Article Text
Abstract
Introduction Primary HPV screening due to its high sensitivity needs appropriate triage to decrease colposcopy referral and overtreatment. Several triage tools have been recommended for diverse resource settings. The present study was conducted to determine the optimum triage strategies for women with positive HPV screening.
Methods This prospective study included women aged 30-65 years referred for colposcopy in view of HPV positivity. All women underwent triage with VIA, HPV 16/18 genotyping (Cobas, Roche), colposcopy with Swede scoring (SS). Outcome of triage with each modality was analysed.
Results 209 HPV positive women were recruited. The common presentations were postcoital bleeding(11.0%), intermenstrual bleeding(9.6%), postmenopausal bleeding(5.7%), persistent vaginal discharge(65.1%) and unhealthy cervix 8.1%). Test positivity rates, and corresponding referral rates, were 76.0%, 26.0%, 43.1% and 16.7% on VIA, HPV 16/18 genotyping and colposcopy with SS ≥5 and ≥8, respectively. On histopathology, 40(19.1%) women were detected with CIN2+(CIN2 2.8%; CIN3 10.0%; invasive cancer 6.2%). False positive rates were 77.9%, 55.1%, 65.6% and 40.0% respectively. False negative rates were 10.0%, 6.4%, 7.5% and 10.9% respectively. The test characteristics of VIA, HPV genotyping and colposcopy SS ≥5 and ≥8 included sensitivity of 87.5%, 70.9%, 77.5% and 52.5%; specificity: 26.6%, 82.8%, 65.1% and 91.7%; positive predictive value (PPV) 22.0%, 49.0%, 34.4% and 60.0%; and negative predictive value (NPV) 90.0%, 93.5%, 92.4%, and 84.2% respectively.
Conclusion/Implications HPV 16/18 genotyping had highest NPV. As a built-in triage, it is easily implementable. Transition to HPV testing will improve efficiency of programs even in low resource settings.