Intraobserver and Interobserver Reliability of the Peritubercle Lucency Sign in Slipped Capital Femoral Epiphysis

BACKGROUND An area of enlargement of the metaphyseal socket around the epiphyseal tubercle, termed the peritubercle lucency sign, has recently been introduced as a possible predictor of contralateral slipped capital femoral epiphysis in patients with previous unilateral slipped capital femoral epiphysis. This study aimed to assess intraobserver and interobserver reliability for detecting the presence or absence of the peritubercle lucency sign. METHODS Thirty-five radiographs were presented to 6 fellowship-trained pediatric orthopaedic surgeons on 2 separate occasions 30 days apart, ensuring that the images were shown in a different order on the second exposure. Both times the reviewers recorded whether the peritubercle lucency sign was present or absent in each of the radiographs. Statistical analysis was performed to determine the intraobserver and interobserver reliability. RESULTS In the intraobserver analysis, percent agreement between the first and second time the radiographs were reviewed varied between 62.9% and 85.7%, for an average intraobserver agreement of 74.8%. κ values for the 6 reviewers varied between 0.34 and 0.716, with an average intraobserver κ value of 0.508. The interobserver percent agreement was 40.0% for the first time the radiographs were reviewed, 42.9% the second time, and the overall interobserver percent agreement was 29%. The interobserver κ value was 0.44 the first time the radiographs were reviewed, 0.45 the second time, and the overall interobserver κ value was 0.45. DISCUSSION On the basis of our findings, the peritubercle lucency has modest intraobserver and interobserver reliability at best and should be used with other currently used factors, such as age, presence of endocrinopathy, status of triradiate cartilage, posterior sloping angle, and modified Oxford score, in determining the need for prophylactic pinning. Further refinement of the definition of the peritubercle lucency sign may be needed to improve agreement and reliability of the sign. LEVEL OF EVIDENCE Level III-prognostic study.