Effects of Adherence to Colposcopy Guidelines on the Detection of Cervical Intraepithelial Neoplasia
Objectives
Practice recommendations for the minimum steps to perform a colposcopy were published in 2017. The goals of this study are to assess whether adherence to these guidelines correlates with detection of cervical intraepithelial neoplasia and to analyze factors associated with rates of biopsy during colposcopy.
Methods
All colposcopies performed at 2 academic medical centers between 2018–2023 were identified by billing code. Patient demographics, cervical cancer screening, medical history, colposcopy details, and pathology were collected from retrospective medical record review. Adherence to guidelines was determined by the number of criteria met on colposcopy documentation. The outcomes assessed were grade of pathology and rate of biopsy.
Results
In univariable regression analysis, no association between adherence to guidelines and detection of cervical intraepithelial neoplasia 2+ was identified (odds ratio [OR] = 1.11, p = .12). When controlling for race, ethnicity, insurance type, Pap test cytology, and human papillomavirus vaccination status in multivariate analysis, there remained no statistically significant correlation (adj. OR = 1.11, p = .14). Black patients (OR = 0.66, p = .039), publicly insured patients (OR = 0.38, p < .001), and patients with HIV (OR = 0.39, p < .001) were each significantly less likely to have a biopsy taken during colposcopies compared to White, privately insured patients, or HIV-negative patients. In multivariate analysis, HIV and public insurance remained factors associated with lower rates of biopsy (adj. OR = 0.46, p < .001 and adj. OR = 0.63, p = .035, respectively).
Conclusions
The study found insufficient evidence to support current colposcopy guidelines. The analysis revealed disparities in the execution of biopsies during colposcopies, particularly in Black, publicly insured, and HIV-positive patients, underscoring the importance of standardization of colposcopic practices.