Background: Cervical cancer remains a leading cause of cancer-related mortality among women in developing countries. Early detection through effective screening is crucial for reducing morbidity and mortality. This study aimed to compare the diagnostic performance of Pap smear and colposcopy as screening tools for detecting cervical premalignant and malignant lesions among symptomatic women at a tertiary care centre serving the Sundarban region of West Bengal. Methodology: A hospital-based prospective study was conducted among 110 sexually active women aged 21-50 years presenting with symptoms suggestive of an unhealthy cervix. Participants underwent Pap smear examination followed by colposcopy and colposcopy-guided biopsy. Histopathological examination served as the gold standard for diagnosis. Sensitivity, specificity, positive predictive value, negative predictive value, and diagnostic accuracy were calculated for both screening modalities, and agreement was assessed using Cohen’s kappa. Results: The mean age of participants was 33.2±4.1 years, with 41.8% aged 31-40 years. White vaginal discharge was the most common presenting complaint (50.9%), followed by intermenstrual bleeding (33.6%). Histopathological examination revealed premalignant and malignant conditions in 13.6% of participants. Cervical pathology was significantly associated with lower socioeconomic status (p<0.001), early pregnancy (p=0.004), multiple sexual partners (p=0.002), history of STI (p<0.001), and prolonged OCP use (p<0.001). Pap test demonstrated 33.3% sensitivity and 97.9% specificity, while colposcopy showed superior performance with 86.7% sensitivity and 80.0% specificity. Agreement between Pap and colposcopy was weak (κ = 0.284). Conclusion: Colposcopy demonstrates significantly higher sensitivity than Pap smear for detecting cervical lesions among symptomatic women, although the two tests demonstrated only weak concordance. While both methods provide useful information where HPV-based screening is not widely available, combined use should be considered cautiously, given resource constraints.
Sinha et al. (Fri,) studied this question.