Evaluation of the Reliability of Classification Systems Used for Distal Radius Fractures

Abstract

The objective of this investigation was to evaluate the reliability of classification systems by determining inter- and intraobserver agreement in displaced distal radius fractures. Radiographs of 32 patients (21 men and 11 women with a mean age of 41.6 years) who presented with a displaced distal radius fracture were classified by 9 orthopedic surgeons (5-25 years experience) using 5 different classification systems (Fernandez, AO, Frykman, Melone, and Universal Classification systems) twice with 20-day intervals. The results were processed with kappa statistics and used in assessment of inter- and intraobserver agreement of the classification systems. When classification systems were compared, the highest kappa coefficient in intraobserver agreement was determined in Universal classification (0.621). Fernandez (0.474), AO (0.309), Frykman (0.305), and Melone classification systems (0.262) followed the Universal system respectively. Kappa statistical results were evaluated using the Landis Koch score system for the assessment of interobserver agreement. According to the Landis Koch score system, the results were insufficient in all classification systems. Fernandez classification system had the highest interobserver agreement (0.235) and Melone classification system had the lowest interobserver agreement (0.056). According to the results of our study, the systems used to classify the displaced distal radial fractures are insufficient. A new classification system that ensures the 3-dimensional assessment of the fracture is more user-friendly and a high inter- and intraobserver agreement is necessary.

Adult distal radius fractures classification systems: essential clinical knowledge or abstract memory testing?

“…1,6 In addition, we suggest that they should be simple, easy to remember and have acceptable interobserver agreement (reliability), intra-observer agreement (reproducibility) and validity. 7 Reliability and reproducibility are demonstrated by the ability of a classification to return the same result for a particular patient's data when shown to multiple observers (reliability) or to the same observer viewing the same patient's data at different time points (reproducibility). Validity is the accuracy with which the classification determines the true fracture type.…”

“…It was shown to have slight to moderate reliability and reproducibility with kappa values of 0.337, 0.4-0.6, respectively, in one study 19 and 0.056, 0.262 in another study. 7 Observers had the greatest challenge in identifying the four fragments on x-ray. 19 …”

“…19 Reliability and reproducibility worsens to fair to moderate when divisions within each type are considered. 7,16,19 Mayo (1992) In 1992, the Mayo Clinic proposed its own classification. It is a four-part classification similar to the Frykman classification in that is focuses on fracture extension into the radiocarpal or radioulnar (Fig 7).…”

“…It has shown fair to moderate reliability (k = 0.226-0.41) but fair to substantial reproducibility (k = 0.49-0.621). 7,20,21 Conclusion Classification systems in their most useful form can be tools for succinct rapid communication in clinical contexts between healthcare professionals. Further, when they elude to treatment, recovery or prognosis, they become an invaluable tool for a clinician and can often guide or polarise treatment strategy.…”

The Reliability of Mobile Multimedia Messaging (MMS) for Decision Making in Distal Radius Fractures: An Effective Alternative

“…Furthermore, there is poor inter and intra-observer reliability among distal radius classification systems. 11,12 A more relevant clinical outcome is whether the management following reviewing an image via MMS is reliable when compared to viewing the images via a PACS system. To that end, a retrospective study was designed to assess whether MMS was reliable in predicting the recommended management when compared to PACS.…”