[Intra- and interobserver reliability of digitally photodocumented findings in wrist arthroscopy]

The aim of this study was to evaluate the intra- and interobserver reliability of findings in wrist arthroscopies documented by digital photography.

102 consecutive wrist arthroscopies were documented by at least 6 photographs: (1) radioscaphoidal joint with radiopalmar ligaments, (2) scapholunate ligament from radiocarpal, (3) TFCC, (4) lunotriquetral joint with ulnar recessus, (5) scapholunate, and (6) lunotriquetral joint from midcarpal. More photographs were taken of additional pathologies. Arthroscopic findings were documented in an arthroscopy form (synovitis: no/yes, articular effusion: no/yes, cartilage lesion: no/yes). Lesions of the scapholunate ligament were classified according to Geissler, scapholunate and lunotriquetral ligament lesions were graded as partial or complete, TFCC lesions were categorised according to Palmer and trampoline effects noted. Three months later the 6 standard photos and then all photos were re-evaluated by the scoping surgeon and two experienced hand surgeons. Cohen's kappa was used as a measure of agreement between findings three months postoperatively and those in the operation report. Further statistical analyses were done, e. g., to compare intra- and interobserver reliability.

In general intraobserver reliability (mean kappa 0.44) was better than interobserver reliability (0.28, 0.30). Assessing all available photos (0.47) did not improve reliability compared to assessing only the six standard photos (0.44). On grading scapholunate ligament lesions as partial or complete the interobserver reliability was fair (0.37, 0.30). Using Geissler's classification the interobserver reliability was κ=0.34 and 0.23. There was substantial intra- (0.61) and interobserver reliability (0.63, 0.60) for cartilage lesions. While intraobserver reliability for TFCC lesions was substantial (0.61), interobserver reliability was only moderate (0.31, 0.48).

Even with wrist arthroscopy, the best diagnostic modality for scapholunate ligament, TFCC and carpal cartilage lesions, hand surgeons do not agree in diagnosis and pathology. Therefore, the documentation of arthroscopic findings has to be improved. Digital video documentation of wrist arthroscopy may illustrate wrist pathologies more adequately. However, this still has to be proven by further studies.