ABSTRACT
BACKGROUND: The development of a terminology system is essential to allow uniformity in reporting serous fluid specimens. An important topic to cover is the issue of specimen adequacy. In the present study, we aimed to evaluate whether there is a correlation between number of mesothelial cells and overall improved sensitivity and adequacy control of tests. METHODS: Cases of negative pleural fluids with concomitant positive pleural biopsies were selected from two referral institutions, with observation of the number of mesothelial cells in 10 high-power fields, comparing the results with a control group (cases with negative biopsies, ie, true negatives). Comparisons were conducted using the nonparametric Mann-Whitney U test. Data were analysed for sensitivity and specificity derived from the receiver operating characteristics curve. For the choice of an optimal cut-off of mesothelial cells, receiver operating curve analysis was constructed and the Youden index was calculated. RESULTS: A total of 112 pleural effusions with paired pleural biopsies were studied. There was no difference in distributions of the number of mesothelial cells between cases with a positive biopsy (false negatives) and the control group (median = 39 vs median = 30, respectively, P-value = .974). However, simple logistic regression found a cut-off of 750 cells per 10 high-power fields as an optimal number for improved sensitivity (72.7%), with fair discriminatory power. CONCLUSIONS: Enumeration of mesothelial cells may improve the sensitivity of the cytological diagnosis of malignant pleural effusion, serving as an internal quality control for the test's overall accuracy.
