Comparison of fluid-based, thin-layer processing and conventional Papanicolaou methods for uterine cervical cytology

This study was conducted to compare the diagnoses and specimen adequacy of the ThinPrep (Cytyc Corporation, Boxborough, MA, USA) and conventional Papanicolaou preparation methods for cervical cytology. Cervical specimens from 972 patients from our gynecology clinic were analyzed. A single sample of the uterine cervix taken from each patient was first used to prepare the conventional Papanicolaou smear. Afterward, the residual tissue on the sampling device was rinsed into a fluid preservative from which two slides were prepared by means of the ThinPrep method. Conventional and ThinPrep slides were examined independently. Cytologic diagnoses and specimen adequacy were classified according to the Bethesda System. The diagnoses based on the ThinPrep and conventional smears matched in 939 (96.6%) cases. Low-grade squamous epithelial or higher grade lesions were detected with at least one of the methods in 59 cases and with both methods in 42 cases. Fifteen more cases of low-grade squamous epithelial or higher grade lesions (58 vs 43; p < 0.001) and 10 more cases of high-grade squamous epithelial lesions or carcinoma (42 vs 32; p = 0.006), were detected with ThinPrep than with conventional smears. The ThinPrep method yielded a higher rate of specimens that were satisfactory for cytologic examination than the conventional smears (89.8% vs 70.9%, p < 0.001). The detection rate of infectious agents was also higher with the ThinPrep method (14.9% vs 6.6%; p < 0.001). The reproducibility rate of the twin slides prepared with the ThinPrep method was greater than 99%. In conclusion, the ThinPrep method yielded a significantly higher positive detection rate of cervical abnormalities and more satisfactory specimens than conventional Papanicolaou smears.