Evaluation of serial cervical cytology in the assessment of preinvasive cervical neoplasia.

Examination was made of the cervical cytology in 107 patients who underwent cone biopsy with a subsequent diagnosis of severe dysplasia or carcinoma in situ (105) or microinvasive carcinoma (2). Multiple smears had been performed for each patient over a period of time. From the total number of smears a 'false-negative' cervical cytology rate of 10.3 per cent was found. The time interval from the initial abnormal smear until definitive surgical treatment was carried out was calculated. The cervical smear prior to cone biopsy was found to correlate with the histological diagnosis in only 46.7 per cent. The implications of performing repeated cervical cytology are discussed and a firm recommendation is made that patients with abnormal cervical smear be promptly referred to definitive treatment.