Acceptability and Concordance of Self- Versus Clinician- Sampling for HPV Testing among Rural South Indian Women

Despite being largely preventable, HPV-related cervical cancers continues to be the second highest cause of cancer deaths among Indian women. HPV testing using self-sampled samples may offer an opportunity to expand cervical cancer screening in India where there is currently a shortage of providers and facilities for cervical cancer screening. The study examines acceptability and concordance of self vs. clinician collected samples for HPV-relted cervical cancer screening among rural South Indian women. Between May and August 2017, eight mobile screening clinics were conducted among 120 eligible women in rural villages in Mysore District, India. Women over the age of 30 underwent informed consent process and then self-sampled a sample for cervicovaginal HPV DNA testing. Next, the women underwent clinical exam where the clinician collected a cervicovaginal HPV DNA sample. Following the clinical exam, all participants answered an interviewer-administered questionnaire to assess their history of cervical cancer screening and acceptability of self- and clinician-sampling methods. To assess diagnostic accuracy, concordance of self- and clinician-sampled HPV DNA specimens was calculated in addition to five measures of acceptability (feeling of caring, privacy, embarrassment, genital discomfort, and genital pain). Study participants had a median age 39 years, about four-in-ten (41.7%) had a secondary education or above, the vast majority (87.5%) were married and only 3.4% reported having screened for cervical cancer. For all measures of participant acceptability, self-sampling was rated significantly higher than clinician-sampling. Cohen's kappa was 0.73 (95% CI: 0.34, 1.00), indicating substantial agreement between self- and clinician-sampling. This study demonstrates that HPV self-sampling for cervical cancer screening is feasible and acceptable in a community setting among South Indian rural women. Concordance between self-sampling and clinician-sampling was adequate for screening in community settings. .