2019 ISAKOS Biennial Congress ePoster #617
AIIS Morphologic Classification: Evaluation of the Accuracy of Plain Radiographs
Guillaume D. Dumont, MD, Columbia, SC UNITED STATES
Andrew Swentik, MD, Columbia, SC UNITED STATES
William Melton, MD, Columbia, SC UNITED STATES
Alfredo Arrojas, MD, Columbia, SC UNITED STATES
Adam Money, MD, Columbia, SC UNITED STATES
J. Benjamin Jackson, MD, Columbia, SC UNITED STATES
Stephen A. Parada, MD, Augusta, GA UNITED STATES
University of South Carolina School of Medicine, Columbia, SC, UNITED STATES
FDA Status Not Applicable
Assessment of plain radiographs for classification of AIIS morphology.
The anterior inferior iliac spine (AIIS) has been described as an extra-articular source of hip impingement, and is commonly found in association with traditional FAI. A three type, CT based, AIIS morphologic classification system has been described by Hetstroni et al. Plain radiographs have also been used to reliably measure distance from the acetabular rim to the AIIS. The purpose of this study was to assess the utility of plain radiographs in classifying AIIS morphology using the previously described CT classification system.
A single surgeon’s case list was retrospectively reviewed to identify patients that underwent primary hip arthroscopy for treatment of FAI and had a preoperative 3D CT in addition to a three view series of hip radiographs including 1) anteroposterior view of the pelvis; 2) Dunn 45 lateral view; and 3) false profile view. Each CT was evaluated to determine the AIIS type. Six reviewers (three attending orthopaedic surgeons and three orthopaedic residents) reviewed the radiographs of each patient twice on non-consecutive days and classified the AIIS morphology. Overall accuracy of radiographic classifications; accuracy on the first and second trials; and classifications by attending and resident physicians were compared. Accuracy of classifying Type 1 versus Type 2 AIIS was also compared.
76 patients were included. Based on CT classification, the series included 26 Type 1, 49 Type 2, and 1 Type 3 AIIS. 479/912(52.5%) of classifications made by plain radiographs were correct. Attending surgeons classified the AIIS correctly on 237/456(52.0%) attempts versus 242/456(53.1%) by resident surgeons (p=0.74). Reviewers classified the AIIS correctly on 247/456(54.2%) first attempts and 232/456(50.9%) second attempts (p=0.32). Type 1 and Type 2 AIIS were correctly identified on 132/312(42.3%) and 342/588(58.2%) of attempts (p<0.0001).
Our study found standardized 3 view plain radiographs not to be reliable for classifying AIIS morphology based on the previously described CT based classification system. Accuracy of classification was similar on first and second attempts, and by resident and attending surgeons. Type 2 AIIS were correctly classified more frequently than Type 1 AIIS. CT remains the gold standard for identification of AIIS variants in patients with symptomatic FAI.