Predicting bacteria causing acute bacterial rhinosinusitis by clinical features

Introduction

Clinicians rely on clinical presentations to select therapeutic agents for acute bacterial rhinosinusitis. Streptococcus pneumoniae and Haemophilus influenzae are common in acute bacterial rhinosinusitis. Drug resistant Streptococcus pneumoniae and Haemophilus influenzae require different antibiotics.

Objective

This study aimed to evaluate the associations between clinical features of acute bacterial rhinosinusitis and pathogenic bacteria.

Methods

Sixty-four patients with acute bacterial rhinosinusitis were enrolled. Clinical features including nasal obstruction, discolored discharge, facial pain, smell disturbance, fever and laboratory findings of patients with acute bacterial rhinosinusitis were collected. The bacterial cultures of endoscopic middle meatal swabs were used as a reference.

Results

Serum C-reactive protein level elevation correlated with the bacterial species (p = 0.03), by which was increased in 80.0% of Haemophilus influenzae rhinosinusitis and 57.1% of Streptococcus pneumoniae rhinosinusitis. The elevated C-reactive protein was the significant predictor for Haemophilus influenzae rhinosinusitis with the Odds Ratio of 18.06 (95% CI 2.36–138.20). The sensitivity of serum C-reactive protein level elevation for diagnosing Haemophilus influenzae rhinosinusitis was 0.80 (95% CI 0.49–0.94).

Conclusion

Elevation of serum C-reactive protein level was associated with and predicted acute bacterial rhinosinusitis caused by Haemophilus influenzae.

EPOS2012 has better specificity compared to IDSA2012 for diagnosing acute bacterial rhinosinusitis

BACKGROUND

Acute bacterial rhinosinusitis (ABRS) is a subtype of acute rhinosinusitis (ARS). To prevent excessive antibiotic prescribing, clinical criteria for diagnosing ABRS are presented in two major international guidelines from European Position Paper on Rhinosinusitis and Nasal Polyps (EPOS2012) and the Infectious Diseases Society of America (IDSA2012). This study aims to assess accuracy of these criteria.

METHODOLOGY

Patients with ARS were recruited. Clinical features were collected including discolored nasal discharge, facial pain, fever, double sickening, symptoms persisting longer than 10 days, and elevated serum C reactive protein (CRP) and erythrocyte sedimentation rate (ERS). Using middle meatal bacterial culture as a reference, accuracy of EPOS2012 and IDSA2012 criteria were analyzed.

RESULTS

Eighty-eight patients (age 43.2+/-14.5 years, 67% female) with ARS were recruited. Using the two criteria for diagnosing ABRS, EPOS2012 and IDSA2012 have sensitivity of 50% (95%CI: 38%-62%) versus 69% (95%CI: 57%-79%), specificity of 63% (95%CI: 43%-79%) versus 46% (95%CI: 28%-65%), and accuracy of 53% versus 63%, respectively.

CONCLUSION

Both EPOS2012 and IDSA2012 had modest accuracy. EPOS2012 had less sensitivity but a better specificity compared to IDSA2012. This suggests that IDSA2012 diagnostic criteria may contribute to inappropriate use of antibiotics due to poorer specificity.