| Objective The X-ray film and CT three-dimensional reconstruction images of the patients with femoral neck fracture were collected and classified into Garden classification.The inter-observer and intra-observer repeatability and consistency of the two methods were compared,and the inter-observer and intra-observer repeatability and consistency of the three-classification Garden classification(typeⅠ+Ⅱ,Ⅲ,Ⅳ)were compared.And to explore the certainty of the existence of Garden type Ⅰ femoral neck fracture.Methods A total of 886 elderly patients with femoral neck fracture admitted to the Affiliated Central Hospital of Shenyang Medical College and Shengjing Hospital of China Medical University from June 2018 to June 2021 were selected in strict accordance with the inclusion criteria,aged 65-96 years.(1)The patient information was removed from the X-ray film data of 886 patients,and the X-ray film data were randomly numbered.Four radiologists(two chief physicians Aand B,two residents C and D)and four orthopedists(two chief physicians E and F,two residents G and H)with more than 15 years of work experience performed Garden classification.All observers were familiar with the Garden classification criteria,and the direction and position of fracture line and soft tissue injury were observed to complete the Garden classification of femoral neck fractures.(2)The patient information was removed from the CT image data of 886 patients,and the CT image data were randomly numbered.The above 8 physicians judged the CT images separately again to complete the Garden classification of femoral neck fractures.All judgments were recorded and then deposited in a database managed by the investigators.(3)Three months later,the X-ray and CT images were classified by Garden classification by the 8 doctors.The results of four-class Garden classification(types Ⅰ,Ⅱ,Ⅲ,and Ⅳ)and three-class Garden classification(types Ⅰ+Ⅱ,Ⅲ,and Ⅳ)were compared.(4)After separate classification,the imaging data of 886 patients were reviewed by 8observers.8 observers combined with X-ray film and CT image data to classify the Garden typeⅠfractures.For 8 cases of Garden type I fractures found by common observers,the medical records and surgical records were reviewed to compare the consistency between intraoperative findings and imaging data.SPSS 26.0 statistical software was used to analyze the data.The inter-observer and intra-observer consistency were analyzed by calculating Cohen kappa value.Fleiss’ s statistical method was used to calculate the inter-observer agreement kappa value of multiple observers.According to Landis and Koch’s guidelines,kappa value was interpreted with reference to kappa coefficient scale: kappa value was 0.00-0.20,which was interpreted as a low degree of consistency;The kappa value was between 0.21 and 0.40,which was interpreted as general consistency.The kappa value was between 0.41 and 0.60,which was interpreted as a moderate degree of consistency.The kappa value was between 0.61 and 0.80,which was interpreted as a high degree of consistency.kappa values ranged from 0.81 to 1.00,which was interpreted as a very high degree of agreement.Results Garden classification was performed twice according to X-ray films.Based on the consistency test of the four-classification Garden classification(type Ⅰ,Ⅱ,Ⅲ,Ⅳ),the kappa values among chief physicians A,B,E,F were 0.24-0.43,and the kappa values among residents C,D,G,H were 0.18-0.35.The kappa values within chief physicians ranged from0.38 to 0.42,and within residents ranged from 0.26 to 0.36.Inter-observer and intra-observer agreement was mostly at a low level.Garden classification was performed twice according to CT images.The consistency test results based on the four-category Garden classification(type Ⅰ,Ⅱ,Ⅲ,and Ⅳ)showed that kappa values among and within chief physicians were 0.56-0.76,and kappa values among and within residents were 0.28-0.35.Most of the findings showed high consistency between and within chief physicians.The number of Garden type Ⅰ cases judged by each observer based on CTimages was smaller than that judged by X-ray films.The number of Garden typeⅡ fractures judged by CT images was more than that by X-ray films,but the total number of Garden type Ⅰ and Ⅱ fractures judged by CT images was basically the same as that by X-ray films.The classification results of Garden type Ⅰ and type Ⅱ fractures were combined into one group,and the consistency was tested again.The kappa values of Garden classification(type Ⅰ+Ⅱ,Ⅲ,Ⅳ)based on X-ray film or CT images ranged from 0.76 to 0.90,and most of them showed very high consistency between and within observers.After the classification experiment,52 patients were confirmed as Garden type Ⅰfractures by 8 observers’ joint interpretation of X-ray films and CT images.The clinical data and surgical records of the patients were reviewed,and 14 cases were treated with closed reduction and internal fixation with cannulated screws,and 38 cases were treated with hemiarthroplasty or total hip arthroplasty.For patients treated with closed reduction,the true condition of the femoral neck fracture could not be determined.In the 38 patients treated with hip arthroplasty,all the femoral neck fractures were complete: 22 cases were anterior-posterior displaced or valgus impacted fractures,and only 16 cases were non-displaced fractures,whose fracture lines also completely involved the lateral and medial cortices.Conclusion The inter-observer and intra-observer reproducibility and consistency of the traditional four-classification Garden classification(types Ⅰ,Ⅱ,Ⅲ,and Ⅳ)were low,while the inter-observer and intra-observer reproducibility and consistency of the three-classification Garden classification(types Ⅰ+Ⅱ,Ⅲ,and Ⅳ)were high.The Garden classification of femoral neck fractures should be combined with the X-ray film and CTimage data of patients,and it is recommended that the senior chief physician judge the classification,which can be more accurate and better guide the decision-making of treatment methods,so as to improve the prognosis of patients.Partial fracture of Garden type Ⅰ femoral neck may not exist in elderly patients.All of them should be considered as Garden type Ⅱ fractures and should be treated according to the treatment plan of Garden type Ⅱ fractures. |
PDF Full Text Request