Back to Annual Meeting Posters
Does the Modified Gartland Classification Clarify Decision-making?
Sophia Leung, MD1; Ebrahim Paryavi, MD2; Martin Herman, MD3; Paul D. Sponseller, MD, MBA4; Joshua Abzug, MD1
1University of Maryland Medical Center, Baltimore, MD; 2University of Maryland School of Medicine, Baltimore, MD; 3St. Christopher's Hospital for Children, Philadelphia, PA; 4Johns Hopkins University, Baltimore, MD
Introduction: The modified Gartland classification system for pediatric supracondylar fractures is often utilized as a communication tool to aid in determining whether or not a fracture warrants operative intervention. This study sought to determine the inter- and intra-observer reliability of the classification system as well as to determine if the study participants agreed that a fracture warranted operative intervention regardless of the classification system.
Materials and Methods: 200 AP and lateral radiographs of pediatric supracondylar humerus fractures were retrospectively reviewed by 3 fellowship trained pediatric orthopaedic surgeons and classified as Type I, IIa, IIb or III. The surgeons were then asked to record whether they would treat the fracture non-operatively or operatively. Kappa coefficients were calculated to determine inter- and intra-observer reliability.
Results: Overall, the modified-Gartland classification had moderate interobserver reliability with kappa coefficients of 0.638 (0.557-0.710), as well as high intraobserver reliability, with coefficient of 0.799. However, a low interobserver rate was found when differentiating between Type IIa and IIb, with a coefficient of 0.240 (0.116-0.372). There was moderate to high interobserver reliability for decision to operate, with a coefficient of 0.691 (0.598-0.773), and high intraobserver reliability, with a coefficient of 0.760. For fractures classified as Type I, the decision to operate was made 3% of the time. If classified as Type IIa, the decision to operate was made 27% of the time, and 99% of the time if classified as Type IIb. The decision was made to operate for 100% of fractures classified as Type III.
Conclusions: There is almost full agreement for the non-operative treatment of Type I fractures and operative treatment for Type III fractures. There is agreement that Type IIb fractures should be treated operatively and that the majority of Type IIa fractures should be treated non-operatively. However, the interobserver reliability for differentiating between Type IIa and IIb fractures is low. Our results validate the Gartland classfication system as a method to help direct treatment of pediatric supracondylar humerus fractures, although the modification of the system, IIa versus IIb, seems to have limited reliability and utility. This suggests that rotational deformity is difficult to assess in pediatric supracondylar humerus fractures. Terminology based on decision to treat would lead to a more clinically useful classification system in the evaluation and treatment of pediatric supracondylar humerus fractures.
Back to Annual Meeting Posters