Diagnostic Accuracy of Colposcopy in Cervical Intraepithelial Neoplasia and Its Influencing Factors: A Retrospective Study

Abstract

Objective: To evaluate the diagnostic accuracy of colposcopy for cervical intraepithelial neoplasia (CIN) and identify its influencing factors. Methods: This retrospective study analyzed 493 cases where colposcopy and guided biopsy were performed, with pathologic confirmation serving as the gold standard. We assessed the sensitivity, specificity, positive predictive value, and negative predictive value of colposcopy, as well as the concordance between colposcopy and biopsy findings. Factors such as HPV types, TCT results, transformation zone types, and cervical lesion area were examined for their impact on colposcopy accuracy. Logistic regression was employed to identify factors significantly affecting diagnostic accuracy. Results: Using pathological diagnosis as the benchmark, colposcopy exhibited a sensitivity of 78.11%, specificity of 81.06%, positive predictive value of 91.86%, and negative predictive value of 57.53% in the 493 evaluated cases (Kappa value 0.525, P < 0.001). The exact agreement between colposcopy and biopsy diagnoses was 70.79% (349/493). Univariate analysis indicated that HPV subtype, transformation zone type, age group, and lesion size influenced the diagnostic accuracy of colposcopy. Logistic regression revealed that types II and III transformation zones were independent risk factors for inaccurate diagnosis of cervical squamous intraepithelial lesions via colposcopy. Conclusion: Colposcopy, while useful, has limitations in diagnosing CIN. Clinical judgment, supplemented by random biopsy, is essential to minimize diagnostic omissions. Particular attention should be given to patients with unsatisfactory colposcopy results, combining clinical signs with diagnostic tools to enhance the accuracy of colposcopy.