Measuring cervical ectopy: direct visual assessment versus computerized planimetry.

Abstract

OBJECTIVE: Cervical ectopy has been identified as a possible risk factor for heterosexual transmission of human immunodeficiency virus. To accurately assess the importance of cervical ectopy, methods for measuring ectopy with precision need to be developed. The objective of this study was to evaluate the reliability of two methods of measuring cervical ectopy: direct visual assessment and computerized planimetry. STUDY DESIGN: Cervical photographs of 85 women without cervical disease were assessed for cervical ectopy by three raters using direct visual assessment and a computer planimetry method. Agreement between the two methods, among the three raters, and among measurements by each rater over time was calculated with use of intraclass correlation coefficients, where 1.0 represents perfect agreement and 0 represents no agreement except by chance. RESULTS: The intraclass correlation coefficient among the three raters (interrater agreement) was 0.58 for direct visual assessment without application of acetic acid to the cervix compared with 0.72 for direct visual assessment with acetic acid and 0.82 for computerized planimetry with acetic acid. The intraclass correlation coefficient among measurements by each rater over time (intrarater agreement) was 0.66 for direct visual assessment without acetic acid compared with 0.77 for direct visual assessment and 0.83 for computerized planimetry after application of acetic acid. When acetic acid was used, the intraclass correlation coefficient between the two methods was 0.69. CONCLUSIONS: Computerized planimetry of cervical photographs may provide the most consistent estimate of the percent of ectopy. However, if time and resources make the use of computer planimetry difficult, direct visual assessment after application of 5% acetic acid appears to provide comparable estimates.