Optimal positive cutoff points for careHPV testing of clinician- and self-collected specimens in primary cervical cancer screening: an analysis from rural China

careHPV, a lower-cost DNA test for human papillomavirus (HPV), is being considered for cervical cancer screening in low- and middle-income countries. However, not a single large-scaled study exists to investigate the optimal positive cutoff point of careHPV test. We pooled data for 9,785 women participating in two individual studies conducted from 2007 to 2011 in rural China. Woman underwent multiple screening tests, including careHPV on clinician-collected specimens (careHPV-C) and self-collected specimens (careHPV-S), and Hybrid Capture 2 on clinician-collected specimens (HC2-C) as a reference standard. The primary endpoint was cervical intraepithelial neoplasia grade 3 or more severe (CIN3+) (n = 127), and secondary endpoint was CIN2+ (n = 213). The area under the curves (AUCs) for HC2-C and careHPV-C were similar (0.954 versus 0.948, P = 0.166), and better than careHPV-S (0.878; P < 0.001 versus both). The optimal positive cutoff points for HC2-C, careHPV-C, and careHPV-S were 1.40, 1.74, and 0.85, respectively. At the same cutoff point, careHPV-C was not significantly less sensitive and more specific for CIN3+ than HC2-C, and careHPV-S was significantly less sensitive for CIN3+ than careHPV-C and HC2-C. Raising the cutoff point of careHPV-C from 1.0 to 2.0 could result in nonsignificantly lower sensitivity but significantly higher specificity. Similar results were observed using CIN2+ endpoint. careHPV using either clinician- or self-collected specimens performed well in detecting cervical precancer and cancer. We found that the optimal cutoff points of careHPV were 2.0 on clinician-collected specimens and 1.0 on self-collected specimens.