Diagnostic performance of nasal cytology.

Purpose: Nasal pathologies are characterized by a symptomatology that hardly allows to distinguish allergic rhinitis (AR), non-allergic rhinitis (NAR), and chronic rhinosinusitis (CRS). Nasal cytology (NC) has shown increasing importance in helping the clinician to differentiate the various phenotypes of rhinitis. NC allows us to evaluate nasal cellularity by distinguishing AR and various types of NAR. The objective of the study is to assess the diagnostic performance of the NC by evaluating its sensitivity, specificity, and predictive value.

Methods: We recruited 387 patients with persistent rhinitis symptoms, and nasal cytology was performed. The rhinocytogram was obtained by reading for fields and the cellular count was made using quantitative and semi-quantitative grading together.

Results: Two hundred and fifteen patients (55.5%; 38 had acute rhinitis, 24 acute sinusitis, 153 chronic rhinosinusitis) out of 387 referred nasal symptoms. Cytological specimen showed a mean of 94 ± 4% ciliated cells, 29 ± 0.2% mucinous cells, 16 ± 0.1% neutrophils, 11 ± 0.08% eosinophils, 4 ± 0.03 lymphocytes, 4 ± 0.03% mast cells, and 4 ± 0.01% other cells. NC was positive in 271 cases (70%). After revision of medical history, 153 patients (39%) were considered positive for NAR. Test sensibility was 100% (95% CI 97-100), specificity was 49.6% (95% CI 43-56%). Positive predictive value (PPV) was 56% (95% CI 50-62%), and negative predictive value (NPV) was 100% (95% CI 96-100%). The positive likelihood ratio was 1.98 (95% CI 1.75-2.25). Accuracy of the test was 69.5% (95% CI 64.6-74.0%).

Conclusion: Our data showed ability to identify the true-positive patients with NAR but a low ability to identify the true-negative patients, with a global accuracy of 69.5%.