RESUMO
INTRODUCTION: Classification systems are used for communication, planning treatment options, predicting outcomes and research purposes. The majority of subtrochanteric fractures are now treated with intramedullary nails and therefore questioning the need for classification. OBJECTIVES: To assess the intra- and inter-observer reproducibility of the Seinsheimer, AO and Russell-Taylor (RT) classification systems and to assess a new simple system (MCG). MATERIALS AND METHODS: The MCG system was developed to alert the surgeon to potential hazards: type 1-subtrochanteric fracture (ST#) with intact trochanters, type 2-ST# involving greater trochanter (entry point for nailing difficult), and type 3-ST# involving lesser trochanter (most unstable). Thirty-two anteroposterior and lateral radiographs of subtrochanteric fractures were classified independently for each of the 4 classification systems by 4 observers on 2 separate occasions. RESULTS: The intra- and inter-observer variation was poor in all systems (highest Kappa 0.35). MCG had the best reproducibility followed by RT, then AO and Seinsheimer. The data were re-analysed to determine whether the findings were due to the presence of too many subgroups and whether the observers could more accurately identify important individual subclassifications: Seinsheimer 3a, AO31-A3.1, RT 1 or 2, RT a or b, and MCG3. The MCG3 had the narrowest ranges for intra- and inter-observer reproducibility. CONCLUSIONS: The classification systems analysed in this study have poor reproducibility and seem to be of little value in predicting the outcome of intramedullary nailing as all of the fractures achieved union. The MCG system may be of some use in alerting the surgeon to potential problems.
