There is poor accuracy in documenting the location of labral and chondral lesions observed during hip arthroscopy.

PURPOSE

To determine and compare the accuracy and interobserver reliability of the different methods for localizing acetabular labral, acetabular chondral, and femoral head chondral lesions with hip arthroscopy .

METHODS

Three cadaver hips were placed in the supine position. Three labral, three femoral chondral, and six acetabular chondral lesions were made in each cadaver using electrocautery. Six surgeons classified the lesions according to different classification systems (clock-face, geographic, Method-G) using hip arthroscopy and standardized portals. Identification of each lesion was performed after conclusion of the study through open dissection and surgical hip dislocation to be used as the "gold-standard." Accuracy was calculated as the number of correct answers divided by total number of responses for a given system. The interobserver reliability was calculated using the kappa coefficient. The different classification methods were compared. All P values were reported with significance set at P < 0.05.

RESULTS

The clock-face method had an accuracy of 74% (95% CI, 60%-85%) and interobserver reliability of 0.19 (95% CI, 0.11-0.26) while the geographic method had an accuracy of 50% (95% CI, 36%-64%) and interobserver reliability of 0.21 (95% CI, 0.05-0.31) for acetabular labral lesion identification (P > 0.05). The acetabular chondral lesion identification accuracy was 56% (95% CI, 46%-65%) for Method G, 66% (95% CI, 56%-75%) for Method G-simp, and 63% (95% CI, 53%-72%) for the geographic system (P > 0.05) with an interobserver reliability of 0.31 (95% CI, 0.27-0.35), 0.34 (95% CI, 0.28-0.40), and 0.40 (95% CI, 0.34-0.45), respectively (P > 0.05). Femoral chondral lesion identification accuracy was 74% (95% CI, 60%-85%) for Method G, 43% (95% CI, 29%-57%) for the geographic method, and 59% (95% CI, 45%-72%) for the geographic-simp system with interobserver reliability of 0.37 (95% CI, 0.27-0.47), 0.34 (95% CI, 0.28-0.40), and 0.40 (95% CI, 0.29-0.51), respectively (P > 0.05). Method G was significantly more accurate than the geographic system (P = 0.001).

CONCLUSIONS

There was poor to fair accuracy and interobserver reliability of the reporting systems for localization of labral, acetabular chondral, and femoral chondral lesions encountered during hip arthroscopy. The study suggests there is a need for a new method that is easy to use, reliable, reproducible and accurate.