Improving sensitivity of cervical cytology by removal of cervical secretions before sampling: a prospective study in Mexico.

UNLABELLED Sensitivity of cervical cytology is suboptimal, especially in developing countries such as Mexico, despite available guidelines aimed at improving this. When obtaining cervical samples, whether the samples are taken from the transformation zone and whether abnormal cells are missing must be considered. Cervical secretions (CS) are always present in variable proportions, and when cleaning the cervix, better samples may be obtained. In this study, we analyzed samples obtained with or without cleaning the cervix, and compared their contents in order to determine the sensitivity and specificity of these two methods. METHODS Of 500 patients who underwent cytology and colposcopy, 271 (54.2%) required a second opinion due to a diagnosis of cervical intraepithelial neoplasia (CIN). CS was removed and compared with the clean, second sample (SS) using in both liquid-based cytology. The quality of samples according to the Bethesda System, the presence of CIN, and inflammatory reactions were recorded. The sensitivity and specificity were calculated using biopsy as the gold standard. RESULTS The SS resulted in a higher proportion of adequate samples being obtained (97.6% vs. 44.8%), and in increased sensitivity (88.2% vs. 58.8%). CIN was detected in the SS 26% more often than in the CS (34 vs. 27 samples), whereas inflammatory reactions were noted more often in the CS (91.4% vs. 74%). CONCLUSION Cervical sampling including CS results in lower sensitivity and CIN detection rates, and in more inflammatory reactions. By excluding CS from cervical samples, the sensitivity could be improved and the false negative rate could be reduced.