Computerized planimetry versus clinical assessment for the measurement of cervical ectopia

OBJECTIVE: As part of a study to assess the role of cervical ectopia in the acquisition of cervical infections, we determined the reliability of cervical ectopia measurements made by computer planimetry and by clinical (visual) assessment. STUDY DESIGN: We conducted pelvic examinations of 1004 women seeking contraceptive services at two health centers in Baltimore. After application of acetic acid, clinicians estimated the relative area of ectopia by visual inspection and took cervical photographs. Two independent raters measured the absolute and relative areas of ectopia from the digitized images by means of an analytic software program. Agreement levels between raters, between multiple readings by the same rater, and between the two measurement methods were quantified by means of the intraclass correlation coefficient and weighted κ. RESULTS: Intrarater agreement was excellent for computer planimetry measurements of the absolute (intraclass correlation coefficient, 0.97) and relative (intraclass correlation coefficient, 0.89) areas of ectopia. Interrater agreement was also high for computer planimetry measurements of the absolute (intraclass correlation coefficient, 0.83) and relative (intraclass correlation coefficient, 0.85) areas of ectopia. Agreement levels were moderate between clinician assessment and computer planimetry measurements of the relative area of ectopia (κ = 0.48), but agreement was better when clinical assessment was limited to observations by a single, experienced clinician. CONCLUSION: Measurement of cervical ectopia by computer planimetry was highly reliable and appears appropriate for assessment of the role of ectopia in the acquisition of cervical infections. Clinical assessment of cervical ectopia may be used when computer planimetry is not available.