Symptoms predictive of Fusobacterium necrophorum pharyngotonsillitis – an observational study of cases presenting to hospitals in Southern Sweden

Abstract

Objectives Fusobacterium necrophorum is a common cause of pharyngotonsillitis. However, no guidelines exist on when to diagnose or treat it. We aimed to investigate associations between clinical criteria and F. necrophorum-positivity in pharyngotonsillitis and assess the predictive potential of a simple scoring system. Methods Pharyngotonsillitis patients who were tested for F. necrophorum (PCR) and presented to hospitals in the Skåne Region, Sweden, between 2013–2020 were eligible. Data were retrieved from electronic chart reviews and registries. By logistic regression we investigated associations between F. necrophorum-positivity and pre-specified criteria: age 13–30 years, symptom duration ≤ 3 days, absence of viral symptoms (e.g. cough, coryza), fever, tonsillar swelling/exudate, lymphadenopathy and CRP ≥ 50 mg/L. In secondary analyses, associated variables were weighted by strength of association into a score and its predictive accuracy of F. necrophorum was assessed. Results Among 561 cases included, 184 (33%) had F. necrophorum, which was associated with the following criteria: age 13–30, symptom duration ≤ 3 days, absence of viral symptoms, tonsillar swelling/exudate and CRP ≥ 50 mg/L. Age 13–30 had the strongest association (OR5.7 95%CI 3.7–8.8). After weighting, these five variables had a sensitivity and specificity of 68% and 71% respectively to predict F. necrophorum-positivity at the proposed cut-off. Conclusion Our results suggest that F. necrophorum cases presenting to hospitals might be better distinguished from other pharyngotonsillitis cases by a simple scoring system presented, with age 13–30 being the strongest predictor for F. necrophorum. Prospective studies, involving primary care settings, are needed to evaluate generalisability of findings beyond cases presenting to hospitals.