Improving the sensitivity of cervical cytologic screening. A comparison of duplicate smears and colposcopic examination of patients with cytologic inflammatory epithelial changes.

During a ten-month period, 264 cervical cytologic specimens were submitted in duplicate to two separate cytology laboratories. An attempt was made to perform colposcopy on all 45 patients reported as having an abnormality by either laboratory. All but one patient with a cytologic diagnosis of cervical intraepithelial neoplasia (CIN) underwent colposcopy, as did 68% of the patients with a diagnosis of nondysplastic atypia (inflammatory epithelial changes [IEC]). Five cases of histologically verified CIN were found by colposcopic study of patients with a cytologic diagnosis of CIN; two additional cases were found by colposcopic study of patients with a cytologic diagnosis of IEC. On the assumption that patients not colposcoped were not systematically different from the others with IEC, the screening sensitivities for both laboratories and for cytology followed by colposcopy of IEC cases were estimated. A statistically significant improvement in screening sensitivity was achieved by colposcopic examination of patients with IEC. This conclusion was tempered by a Bayesian analysis that suggested that some of the apparent improved sensitivity could be due to falsely positive biopsy reports. Despite potential benefits, it is premature to recommend universal colposcopic examination of patients with cytologic reports of inflammatory epithelial changes.