Interobserver Variability Of Visual Inspection Methods Between Physician And Nurse For Cervical Cancer Screening: A Cross Sectional Study

Abstract

Objective: To evaluate and compare predictive value of visual inspection methods and assess the concordance of results between physician and nurse for cervical cancer screening as compared to colposcopy and histopathology. Study design: Cross sectional study Study place: Colposcopy Clinic at Gynaecology Outpatient department ( OPD ) , K.L.E. University Teaching Hospital, Belgaum Source of data: All sexually active women from age 20 to 50 years attending Colposcopy Clinic with complaints of persistent vaginal discharge, inter menstrual bleeding, post coital bleeding, suspicious looking cervix and positive screen test. Study interventions: VIA and VILI by both physician and nurse and the findings were interpreted independently. This was followed by colposcopy done by a gynaecologist blinded to the results of VIA and VILI and directed biopsy was taken if indicated. Negative colposcopy and histopathology were considered as the reference standard. Test positivity was defined by histopathology report of cervical intraepithelial neoplasia 2 ( CIN 2) and above. Results: VIA by doctor had a higher sensitivity ( 100% versus 88.23% ) and higher specificity ( 57.5% versus 56.2% ) when compared with the nurse. There was almost perfect agreement between their VIA findings ( kappa=0.83) . VILI findings of doctor had a sensitivity and specificity of 82.3% and 59.65% respectively, while nurse had a sensitivity of 76.47% and specificity of 58.36%. IV There was almost perfect agreement ( kappa=0.87) between VILI by the doctor and nurse. Conclusion: Visual inspection can be performed reliably by trained nurses and doctors and is an effective screening option in low resource settings. Intensive training and periodic reinforcement sessions are needed so as to reduce the false positive results.