Clinical relevance of objectifying colposcopy

Abstract

Purpose: To evaluate the clinical value of four objective colposcopic criteria inner border, ridge sign, cuffed crypt openings and rag sign to diagnose cervical intraepithelial neoplasia grade 2 or worse (CIN 2+), using video exoscopy and to compare it to subjective graduating signs. Methods: Retrospective evaluation of video recordings of 444 patients, referred for diagnostic colposcopy, who underwent cervical biopsies, and if indicated loop excisions. Most severe histological diagnosis was recorded. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and likelihood ratios (LR) with 95 % confidence interval, for CIN 2+ were calculated. Results: Single biopsy, two biopsies and magnification-guided loop excision were performed in 60.8, 39.2 and 70.5 % of patients, respectively. Sensitivity, specificity, PPV and NPV to detect CIN 2+ were 19.3, 99.2, 98.3 and 35.8 %, for inner border sign; 53.1, 93.5, 94.7 and 47.6 %, for ridge sign; 51.5, 84.9, 88.2, and 44.3 %, for cuffed crypt openings, and 40.7, 96.4, 96.1 and 42.5 %, for rag sign, respectively. The positive likelihood ratio (LR+) was 26.7 and the negative likelihood ratio (LR−) was 0.81, for inner border sign; 8.2 and 0.5, for ridge sign; 3.41 and 0.57 for cuffed crypt openings; and 11.3 and 0.62 for rag sign, respectively. 90 % of CIN 2+ had at least one objective sign. Combination of any two objective signs significantly increased the LR of the presence of CIN 2+, and was clinically superior to any combination of graduating signs. Conclusion: Objective colposcopic criteria are clinically useful and significantly associated with CIN 2+.