Andrea Schwarz Doria MD, Abstract Co-Author: Nothing to Disclose

Bjorn Lundin MD, Abstract Co-Author: Nothing to Disclose

Stephen Franklin Miller MD, Abstract Co-Author: Nothing to Disclose

Sylvia Thomas, Abstract Co-Author: Nothing to Disclose

Amy Dunn, Abstract Co-Author: Nothing to Disclose

Paul S. Babyn MD, Presenter: Nothing to Disclose

This study complements a series of retrospective studies conducted by the Expert MRI Working Group of the International Prophylaxis Study Group which aimed at testing measurement properties of the Compatible MRI scale for evaluation of hemophilic joints. The purpose of this study was to assess the reliability and construct validity of the MRI scale for evaluation of elbows, comparing the diagnostic performance of MRI and plain film radiographs.

29 MR examinations of elbows from 28 boys with hemophilia A and B and age range of 5-17 (mean, 11.5) years were independently reviewed by 4 blinded radiologists on 2 occasions. Variations of 3-plane SE T1-weighted, TSE T2-weighted, fat-saturated 3D gradient-echo and STIR MRI protocols were obtained at 1.5 T in 3 centres (Toronto,Canada, n=24 examinations; Atlanta,US, n=3; Cuiaba,Brazil, n=2). Number of previous joint bleeds and laboratory level of severity of hemophilia were the reference standard measures for assessment of discriminant validity of MRI and radiography.

Inter-reader reliability was substantial (ICC=0.73; 95% confidence intervals [CI], 0.64-0.81) for the A-scale and excellent (ICC=0.83; 95% CI, 0.77-0.88) for the P-scale. Intra-reader reliability was excellent for both P-(ICC=0.91; 95% CI, 0.87-0.94) and A-(ICC=0.93; 95% CI, 0.89-0.95) scores. Convergent validity: Final P-and A-scores correlated positively but poorly (r=0.36, P=0.06) or moderately (r=0.54, P=0.004) with clinical-laboratory measurements. Discriminant validity: Total P- and A-MRI scores and radiographic scores demonstrated excellent accuracy for discrimination of severe disease (P-scale, 0.83±0.09; A-scale, 0.87±0.09, P=0.12), but non-diagnostic accuracy for discrimination of mild disease.

Both MRI scales demonstrated excellent reliability, poor to moderate convergent validity for final scores, and excellent accuracy for discrimination of severe arthropathy, and did not perform better than radiography for discrimination of mild or severe arthropathy.

Standardizing and testing a scale for MRI measurements is paramount for comparison of results of clinical trials conducted in different countries while assessing hemophilic joint damage.

Doria, A, Lundin, B, Miller, S, Thomas, S, Dunn, A, Babyn, P, Measurement Properties of the Compatible MRI Scoring System for Evaluation of Hemophilic Elbows in Children.  Radiological Society of North America 2007 Scientific Assembly and Annual Meeting, November 25 - November 30, 2007 ,Chicago IL. http://archive.rsna.org/2007/5015474.html