A website http://www.mcareu.com was developed by our team with patented copyright for cervical cancer screening provided by Peking University Shenzhen Hospital (PUSH). The website was supervised under the certificated network security level protection of PUSH. The initial version of the website was developed to include modules for public education, self- registration, self-sampling, and result reporting, which are purposed to provide women with the screening project information and the background information on cervical cancer prevention, to facilitate women to complete screening registration by inputting personal data and sign informed consent form online, to facilitate eligible women to get sampler, collect sample for themselves, and apply logistic for shipment of the samples, and to enable women to check their primary testing results, respectively. The website can be log in on computers or mobile devices. Although accessible to all women in China, the project was only open for the participation of women who lived in Shenzhen.

After developing the website and completing its testing, we organized an online cervical cancer screening project conducted in 2 stages from March 2013 to May 2015 and August 2019 to June 2020, in which two deferent HPV assays and the liquid-based and card-based specimen processing media were adopted respectively in the first and second stages. The primary objectives of the project are to test (1) whether it is reliable to have none-professionally trained women to input their own data (self-data-input), (2) if women can collect qualified vaginal sample for themselves in referring to a graphic sampling guidance without instruction of on-site professionals, and (3) whether women can correctly follow instructions to order sample shipment to the lab. The project was divided into two stages because the first stage was approved to provide evidence for the safety and applicability of the website. The second stage was conducted after the website was officially approved for public services. When the second stage of the project was launched, an online questionnaire on participants’ background information was added to collect data for analysis of (the secondary objectives) (1) the willingness of the participants to respond a questionnaire for background information, and (2) the potential impacts of the background social elements on the online participation, in addition of the primary objectives of the study.

The program was designed to screen women who was eligible and registered successfully, with self-collected HPV testing as the primary screening test, and to biopsy all women who were primarily tested positive of hrHPV and returned for diagnostics. All those procedures were free of charge to eligible women. Anyone who was diagnosed as grade 2 cervical intraepithelial neoplasia and above (CIN2+) were referred to PUSH or other qualified hospitals for treatment. This study was approved by the Ethics Committee of the Peking University Shenzhen Hospital (G2014-1).

With referring the project notification and the background information on cervical cancer prevention that was releases on the website, women who were living in Shenzhen could register for participation in the project by inputting their personal data online. Women would be eligible for participation and successful for registration if they were 25–59 (Stage 1) or 30–59 (Stage 2) years of age, none-pregnant, without history of hysterectomy or pelvic radiative therapy, willing to participate by signing an online version of informed consent form. Women were informed that they could withdrew from their participation at any time during the procedures without any impacts on their further medical cares. Women aged 25–29 years of age were included in the first stage of the project because (1) the screening guideline at that time allowed provision cervical cancers screening for women at that age per their willingness, and (2) we tried to include some younger women to test whether internet facilitation worked equal to women aged 55–59 as to young people aged to 25–29. This age group were not targeted in the second stage when we found no significant difference between age 25–29 and 55–59 in errors from self-input data.

The website was accessible for registration during the program stages. With accessibility control mechanism, only women who were living in Shenzhen can register for participation. This limit was reset in consideration of the feasibility of positive triages. No advertisement on public media was released other than the announcement of the program on the website, and no offline contact such as test message or phone call was given to the community women. With this regard, community women could just get known about the project via searching websites with cervical cancer screening related words and oral communications. In this way, we tried to keep most of the participants be screened per their original willingness based on their recognition on cervical cancer prevention without impacts from information resources other than the website. To keep women’s willingness for participation be based on same level of website information, no offline consultation was available before screening results were reported.

Successful registration for participation needed a woman to visit the website (mcareu.com) to “apply for screening” as an applicant. And the online version of informed consent form needed to be reviewed and signed by applicants by checking the reset statement in Chinese, which meant in English as: “I have read the contents of the informed consent and I agree to participate with full understanding of all the items.” The applicant would then get into a field for her to input her essential demographic information (EDI) required for screening tests. Then the registration module of the system would automatically match the applicant’s information with the eligibility criteria that had been reset into the system. Applicants who were eligible for participation would get a reminder for successful registration immediately after the EDI was input, while ineligible applicants would be informed for unsuccessful registration.

EDI required for registration included the name, State ID, living address, and contact details (cellphone number) of the applicants. Only the applicant who provided full items of the required EDI could be enrolled in the screening program as a participant (the participant). A barcode would be created for each of the participant, which was unique to a specific participant and used for identification of the participant in the whole program procedures. All participants could receive a sampling kit shipped to their registered addresses via regular logistics services.

In the second stage of the program, a questionnaire about the social background were provided to the participants to be completed per their willingness. The questionnaires included several questions including cervical cancer screening history, health insurance status, educational background, monthly family income, age of first sex, number of sexual partners, and smoking history, et al.

When the screening events were launched, a sampling kit was provided to each of the participants, which contains a sampling brush, a piece of self-sampling guidance in graphics, an envelope, and a liquid-based specimen vial (for Stage 1) or a specimen processing card (for Stage 2). Participants were instructed to watch a video guidance on the website to learn how to get sample. A graphic guidance was available for each of the participants when they collected sample for themselves. After sample collection, participant (in the first stage) who used a kit containing liquid-based vial needed to put the brush with specimens into the vial and keep the brush head in the vial by breaking the handle off, while that (in the second stage) who used a kit containing a card needed to apply the specimen on the brush head onto the card. With specimen being put into the envelop, participants can apply logistics service on the website to ship the specimen to PUSH for hrHPV testing in the Obstetrics and Gynecology Laboratory of PUSH (Stage 1) or BGI lab under supervision of the investigators (Stage 2). No professional instruction was available online and offline during the whole sampling procedures (Fig. 1).

Samples from Stage 1 were tested for hrHPV on Cobas4800 assay (Roche, USA), and the samples from Stage 2 were tested on SeqHPV assay (BGI, Shenzhen, China). Cobas4800 is a multi-PCR based HPV assay detecting 14 types of hrHPV in three channels respectively for HPV16, HPV18, and 12 other HPV types in pool (pooled 12-HPV) (HPV-31, -33, -35, -39, -45, -51, -52, -56, -58, -59, -66, and -68). SeqHPV is a high-throughput assay based on multiplex PCR for DNA extraction and amplification and next generation sequencing for HPV genotyping. It is configured to detect and report 14 hrHPV types (HPV-16, -18, -31, -33, -35, -39, -45, -51, -52, -56, -58, -59, -66 and -68) individually. It had been validated in a cross-sectional clinic trial to have the same sensitivity and specificity with Cobas4800 in detection CIN2+ and CIN3+ lesion using both physician- and self-collected specimens [20]. To conduct comparative analysis, we categorized hrHPV positive results to be three hierarchies as HPV16, HPV18, and 12-HPV in-pool, which refers to positive of HPV16 only and multi-type of hrHPV including HPV16, positive of HPV18 and multi-type hrHPV including HPV18 but excluding HPV16, and positive of single-type of hrHPV other than HPV16 or 18 and multi-type of hrHPV excluding HPV16 and/or 18, respectively [19].

HPV testing results were uploaded on to the website, which were individual-specific for access by the participants with a validate password and username or ID number. Participants who tested positive of hrHPV (the positives) could schedule a diagnostic visit to PUSH on the website. Offline contacts through phone call were given to the positives participants who did not return for triage in one month after the results were reported online, to encourage them to visit the hospital for management.

Colposcopy and biopsies were performed on all positives returned for management, which were conducted by trained clinicians from PUSH following a POI-protocol developed by Preventive Oncology International (POI). According to that protocol, colposcopy-directed biopsy was taken at all sites with visible lesion and random biopsies were taken at all sites without visible lesion in any or all the 4 quadrants plus endocervical curettage (ECC) [20]. All the biopsied specimens were processed in PUSH lab and analyzed by a gynecological pathologist from PUSH, who was blind of HPV genotype information and the colposcopy impressions. Pathology diagnoses were reported according to the 2014 WHO Classification of Female Genital Tumors [21].

To evaluate the effectiveness of the internet platform in self-sampling based cervical cancer screening program, analysis of the data will be focused on: (1) the rate of qualified registration data among the applicants, (2) the rate of sample provision among those who got the sampler, (3) the rate of qualifies samples, (4) the rate of questionnaire response among enrolled participants, and (5) potential elements that impact on the online screening. To analyze the composition of the participants in internet screening, the demographic and background data of the participants and the clinical records were compared by grouping the participants in variety of dimensions. Univariate regression analysis was conducted to identify the association of demographic and historical details with the participation and sampling rate. The chi-square test was used to conduct comparative analyses. SPSS v.26.0 software (IBM, Armonk, NY, USA) was used for all data analysis in this study with the significance level to be p < 0.05.