Diagnosis of union of distal tibia fractures: accuracy and interobserver reliability

Abstract

Decisions about return to activity and additional surgery are often made on the basis of radiographs obtained 3 months after injury. If radiographs 3 months after injury cannot reliably and accurately diagnose union, then patients may be needlessly disabled and might receive unnecessary treatments including surgery. We evaluated the accuracy and the reliability of the diagnosis of union or eventual union on radiographs obtained 3 months after open reduction and internal fixation of a fracture of the distal tibia by having 69 trauma surgeons evaluate radiographs of 33 consecutively treated patients in an online survey. Observers were also asked to judge specific criteria that are commonly used to diagnose fracture union. There was moderate interobserver reliability for the diagnosis of union or diagnosis of "eventual union". The interobserver agreement for the various specific radiographic signs of union varied between fair to moderate. The sensitivity of radiographs for diagnosis of "union or eventual union" of distal tibia fractures was 47%, the specificity was 73% and the accuracy was 68%. The prevalence adjusted positive predictive value was 25% and the negative predictive value was 88%. Diagnosis of union based on radiographs 3 months after injury is only moderately reliable and accurate but has a high negative predictive value. Decisions about activity level and additional treatment 3 months after injury should not be based on radiographs alone.