Clinical study on new risk assessment and prediction system for early osteonecrosis of the femoral head

Abstract

To establish a risk assessment and prediction system for early osteonecrosis of the femoral head (ONFH) in order to predict the collapse risk. The risk assessment system for early necrosis and collapse of femoral head was established based on the combination of Steinberg stage, ABC typing and the proportion of the proximal sclerotic rim. Firstly, Steinberg stage system was applied. ABC typing was applied to predict risk in stage I, type C was risk free, type B was low risk, type A and type BC were medium risk, type A-C and type AB were high risk. The classification of proximal sclerotic rim was first applied when the Steinberg stage was Ⅱ-Ⅲ, and type 2 was expected to be low risk. If the classification of proximal sclerotic rimwas type 1, then the ABC typing was applied, type C was risk-free, type B was low risk, type A and type BC were medium risk, type A-C and type AB were high risk. According to this prediction system, the collapse risk of femoral head in 188 cases(301 hips) were predicted by retrospective analysis. All the hips were enrolled at the out-patient department of orthopedic in Guang'anmen Hospital attached to China Academy of Chinese Medical Science. The consistency of the prediction results of three doctors and one doctor at different times were evaluated. Among them, 136 cases were male, 52 were female. 75 cases were single hip, 113 were double hip. The age of the patients wa 19 to 64(42.61±12.07) years. The natural course of disease was 0.33 to 5.00(3.62±1.93) years. 206 hips in 301 hips had collapsed, with a collapse rate of 68.44%. In the risk-free group, none hip had collapsed, with a collapse rate of 0%. In the low-risk group, 9 hip in 91 hips had collapsed, with a collapse rate of 9.89%. In the medium-risk group, 12 hip in 19 hips had collapsed, with a collapse rate of 63.16%. And in the high risk group, 185 hips in 190 hips had collapsed, with a collapse rate of 97.37%. They were significantly differences in their collapse rate (P=0.00) in the following order:high-risk group> medium-risk > low-risk group > risk-free group. The prediction value of the system was high (AUC=0.95, P=0.00). The results predicted by different doctors were consistent (ICC=0.94, P=0.00), and the results predicted by the same doctor at two different times were consistent (Kappa coefficient =0.90, P=0.00). The risk assessment and prediction system for early ONFH selects different methods to predict the risk of collapse according to the imaging characteristics of different stages, which is combines with the comprehensive assessment of multiple risk factors. The system is applicable to a wide range, simple operation and convenient for clinical application.