Overutilization of Computed Tomography for Odontogenic Infections

2018 
Purpose There is clear overuse of computed tomography (CT) in the emergency setting, which is associated with a long-term potential for malignancy. This study aimed to determine the rate of overuse of CT for odontogenic infection workup and the correlation of history and physical examination (H&P) findings to diagnose moderate- to high-risk infections. Materials and Methods A retrospective cross-sectional study was implemented to determine the rate of overuse of CT for odontogenic infections. Included patients presented through the emergency department for evaluation of an odontogenic infection. CT was deemed unnecessary if patients presented without “red-flag” signs at physical examination, which included voice change, elevated floor of mouth, signs of inflammation of deep fascial spaces, periorbital edema, nonpalpable inferior border of the mandible, dyspnea, dysphagia or odynophagia, and trismus. Patients could have no evidence of involvement of a moderate- to high-risk space or airway change at CT. Infection severity was low, moderate, or high risk based on anatomic proximity to the airway and critical structures. Sensitivity, specificity, and positive predictive and negative predictive values of H&P findings to predict moderate- to high-risk infections were calculated and included 95% confidence intervals. Results For the 470 included patients, 389 CT scans were performed, with 220 (56.6%) deemed unnecessary. Unnecessary scans were most prevalent in patients with low-risk infections, in whom 284 CT scans were performed, with 222 (78.2%) deemed unnecessary. There was a strong correlation between red-flag signs and moderate- to high-risk infections. Conclusion There is overuse of CT for odontogenic infections that is most prevalent in low-risk infections without indicative findings in the workup. H&P findings can help accurately diagnose a higher-risk infection before subjecting a patient to CT.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    14
    References
    3
    Citations
    NaN
    KQI
    []