Reliability of the Penny and Beit CURE Radiologic Classifications of Pediatric Patients with Chronic Hematogenous Osteomyelitis in the Philippine General Hospital

Abstract

Objective. This study aimed to evaluate the inter- and intraobserver reliability of the Penny and Beit CURE radiologic classifications of pediatric patients with Chronic Hematogenous Osteomyelitis (CHOM) in the Philippine General Hospital (PGH).

Methods. Thirty-four pre-operative radiographs of PGH pediatric patients with CHOM were classified by seven orthopedic surgeons using both Penny and Beit CURE Classification systems. Two sets of radiographs were sent to the surgeons twice, four weeks apart, to classify. The Fleiss and Cohen κ statistics were used to determine inter- and intraobserver reliabilities, respectively.

Results. The Penny Classification had a slight to fair interobserver reliability (Fleiss κ = 0.17 and 0.24) and a fair intraobserver reliability (Cohen κ =0.35) with a 49.58% average intraobserver agreement. The interobserver reliability when including all Beit CURE classification subtypes was fair (κ = 0.28 and 0.31). This improved to moderate (κ = 0.41 and 0.54) when using only the four main types of the Beit CURE classification with a 77.31% intraobserver agreement.

Conclusion. The Beit CURE classification for pediatric CHOM had higher inter- and intraobserver agreement rates than the Penny classification. Further improvement in reliability can be made by combining B2 and B3 subtypes under the Beit CURE classification.