Artikel
The ageement rate and accuracy of classification of distal radius fracture by plain roentgenography
Suche in Medline nach
Autoren
Veröffentlicht: | 6. Februar 2020 |
---|
Gliederung
Text
Objectives/Interrogation: Plain roentgenography is primarily used to classify fracture types of distal radius fractures. Although the AO classification is also frequently used to classify distal radius fractures, many reports have pointed out that the AO classification is not reproducible. We examined the following two points regarding distal radius fractures: how far the results of plain roentgenography diverge from those of CT scan, and how similar the interpretations of plain roentgenography are among doctors.
Methods: We examined 113 patients with distal radius fractures (115 wrists) who had visited our hospital and undergone both plain roentgenography and CT scan between 2012 and 2014. Three orthopedic surgeons interpreted biplanar plain X-ray images and evaluated them by the AO classification. Classification methods for the CT images were discussed by the concerned doctors to standardize the results. The agreement rates among the doctors who had classified the plain X-ray images were calculated using the Kappa coefficient between every two doctors.
Results and Conclusions: The correct answer rates of plain X-ray diagnoses according to the AO classification when considering CT diagnoses as correct answers were as follows: The correct answer rates according to the Type classification (A, B, and C) were high (66%-75%). The correct answer rates according to the Group classification (A2, A3, B1, B3, C1, C2, and C3) were lower (33%-52%). The correct answer rates according to the subgroup classification were very low (24%-38%). When analyzing the correct answer rates of each group, we found out that those of the type C Subgroups were lower and that those of C1 and C2 were especially low (27% and 25% respectively). In particular, 25% of the fractures diagnosed as type C by CT scan were misclassified as Type A by plain roentgenography. The agreement rates between the examiners using the Kappa coefficient were moderate for the type classification and poor for the Group classification.
Many reports have pointed out that the agreement rates of X-ray diagnoses among examiners are low. Our investigation also revealed that the agreement rates among the doctors who had classified the plain X-ray images were low. In addition, 25% of the type C fractures were misclassified as type A, resulting in overlooks of intraarticular fractures. We thus concluded that combined adoption of CT scan and roentgenography is desirable when deciding on treatment methods for distal radius fractures.