Comparison of conventional Papanicolaou smears and a fluid-based, thin-layer system for cervical cancer screening

Abstract

To compare the cytologic diagnoses and specimen adequacy of a fluid-based, thin-layer preparation and conventional Papanicolaou tests. A total of 7360 women from six separate sites-three screening centers and three hospitals-participated in split-sample/matched-pair, double-masked clinical trials. A conventional Papanicolaou test was performed, after which residual material on the sampling device was rinsed into a fluid preservative from which a thin-layer slide (ThinPrep) was prepared using the ThinPrep 2000 automated slide processor (Cytyc Corp., Boxborough, MA). Conventional and ThinPrep slides were read independently. Cytologic diagnoses and specimen adequacy were classified using the Bethesda system. For the three screening centers, 65% more diagnoses of low-grade squamous intraepithelial lesions and higher were made on the ThinPrep slides (P < .001); for the three hospital centers, 6% more of these diagnoses were made with the ThinPrep method (P = .294). Comparison of specimen adequacy showed a significant increase in satisfactory specimens and a reduction in the number of "satisfactory but limited by" specimens using the ThinPrep method (P < .001). The fluid-based, ThinPrep method of sample preparation resulted in a statistically significant increase in cytologic diagnosis of cervical cancer precursors and in specimen adequacy compared with the conventional Papanicolaou testing method.