An Analysis of 84,244 Patients from the British Colombia Cytology–Colposcopy Program

This paper presents the data from the integrated provincial screening model (cytology, colposcopy, biopsy) used in British Columbia during the years 1986 to 2000. The British Columbia Cancer Agency Cervical Screening Program was used for cytologic reporting. This system is closely derived from one instituted by the British Society of Clinical Cytology in 1986 and is similar to the Bethesda System 2001. All patients (n = 84,244) who had their primary diagnostic evaluation for cervical abnormalities (96%) or suspicious clinical findings (4%) in the 15 years of the study period were included for analysis. Thirty percent of women were 25 years of age or younger, whereas 41% were over the age of 35 years. Cytologic diagnoses of both minor lesions and malignant disease have been consistent throughout the study period. Diagnoses of mild or moderate squamous dyskaryosis, the most common cytologic finding, increased from 20% and 27.4% in 1986 to 35.5% and 40.2% in 2001, respectively. More severe findings, marked or suspicious squamous dyskaryosis, decreased in frequency from 32% and 15.8% to 14.7% and 3.3%, respectively. The data from the final 3 years of this analysis reveal the most pronounced change, with only 4.2% and 0.9%, respectively, of cytologic findings showing either mild or moderate and higher dyskaryosis. Within one grade of disease, the referring cytologic diagnosis and the colposcopistʼs impression of pathology were in agreement for 97.8% of all patients. Among patients whose pathology showed mild or reactive changes, 1166 (4.5%) had colposcopic impressions of more severe lesions, including 23 (0.08%) with changes suggestive of invasive cancer. In patients whose cervical smears were diagnosed as moderate dyskaryosis or worse, 71% had colposcopic findings in agreement with the cytologic diagnosis, 20.3% had colposcopic impressions of low-grade disease, and 8.7% had benign or nondysplastic changes. Agreement between colposcopy and cytology, within one grade of disease, was statistically significant (P < 0.001). Similarly, there was significant agreement between histology and cytology, within one grade of disease (82%, P < 0.0001). Cervical biopsy diagnoses were worse than cytology in 2.3% of patients and less dysplastic in 16.1%. Among patients with mild or reactive changes diagnosed in the cervical smear, 2.3% had biopsy-proven diagnoses of more severe disease, including 9 women with invasive cancer. When each cytological category was analyzed individ