of osteoporosis (OP). Aims of this study were to assess diagnostic agreement, prevalence of OP, and diagnostic misclassification between Caucasian, Japanese, and Thai normal databases. The cross-sectional study of 3181 Thai women who had bone mineral density (BMD) measurement between January 2008 and December 2010 was performed. BMDs at lumbar spine (LS), femoral neck (FN), and total hip (TH) were derived to T-score by using Caucasian, Japanese, and Thai standard references. Kappa statistic was used to assess diagnostic agreement and misclassification. Diagnostic agreements between Caucasian and Thai reference databases were 0.39 for LS and 0.90 for FN. No statistical agreement was found in TH region (0.01, p value=0.264). Applying the Japanese reference, diagnostic agreements were 0.71 for LS, 0.76 for FN, and 0.94 for TH regions. Prevalence of OP in postmenopausal women was 64.1%, 37.7%, and 41.4% using Caucasian, Japanese, and Thai standard references. Percentage of misclassification was varied by menopausal status and reference database from 11.2% to 48.7%. When applying Japanese databases instead of Caucasian normal databases, overall diagnostic misclassification decreased from 35.1% to 16.1%. Choice of reference database has a significant effect on the diagnosis of low bone mass and OP. Japanese reference database has better diagnostic agreement with previously studied Thai reference database in 1999 than Caucasian reference database.