A novel classification of intraoperative ulnar nerve instability to aid transposition surgery.

Abstract

The purpose of this study was to compare inter- and intraobserver agreement of a novel intraoperative subluxation classification for patients undergoing ulnar nerve surgery at the elbow. We hypothesize there will be strong inter- and intraobserver agreement of the 4-category classification system, and reviewers will have substantial confidence while reviewing the classification system. Four blinded fellowship-trained orthopedic hand surgeons reviewed 25 videos in total on 2 separate viewings, 21 days apart. Variables collected were ulnar subluxation classification (A, B, C, or D) and a confidence metric. Subsequent to primary data collection, classification grading was stratified into A/B or C/D subgroups for further analysis. Cohen κ scores were used to evaluate all variables collected in this study. The interpretation of κ scores included ≤0.0 as no agreement, 0.01-0.20 as none to slight, 0.21-0.40 as fair, 0.41-0.60 as moderate, 0.61-0.80 as substantial, and 0.81-1.0 as almost perfect agreement. Interobserver agreement of subluxation classification as a 4-category scale demonstrated a moderate agreement on first viewing, second viewing, and when both viewings were combined (κ = 0.51, 0.51, and 0.51 respectively). Seventy-five percent (3 of 4) of reviewers had moderate intraobserver agreement for ulnar nerve subluxation classification, whereas 1 reviewer had substantial intraobserver classification (κ = 0.72). Overall, there was high confidence in 65% of classification scores in the second round of viewing, which improved from 58% in the first viewing round. When ulnar subluxation classification selections were regrouped into classes A/B or C/D, 100% of reviewers had substantial interobserver (κ = 0.74-0.75) and substantial to almost perfect intraobserver (κ = 0.71-0.91) agreement. The 4-category classification was reproducible within and between reviewers. Agreement appeared to increase when simplifying the classification to 2 categories, which may provide guidance to surgical decision making. The validation of a reproducible classification scheme for intraoperative ulnar subluxation may aid with decision making and further postoperative outcomes research.