Stress Fractures: Diagnosis with MRI, Bone Scintigraphy, and Radiography

Kiuru MJ,^1,2 Pihlajamäki HK,³ Ahovuo JA²

1. Research Institute of Military Medicine, Helsinki, Finland

2. Department of Radiology, Helsinki University Central Hospital, Helsinki, Finland

3. Department of Surgery, Military Central Hospital, Helsinki, Finland

Purpose: To compare MR imaging (MRI), bone scintigraphy, and radiography in the diagnosis of stress fractures of the pelvis and lower extremities.

Materials and Methods: A total of 50 consecutive military recruits clinically suspected of having a stress fracture underwent MR imaging and scintigraphy. Of these patients, 43 were studied also with plain film radiography, and follow-up radiographs were taken of 24 patients. A total of 41 stress fractures were diagnosed in 32 patients using MR imaging.

Results: Of the patients, 36% had a false-positive clinical diagnosis of stress fracture. The sensitivity of radiography was 56%, specificity = 94%, accuracy = 67%, positive predictive value (PPV) = 95%, and negative predictive value (NPV) = 48%. In the follow-up, the sensitivity increased to 72%, NPV to 60%, accuracy to 79%, and PPV to 96%, and specificity remained the same. The kappa value for MR imaging and radiography was fair (0.39) initially, and moderate (0.57) in the follow-up. The sensitivity of bone scintigraphy was 92%, specificity = 90%, accuracy = 92%, PPV = 95%, and NPV = 86%. The kappa value for bone scintigraphy and MR imaging was very good: 0.81. Three false-negative and two false-positive scintigrams were identified.

Conclusions: The clinical diagnosis of stress fractures of the pelvis and the lower extremity is not reliable. If the primary radiography is negative, the use of MR imaging is preferrable to bone scintigraphy and follow-up radiographs. The diagnostic value of MR imaging is higher than bone scintigraphy.

Keywords: diagnosis; fractures, stress; MRI; scintigraphy; x-ray

Prehosp Disast Med 2002;17:s??.