A maximally sensitive clinical decision rule to reduce the need for radiography in mandibular trauma

To identify clinical symptoms and signs that exclude the presence of mandibular fracture in patients presenting with mandibular trauma and thus devise a clinical decision rule that will rule out the need for radiography in some patients.

A prospective study was conducted of consecutive patients with a possible diagnosis of mandibular fracture who attended a city-centre emergency department between July 2000 and December 2001. Clinical symptoms and signs were recorded for each patient using a predesigned proforma. Radiographic evaluation of a suspected mandibular fracture consisted of lateral-oblique and postero-anterior mandibular views. The presence of a fracture was based on the interpretation of the X-rays by a radiologist who was blinded to the clinical probability of a fracture. Data were initially analysed using the chi-square test. Recursive partitioning was then performed to create a maximally sensitive decision tree.

280 patients were included in the study, 65 of whom had a mandibular fracture. A maximally sensitive decision rule was found that identified 5 parameters (malocclusion, trismus, broken teeth, pain with mouth closed, step deformity) whose absence excluded mandibular fracture. This rule has a sensitivity of 100% and specificity of 39% in identifying patients with mandibular fracture. If applied to our patient cohort, this rule would have saved 83 radiographs without missing any fractures.

A simple decision rule is presented that can be used to exclude the need for radiography in a subset of patients with mandibular trauma.