SLR - May 2020 - Olga Drogomiretskiy

Inter- and Intraobserver Reliability of Non-Weight-Bearing Foot Radiographs Compared With CT in Lisfranc Injuries

Reference:  Ponkilainen, V, Partio N, Salonen E.  Inter- and Intraobserver Reliability of Non-Weight-Bearing Foot Radiographs Compared with  CT in Lisfranc Injuries. Archives Orthopedic Trauma Surgery. 2020 March 5th. 

Scientific Literature Review

Reviewed By: Olga Drogomiretskiy, DPM
Residency Program: Franciscan Health System-St. Francis Hospital – Federal Way, WA 

Podiatric Relevance: The diagnosis of Lisfranc injuries can be challenging since they range from a wide variety of purely ligamentous to osseous injuries. Often, it is suggested to initially obtain weightbearing radiographs and possibly magnetic resonance imaging (MRI) to rule out ligamentous sprain or more commonly a CT scan for more precise detection of bony injuries. At times, initial diagnostic imaging includes non-weightbearing radiographs due to patients' inability to apply weight on injured foot. With exception of gravity sagittal stress exam, current literature is limited on the accuracy of these initial diagnostics and their reliability and reproducibility among observers. 

Methods: This is a level III retrospective review from University of Tampere where 100 sets of non-weightbearing radiographs were evaluated  for Lisfranc injury following acute trauma by three senior orthopedic surgeons and three residents from January 2012 to December 2016 and then confirmed with a CT scan. Patients with intra-articular fractures and avulsion fractures around the tarsometatarsal joint complex were included, while those with extra-articular metatarsal injuries were excluded. The patients were then divided into displaced and non-displaced injuries with <2 millimeters considered non-displaced. Intra- and interobserver reliability was then assessed and repeated in three month intervals. Fleiss kappa was used to evaluate interobserver reliability and Cohen kappa used to assess intraobserver reliability. Mean sensitivity and specificity values of all observes were also recorded. 

Results: Of the 100 Lisfranc injuries assessed on non-weightbearing x-ray films between the six observers, the Fleiss kappa (κ) coefficient for interobserver reliability resulted in moderate correlation κ = 0.50 during first evaluation and κ = 0.58 during second evaluation. After three months, the evaluation was repeated and the Cohen's kappa (κ) coefficient for intraobserver reliability showed substantial correlation κ = 0.71. The mean sensitivity was 76.1 percent and specificity was 85.3 percent for all observers. The sensitivity of subtle injuries was lower than severe injuries (65.4 percent vs 87.1 percent p = 0.003). The number of missed cases was higher among the non-displaced injuries in comparison to the displaced variants (11 vs 4 p=0.002). The false-negative rate was 23.9 percent and the false-positive rate was 14.7 percent. 

Conclusions: The authors concluded that Lisfranc injuries assessed had a moderate agreement between observers and substantial agreement by the same observer at different time points. They also noted a significant number (23.9 percent) of injuries that were missed when evaluated by plain radiographs with more not diagnosed that were minimally / non-displaced. To date, this is the first study that assessed Lisfranc injuries on non-weightbearing films and evaluated inter- and intraobserver reliability. Limitations include the retrospective design of this study, lack of MRI use for evaluating missed ligamentous injuries and no weightbearing radiographic films. An additional  limitation includes the fact that all radiographic films were evaluated by orthopedic surgeons and residents, yet most of the time initial assessment is done by our emergency medicine or medicine colleagues. The results of this study confirm that a significant number of injuries that are missed by radiographs and to truly confirm the diagnosis by CT scan if available. 

Educational Opportunities