Nasal cytologies--impact of sampling method, repeated sampling and interobserver variability

The influence of sampling technique, repeated sampling and the interobserver variability on quantitative analysis of nasal cytologies were explored. Nasal cytologies from 12 healthy and 4 allergic volunteers were repeatedly collected with a total of three different sampling techniques. Cells were demonstrated by May-Grünwald-Giemsa staining and by immunostaining for CD45 and eosinophil peroxidase (EPO). To determine the interobserver variability specimens were evaluated by three observers with various levels of nasal cytology experience. When adjusted for the variability due to sampling method and repeated sampling, the staining index for CD45+ cells was 53.4 +/- 11% (interindividual coefficient of variation: 0.21) with an average intraindividual coefficient of variation of 0.14. For CD45, staining indices differed significantly between the washing techniques and the nasal swabs (p<0.01). Repeated sampling did significantly influence the staining indices of specimens collected with nasal swabs (p<0.05). Subjects with nasal allergy had less CD45+ cells in their nasal cytologies than the non-allergic subjects (p=0.02). A high level of compliance for the quantitative analysis of nasal cytologies between three observers was found (kappa(w) range 0.95-0.98). For studies that require repeated sampling, the lavage will be recommended as best suitable to obtain nasal cytologies.