Interobserver variabilty in SPECT/CT, MRI, CT, bone scan and plain radiographs in patients with unspecific wrist pain

Objectives To investigate the interobserver variability of SPECT/CT, MRI, CT, bone scan (BS) and plain radiographs (PR) in patients with unspecific wrist pain.

Methods Thirty-two patients (median age 38 years, range 18 to 73 years; 19 female, 13 male) with unspecific pain of the wrist were evaluated retrospectively. The diagnosis was made by the referring hand surgeon based on patient history, clinical examination, PR and clinical guidelines. All patients were imaged by PR, planar early-phase imaging (BS) and late-phase SPECT/CT imaging including high-resolution CT, and MRI. Two experienced readers analyzed the images based on a standardized read-out protocol. The interobserver variabilities were determined for detecting the relevant lesion, its location and type and etiology of the underlying pathology. Kappa values of the two readers for all analyzed imaging characteristics were calculated. The diagnostic accuracy of each modality was calculated according to the standard of reference (complete clinical examination and imaging) and a mean follow-up of 20 months.

Results Accuracy of PR, BS and CT was generally poor for both readers, but there was partly good agreement on lesion detection (PR 0.73, CT 0.87) and location (PR 0.69, BS 0.63, CT 0.87), and type of pathology (CT 0.74). The other modalities with higher diagnostic accuracy yielded generally high agreement, with SPECT/CT showing mostly superior results (lesion detection: 0.93 vs. 0.72 (SPECT/CT vs. MRI); localisation: 0.91 vs. 0.75; etiology: 0.85 vs. 0.74), while MRI yielded better results in determining the type of the underlying pathology (0.69 vs. 0.75).

Conclusions SPECT/CT may be a reliable diagnostic tool with consistent image reading in patients with unspecific pain of the wrist.