Background: The aim of this study was to assess the accuracy and intraobserver reliability of the technique of penlight transillumination of simulated hand tumors as well as the rationale for the technique.
Methods: Eight observers examined small (9.5 mm) plastic spheres in a fresh frozen cadaveric human hand 3 weeks apart in a blinded manner. The observers were divided into two overall groups based on their level of training. Four spheres simulating hand tumors (two dorsal and two palmar) were placed subcutaneously. The spheres were known to either transilluminate or to be opaque. The observers noted their impression as to whether the spheres either did or did not transilluminate. Accuracy and multi-rater-kappa-statistical analysis were performed.
Results: The overall accuracy was 87.5%: 95% for senior group, 81% for junior group (P = .388, not significant). The average kappa of the intraobserver reliability overall was 0.46. The senior group had a kappa value of 0.67 (substantial agreement), the "junior" group: 0.25 (fair agreement).
Conclusions: Accuracy at correctly determining whether or not a small hand tumor transilluminated was high. The senior group was more accurate overall in correctly determining transillumination, though not with statistical significance. Intraobserver reliability was high for the senior group and less robust for the junior group.
Keywords: Hand tumor; Transillumination.