A revised method to detect low bone mineral density in plain x-rays of the hip
Background Osteoporosis is associated with hip fractures. The gold standard to assess osteoporosis is to measure bone mineral density (BMD) by dual-energy x-ray absorptiometri, which is quick and reliable but not available in most orthopaedic departments. A fast and feasible way to assess bone quality is using plain x-rays but previous attempts to correlate plain x-rays to osteoporosis are inconclusive. Aim of Study Evaluate the reliability of the Cortex-Marrow Ratio (CMR) measure and precision for CMR as a screening test for osteoporosis. Materials and Methods From a previous study we extracted all consecutive patients with femoral neck fractures referred for BMD measurements, n=133 patients (mean 78.1 years, 100 female). On preoperative digital AP hip x-rays two observers measured the outer cortex diameter just below the minor trochanter (TM) on a 20 inch screen or larger with maximal possible zoom. Two diameters distal to TM the outer and inner cortex distance were measured. The outer cortex divided by the inner cortex is defined as the CMR. One observer measured twice, 3 months apart. Intra- and interrater reliability was assessed with Intra Class Correlation (ICC) and smallest detectable change (SDC). Osteoporosis was defined as a Total-hip T-score below -2.5, which 53.5 % of the patients had. Results Mean CMR 1.81 (1.76;1.87), range 1.19-2.52. Intrarater ICC 0.97 (0.96;0.98) and interrater ICC 0.84 (0.78;0.88). SDC was 0.16 and 0.10. No indication of systematic bias (Bland-Altman Plot). With a cutpoint at CMR≤1.5 specificity is 95.5 % for having osteoporosis (16.5 % of the patients), at CMR≥2.05 specificity is 95.5 % for not having osteoporosis (20.3 % of the patients). Conclusions CMR is a reliable measure and osteoporosis status is settled for about 37 % of patients of having either osteoporosis or not. Further exploration of CMR is needed.