Cervical cytology: a randomized comparison of four sampling methods.

Abstract

The purpose of this study was to compare smear quality and endocervical cell recovery of four cervical smear sampling devices. Two thousand fifteen patients undergoing routine cervical smears at the University of Tennessee Obstetrics and Gynecology clinics were randomly assigned to a cotton swab-spatula, Cytobrush-spatula, Cervex-Brush, or Bayne Pap Brush. The cytopathology laboratory, blind to method, used specific criteria to grade smears as being optimal, adequate, marginal, or inadequate. Statistical analysis was by the chi 2 and analysis of variance tests. No statistical differences occurred among the groups for nonpregnant patients. For pregnant patients smear quality was improved with both Cytobrush-spatula and Bayne Pap Brush versus cotton swab-spatula (p = 0.0301 and 0.0004, respectively); cotton swab-spatula had fewer endocervical cells than the Cytobrush-spatula (p = 0.0001), Cervex-Brush (p = 0.0288), and Bayne Pap Brush (p = 0.0081). The cotton swab-spatula and Cytobrush-spatula appear to be the most effective screening methods for nonpregnant and pregnant patients, respectively.