ObjectiveTo establish a vaginal self-sampling HPV cervical cancer screening model in Hainan Province, to analyze the application of p16 protein detection in HPV positive and non-HPV16 /18 shunt screening. MethodsFrom January 2019 to September 2022, a total of 200 women from the targeted population was randomly selected for vaginal self-sampling HPV typing test to screen cervical cancer using randomized numeric table method, followed by cervical cytology sampling for cytology p16 protein detection. Postoperative pathological examination was used as the gold standard. Multivariate logistic regression analysis was used to analyze the influencing factors of HPV positive detection rate in cervical lesions, and the nomogram model was constructed simultaneously. The receiver operating characteristic(ROC) curve and calibration curve were used for evaluating the accuracy of the nomogram model. Differences in the distribution of self-sampled HPV-positive and HPV infected genotypes were recorded, and the application of p16 protein detection in HPV-positive and non-HPV16/18 shunt screening was analyzed. ResultsAged ≥40 years, BMI ≥28.00 kg·m^-2, number of sexual partners ≥2, frequency of sexual life ≥10 times·month^-1, bleeding from sexual intercourse, and age of first sexual intercourse <22 years were the risk factors for HPV positive of cervical lesions (all P<0.001). The results of ROC curve and calibration curve showed that the area under ROC curve (AUC) was 0.874 (95%CI: 0.823‒0.907, P<0.05), the sensitivity was 0.835, the specificity was 0.847, and the Youden index was 0.672, indicating a good fit of the model. Results of vaginal self-sampling HPV test showed that the positive rate of HPV was 86.50% (173/200). HPV high-risk infection types mainly included HPV16, 18, 31, 33, 35, 39, 45, 52, 56, 58, 59, 68, 73, and 82. Single HPV infection accounted for 95.95% (166/173), 2.89% (5/173) were infected with two types of HPV, and 1.16% (2/173) were infected with three or more types of HPV. Colposcopic pathologic diagnosis was used as the gold standard, and the results showed that the accuracy of p16 protein detection in the diagnosis of cervical cancer was 93.50% (187/200), with a sensitivity of 96.53% (167/173), and a specificity of 74.07% (20/27). The negative and positive predictive value were 76.92% (20/26) and 95.98% (167/174), respectively. The results of shunt screening showed that there were 80 cases infected with HPV16, 79 cases infected with HPV18 and 41 cases of non-HPV16/18, with a sensitivity of 90.91%, 90.32% and 86.67%, a specificity of 71.43%, 64.71% and 72.73%, a negative predictive value of 62.50%, 64.71% and 66.67%, a positive predictive value of 93.75%, 90.32% and 89.66%, and an accuracy of 87.50%, 84.81% and 82.93%, respectively. The specificity and accuracy of p16 positive screening for cervical cancer were significantly higher than that of HPV positive detection, but the false positive rate was significantly lower than that of HPV positive detection. The AUCs of HPV positive, p16 positive and combination of the two detection methods for cervical cancer were 0.603, 0.822 and 0.907, respectively. ConclusionVaginal self-sampling HPV testing is a widely accepted mode for cervical cancer screening. Cervical cytology p16 protein detection is important for self-sampled HPV positive and shunt screening of non-HPV16/18.