Diagnostic accuracy of history and physical examination in bacterial acute rhinosinusitis

Abstract

OBJECTIVES/HYPOTHESIS: To evaluate the diagnostic accuracy of symptoms, the symptom progression pattern, and clinical signs in identifying bacterial acute rhinosinusitis (ARS). STUDY DESIGN: We conducted an inception cohort study among 50 military recruits with ARS. METHODS: We collected symptoms daily from the onset of symptoms to approximately 10 days. At 9 to 10 days, standardized data on symptoms and physical findings were gathered. A positive culture of maxillary sinus aspirate was considered to be the reference standard for bacterial ARS. RESULTS: At 9 to 10 days, the presence or deterioration after 5 days of any of the symptoms could not be used to diagnose bacterial ARS. Toothache had an adequate positive likelihood ratio (positive likelihood ratio [LR+] 4.4) but was too rare to be used for screening. In contrast, several physical findings at 9 to 10 days were of more diagnostic use and frequent enough for screening. Moderate or profuse (vs. none/minimal) amount of secretion in nasal passage seen in anterior rhinoscopy satisfactorily either ruled in, if present (LR+ 3.2), or ruled out, if absent (negative likelihood ratio 0.2), bacterial ARS. If any secretion was seen in the posterior pharynx or middle meatus, the probability of bacterial ARS increased markedly (LR+ 5.3 and LR+ 11.0, respectively). CONCLUSION: We found symptoms or their change to be of little use in identifying bacterial ARS. In contrast, we observed several clinical findings after 9 to 10 days of symptoms to predict bacterial ARS quite accurately.