Our objective was twofold: (1) to determine whether the site of patients' pain correlates with abnormal findings on three-compartment wrist arthrograms and (2) to determine whether the directional nature (uni- vs bidirectional) of communication(s) between adjacent wrist joint compartments shown on three-compartment wrist arthrograms correlates with the presence and site of pain.

We retrospectively studied 462 three-compartment wrist arthrograms of 458 patients. The patients were divided into four groups according to the site of pain, the correlation of that site with the site of the communication, and the directional characteristics of communications shown on the arthrogram: (1) exact correlation between site of pain and arthrographic findings, (2) correlation only between side of pain (ulnar or radial) and arthrographic findings, (3) ambiguous correlations (e.g., multifocal, midline, or diffuse pain), and (4) no correlation between site of pain and arthrographic findings.

In 191 patients (42%), 226 intercompartmental communications were detected (51 uni- or 173 bidirectional). One hundred sixty-one patients had only one communication (uni- or bidirectional), and 30 patients had more than one (uni- and/or bidirectional) communication for a total of 65 defects. Regardless of the directional nature of the communication, no positive association between sites and/or the side of symptoms and sites of communications was demonstrable. However, a potential association appears to exist between ulnar-sided pain and combined lunotriquetral and triangular fibrocartilage tears.

Our results suggest that the directional nature of interjoint communications has little or no clinical significance. Thus, with the exception of seeking the cause of ulnar-sided pain, our results question the usefulness of routine arthrography for correlating sites of patients' symptoms and individual interjoint communications.