Data and methods

Women with HIV aged 45 and older who attend routine appointments at the HIV clinics at Karolinska University Hospital, Södersjukhuset and Sahlgrenska University Hospital are invited, about 1,050 eligible women in total. People who have had anal cancer or rectal amputation, or who are being investigated for suspected anal cancer, are not included. Study information is available in the ten most common languages among people with HIV in Sweden.

Included women take an anal and a vaginal self-sample with a Copan FLOQSwab. Samples are tested with BD Onclarity, which detects 14 high-risk HPV types. Participants also give a blood sample for analysis of circulating HPV DNA. People with a negative anal HPV test are offered a new test after 12 months.

Women with anal high-risk HPV are referred for examination of the anal canal, either by high-resolution flexible endoscopy (HR-FEA) at Ersta Hospital in Stockholm or by high-resolution anoscopy (HRA) in Gothenburg. The examination includes a new HPV test, cell samples (cytology) and samples for DNA methylation analysis. Visible changes are removed and sent for tissue analysis. People with vaginal high-risk HPV are referred for colposcopy according to routine care. Participants treated for precancer are offered HPV vaccination.

Cell samples are assessed according to the Bethesda system, and tissue samples are classified as normal, AIN1, AIN2, AIN3 or invasive cancer. Circulating HPV DNA is analysed with droplet digital PCR at SciLifeLab, Örebro University. DNA methylation is analysed with quantitative multiplex methylation-specific PCR at Karolinska Institutet.

• InfCareHIV, the Swedish national HIV quality register: demographic and clinical HIV data.
• The National Cervical Screening Registry (NKCx): history of HPV-related gynaecological changes.
• Medical records: treatment of earlier gynaecological changes.
• Questionnaire: education, employment, sexual history, smoking, and acceptability of screening. Answered via QR code on the participant's own phone. Developed together with patient reference groups.

Data are entered in REDCap by research nurses at each site. Data are pseudonymised at each clinic, and the code key is kept at the respective hospital. Researchers analysing the data only see a study number. Pseudonymised data are stored on a secure server at Karolinska Institutet, with two-factor authentication and access limited to researchers involved in the analyses. The data management plan has been reviewed and approved by Karolinska Institutet.

Each sub-study has its own predefined analysis and sample size calculation. Methods include positive predictive values with Wilson confidence intervals, agreement between sampling methods with Cohen's kappa, McNemar tests for paired samples, exact confidence intervals for proportions, ROC analysis, a Markov model for cost-effectiveness, and deep learning models (convolutional neural networks and transformer-based architectures) for image analysis.