The quality of community colposcopic practice.

Abstract

To estimate the quality of community colposcopic practice in British Columbia through an assessment of the degree of correlation between colposcopy, cytology, and histology. We reviewed all new-patient colposcopies in British Columbia during 2001 by 37 gynecologists in 24 hospital-based clinics. Colposcopic impression closely mirrored the referral cytology diagnosis in 89.8% of cases. As with cytology-biopsy comparisons, discordant cases were more likely to be overestimates of disease rather than underestimates, 18.8% versus 1.8%. Overestimates were usually biopsy sampling errors rather than false positive cytology. The overall correlation between cytology and biopsy was considered satisfactory in 79.4% of cases. Satisfactory agreement between the colposcopic diagnosis and accompanying biopsies occurred in 86.8% of patients. Five colposcopists had performance scores below this standard. Colposcopy with a sensitivity of 90.3% and a specificity of 57.3% as practiced in this provincial program would appear to be of a satisfactory level. The rate of intraepithelial or invasive disease increased from 40.6% in patients with low-grade squamous intraepithelial changes to 91.9% in patients with suspicious or malignant cytology. The value of the colposcopic impression to identify disease correlated best with the higher the grade of disease predicted (64.6% to 92.6%). A measure of the colposcopic proficiency in the community can be estimated by comparing the level of agreement between the presenting cytology, colposcopic impression, and corresponding directed biopsies. The results of this study would indicate that 5 individuals had practice standards that were below average. An integrated cytology-colposcopy program facilitates the assessment and identification of below-average practice standards in a community.