Treatment pathways of fifth metatarsal fractures are commonly directed based on fracture classification, with particular proximal fractures typically requiring closer observation and possibly more aggressive management. Our aim is to investigate the interobserver reliability of assessment of subtypes of fifth metatarsal fractures. We included all patients referred to our virtual fracture clinic with a suspected or confirmed fifth metatarsal fracture. Plain anteroposterior radiographs were reviewed by 2 novice observers, initially trained on the fifth metatarsal classification identification. Eight different zones were defined based on anatomical location. Patients were excluded from analysis if neither observer could identify a fracture. An interobserver reliability analysis using Cohen κ coefficient was carried out, and degree of observer agreement described using Landis and Koch's description. All data were analyzed using IBM SPSS, version 27. A total of 1360 patients who met the inclusion criteria were identified. The 2 observers had moderate agreement when identifying fractures in all zones, apart from zone 1.2 and distal metaphysis (DM) fractures, which only achieved fair agreement (κ = 0.308 and 0.381 respectively). Zone 3 has slight agreement with zone 2 proximally, and there is an apparent difficulty with distal diaphyseal shaft (DS) fractures, resulting in a lot of crossover with DM, achieving a fair level of agreement (DS 312 vs 196; DM 120 vs 237; κ = .398, P < .001). Reliability of subcategorizing fifth metatarsal fractures using standardized instructions conveys moderate agreement in most cases. If the region of the fracture is going to be used in an algorithm to guide a management plan and clinical follow-up during a virtual clinic review, defining fractures of zones 1-3 needs careful consideration. Level III, retrospective cohort study.
Read full abstract