Abstract
Introduction. This study aims to investigate the distribution of high-risk HPV genotypes in different cervical pathology and identify the diagnostic agreement between colposcopy and cervical biopsy in Kazakhstan. Methods. This is a cross-sectional cohort study that was conducted in a tertiary care hospital in Astana, Kazakhstan from November 2024 to March 2026. Women of reproductive age aged 18–45 years underwent HPV testing, a video-colposcopy and colposcopy-guided cervical biopsy. Results. The study included 483 participants who were not previously vaccinated from HPV. Colposcopy was done for 285 women. 48.5% of participants had normal colposcopic impression, 40.1% had abnormal grade I impression, and 10.6% had abnormal grade II impression. Among these 285 patients, cervical biopsy results were available for 127 participants and the distribution of biopsy results varied significantly among three colposcopic impressions. The agreement percentage between colposcopic impressions and biopsy results is 37.5%, unweighted Cohen's κ = 0.083 (bootstrap CI -0.097-0.269) indicating weak and the quadratic-weighted κ = 0.342 (bootstrap CI [0.178, 0.477]) indicating a moderate degree of agreement. Conclusion. Although discordance was present between colposcopy and cervical pathology, our results demonstrated that colposcopy has potential to increased diagnostic value in detection of high grade cervical changes. Key words: cervical precancerous lesions, colposcopy, human papillomavirus, cervical biopsy, cervical cancer.