A randomized control trial of two cervical spatulas.

The sampling efficacy of the Ayre and Rocket cervical spatulas was compared in a randomized controlled trial involving 533 women who were attending a colposcopy clinic either for investigation of cytological abnormality or for review after treatment of CIN. Smears were compared with respect to the presence of cytological abnormality and two types of normal cervical epithelial cells (endocervical and immature metaplastic cells) which are considered indicators of adequate cytological sampling. Smears taken with the Rocket spatula were significantly less likely to contain immature metaplastic cells (95% CI on difference in proportion; 5-30%), but more likely to contain endocervical cells (95% CI 16-36%). No significant difference was found in the proportion of smears containing one or other type of indicator cell or in the yield of abnormal smears. It is concluded that the Rocket spatula offered no advantage when used in these circumstances. This study illustrates how spurious conclusions of sampling efficacy can be drawn if only one type of indicator cell is used as an endpoint.