Julio Urrutia1, Tomas Zamora2, Pablo Besa2, Maximiliano Zamora2, Daniel Schweitzer2, Ianiv Klaber2. 1. Department of Orthopaedic Surgery, School of Medicine, Pontificia Universidad Catolica de Chile, Marcoleta 352, Santiago, Chile. Electronic address: jurrutia@med.puc.cl. 2. Department of Orthopaedic Surgery, School of Medicine, Pontificia Universidad Catolica de Chile, Marcoleta 352, Santiago, Chile.
Abstract
INTRODUCTION: We performed an agreement study of the AO and the Tronzo classifications of fractures of the trochanteric area to determine if they allow communication among practitioners with different levels of expertise. MATERIAL AND METHODS: Complete radiographs of 70 patients with trochanteric fractures were classified by nine evaluators (three hip sub-specialists, three orthopaedic surgery residents and three medical interns) using the AO and the Tronzo classifications. After a six-week interval, all cases were presented in a random sequence for repeat evaluation. The Kappa coefficient (k) was used to determine inter- and intra-observer agreement. RESULTS: Inter-observer: considering the main AO fracture types, the agreement was moderate for sub-specialists (k = 0.60 [0.50-0.70]), residents (k = 0.58 [0.48-0.69]) and medical interns (k = 0.56 [0.45-0.69]). Using AO sub-types, all groups achieved fair agreement (sub-specialists: k = 0.31 [0.25-0.38]; residents: k = 0.32 [0.26-0.38]; medical interns: k = 0.30 [0.24-0.36]). For the Tronzo classification, sub-specialists (k = 0.56 [0.48-0.65]) and residents (k = 0.47 [0.39-0.55]) obtained moderate agreement; medical interns reached fair agreement (k = 0.33 [0.25-0.41]). Intra-observer: considering the main AO fracture types, sub-specialists (k = 0.79 [0.69-0.89]), residents (k = 0.71 [0.60-0.81]) and medical interns (k = 0.70 [0.59-0.82]) obtained substantial agreement. Considering AO sub-types, sub-specialists (k = 0.50 [0.45-0.56]) and medical interns (k = 0.54 [0.48-0.69]) achieved moderate agreement; residents (k = 0.39 [0.33-0.45]) achieved fair agreement. Using the Tronzo classification, all groups obtained substantial agreement (sub-specialists: k = 0.66 [0.58-0.74]; residents: k = 0.63 [0.55-0.71]; medical interns: k = 0.68 [0.60-0.76]). CONCLUSION: The AO classification allows an adequate communication when considering the main fracture types; the agreement within sub-types is not satisfactory. The Tronzo classification does not allow reliable communication between medical professionals.
INTRODUCTION: We performed an agreement study of the AO and the Tronzo classifications of fractures of the trochanteric area to determine if they allow communication among practitioners with different levels of expertise. MATERIAL AND METHODS: Complete radiographs of 70 patients with trochanteric fractures were classified by nine evaluators (three hip sub-specialists, three orthopaedic surgery residents and three medical interns) using the AO and the Tronzo classifications. After a six-week interval, all cases were presented in a random sequence for repeat evaluation. The Kappa coefficient (k) was used to determine inter- and intra-observer agreement. RESULTS: Inter-observer: considering the main AO fracture types, the agreement was moderate for sub-specialists (k = 0.60 [0.50-0.70]), residents (k = 0.58 [0.48-0.69]) and medical interns (k = 0.56 [0.45-0.69]). Using AO sub-types, all groups achieved fair agreement (sub-specialists: k = 0.31 [0.25-0.38]; residents: k = 0.32 [0.26-0.38]; medical interns: k = 0.30 [0.24-0.36]). For the Tronzo classification, sub-specialists (k = 0.56 [0.48-0.65]) and residents (k = 0.47 [0.39-0.55]) obtained moderate agreement; medical interns reached fair agreement (k = 0.33 [0.25-0.41]). Intra-observer: considering the main AO fracture types, sub-specialists (k = 0.79 [0.69-0.89]), residents (k = 0.71 [0.60-0.81]) and medical interns (k = 0.70 [0.59-0.82]) obtained substantial agreement. Considering AO sub-types, sub-specialists (k = 0.50 [0.45-0.56]) and medical interns (k = 0.54 [0.48-0.69]) achieved moderate agreement; residents (k = 0.39 [0.33-0.45]) achieved fair agreement. Using the Tronzo classification, all groups obtained substantial agreement (sub-specialists: k = 0.66 [0.58-0.74]; residents: k = 0.63 [0.55-0.71]; medical interns: k = 0.68 [0.60-0.76]). CONCLUSION: The AO classification allows an adequate communication when considering the main fracture types; the agreement within sub-types is not satisfactory. The Tronzo classification does not allow reliable communication between medical professionals.