Objective: To examine a cohort of women with positive cervical smears, but negative colposcopy, in order to ascertain whether there is a subsequent difference in the incidence of squamous dyskaryosis and cervical intraepithelial neoplasia when compared with a control group.
Design: Prospective follow up study.
Setting: Colposcopy clinics, antenatal clinics, GP surgeries.
Methods: A study group of 255 women with reported abnormal cervical smears but negative colposcopy was subdivided into three groups according to referral smears suggesting high grade dyskaryosis (n = 34), mild dyskaryosis (n = 120) and borderline changes (n = 101). They were followed for at least five years and were compared with a control group of 726 women followed up after a negative smear, using the first and worst follow up smears over a five year recall period.
Main outcome measures: Incidence of subsequent cervical cytological and histological abnormalities.
Results: The control group had a similar incidence of squamous dyskaryosis as that expected in the screening population. Forty-six per cent of the study group with colposcopically unconfirmed ('false positive') cervical smears subsequently had abnormal smears. When the three groups were compared with controls using a chi2 test, their incidence of abnormal smears was significantly increased. Cervical intrepithelial neoplasia was found in 19% of the study group, and in 3% of the control group (P < 0.0001).
Conclusions: The analysis demonstrates that women with so-called 'false positive' smears defined by negative colposcopy have an increased risk of subsequent abnormal smears and cervical intrepithelial neoplasia, suggesting that lesions may have been missed on colposcopy. However, in a significant proportion of women, further abnormalities were not detected during the follow up period, indicating that there may be other causes for positive smears and negative colposcopy.