The effect of labeling photo documents in wrist arthroscopies on intra- and interobserver reliability

Abstract

The reproducibility of diagnoses based on photo documents in wrist arthroscopies is limited and is expected to improve if the photos are labeled with illustrated structures. The purpose of this study was to determine the effect of labeling photo documents on intra- and interobserver agreement and reliability of standard photo documentation in wrist arthroscopies. Digital photographs of 50 arthroscopies were re-evaluated by the surgeon as well as by two independent hand surgeons. First the photos were presented unlabeled in a random order, then the labeled photos in a uniform order. Intra- and interobserver reliability was assessed, and expressed by kappa coefficients. Overall, labeling the photos resulted in a slight improvement in intra- and interobserver reliability (0.573/0.444/0.420 vs. 0.518/0.412/0.212). The time needed to conceive the photo documents, however, was shortened when the photos were labeled. The cartilage status was assessed considerably more accurately if the photos were labeled (0.556/0.560/0.422 vs. 0.459/0.326/0.240; t test: P=0.094). Whereas the SL ligament was assessed more accurately according to labeled photos (P=0.100), the agreement rates for the assessment of other ligament structures (TFCC, LT and radiopalmar ligaments) were not substantially affected by labeling the photos. On re-evaluation of the unlabeled as well as the labeled photos, intraobserver reliability was better than interobserver reliability (0.518 vs. 0.412/0.212 and 0.573 vs. 0.444/0.420). Labeling simplifies but does not necessarily improve the reproducibility of photo documents in wrist arthroscopies. To display the cartilage status and the integrity of the SL ligament, digital photo documents should be labeled with the illustrated structure or joint surface.