Increase in Response by Offering Self-sampling Devices in Belgian GP Practices to Non-screened Women (BELGSSAR)

Randomized Participation Trials: Increase in Response by Offering Self-sampling Devices in Belgian GP Practices to Non-screened Women

Small scale data indicate that cervical cancer screening participation may increase when self-sampling (SS) devices are offered directly by health care workers to non-screened woman, when those woman contact health services for whatever reason. The purpose of the current research is to reproduce the early findings of a MSc project conducted in a general practitioners (GP) group practice in Brussels, where women not screened since >3 years randomized to direct reception of a SS kit yielded a response of 78% vs 51% in the control arm. BELGSSAR will also investigate whether GP's have the information on the most important risk factors for cervical cancer available in their patient files.

Screening with clinically validated high-risk (hr) HPV tests is more effective in reducing cervical cancer incidence than cytology. Another advantage of hrHPV testing is that it can be performed on self-samples whereas cytology on self-samples shows poor clinical accuracy. hrHPV DNA testing on vaginal self-samples is as accurate on self- as on clinician-taken samples under the condition to use clinically validated PCR-based assays that target DNA sequences of the viral genome. Offering kits for vaginal self-sampling is more effective than conventional invitations / reminders sent to under-screened women to have a cervical specimen taken by a clinician. Small scale data indicate that screening participation may increase when self-sampling (SS) devices are offered directly by health care workers to non-screened woman, when those woman contact health services for whatever reason. The purpose of the current research is to reproduce the early findings of a MSc project conducted in a general practitioners (GP) group practice in Brussels, where women not screened since >3 years randomized to direct reception of SS kit yielded a response of 78% vs 51% in the control arm. Risk based cervical cancer screening takes risk factors into account to differentiate screening policies. GPs have knowledge of these risk factors. Demonstration of the knowledge of these risk factors from GP's patient files will be demonstrated through the BELGSSAR trials. Two self-sampling methods will be used in BELGSSAR: vaginal self-sampling using the Evalyn® Brush (Roovers, Oss, NL) and first-void urine collection with the Colli-PeeTM (Novosanis, Wijnegem, BE). Depending on the GP practice, collection will take place at home or at the GP office (predefined by the respective GP). The inclusion of urine collection is motivated by the possible lower threshold for women who dislike genital examinations. The following hypotheses will be investigated: Participation in cervical cancer screening is substantially higher when a GP offers a self-sampling kit compared to control interventions. Risk factors for cervical cancer can be identified accurately from GP's patient files. To test these hypotheses, a two-arm participation trial will be conducted in eight GP practices in Belgium, spread over four locations (Ghent, Leuven, Brussels and Liège). In each GP practice the enrolled patients will be randomized to one of the two study arms. In the experimental arms (Arm 1), the woman will be asked to collect a vaginal sample with the Evalyn® Brush or a urine sample with the Colli-PeeTM. Women shall receive this self-sampling device directly from the GP or shall pick it up at a close by pharmacist with a prescription of the GP. In four GP practices (1, 2, 5, 6) the patient has to collect the SS preferentially in the GP practice (at home if the GP practice is not well equipped) whereas in the other four GP practices (3, 4, 7, 8), the patient has to collect the SS at home and send it to the laboratory by using a prepaid envelope. Women in the control group (Arm 2) will receive an oral recommendation given by the GP with a reminder of the current screening policy to have a cervical sample taken by a physician chosen by the woman. This clinician-taken cervical sample will be sent to a laboratory for processing with the usual screening test. Today the usual screening test still is cytology, but this will change in the future (date to be defined) to an HPV test. Double blinding of patient and investigator is not possible, because the participants will be aware to which study arm they belong when they receive or offer the self-sample. All samples collected under the experimental arms will be handled and tested for presence of hrHPV types at the AML laboratory in Antwerp. AML is the national reference laboratory for HPV for Belgium. The used HPV assay will be the RIATOL qPCR, that detects E6/E7 HPV genes, with measurement of type-specific viral load of all high-risk HPV types. This test is clinically validated for HPV-based cervical cancer screening. The tests performed in the control arm will be processed in the laboratories connected with the respective GP practices.

In the experimental arms, the woman will be asked to collect a vaginal sample with the Evalyn® Brush or a urine sample with the Colli-PeeTM. Women shall receive this self-sampling device directly from the GP or shall pick it up at a close by pharmacist with a prescription of the GP. In four GP practices (1, 2, 5, 6) the patient has to collect the SS preferentially in the GP practice (at home if the GP practice is not well equipped) whereas in the other four GP practices (3, 4, 7, 8), the patient has to collect the SS at home and send it to the laboratory by using a prepaid envelope.

Women in the control group will receive an oral recommendation given by the GP with a reminder of the current screening policy to have a cervical sample taken by a physician chosen by the woman. This clinician-taken cervical sample will be sent to a laboratory for processing with the usual screening test. Today the usual screening test still is cytology, but this will change in the future (date to be defined) to an HPV test.

The GP will hand over the self-sampling kit to the women, after obtaining informed consent, and the women then takes the self sample at home or in the doctor's office (predefined), according to the respective users manual.

Difference in response between intervention and control arm (per protocol (PP) and intention to treat analyses [ITT]).

Difference in response in women included in the self-sampling arms vs the control arm (per protocol (PP) and intention to treat analyses [ITT]). The absolute difference in response rate = proportion with screening test in experimental arm - proportion with screening test in control arm, within 6 months after enrolment. In the PP analysis only women with a screening test on the self-sample according to the specific experimental intervention will be taken into account, whereas in the ITT analysis all screening tests performed within 6 months after enrolment will be considered (also those performed on samples taken by a clinician).

Heterogeneity in response rate differences, over the eight GP practices (assessed by the Cochran Q test for heterogeneity)

net difference in response rate between practices offering vaginal self-sampling devices vs offering urine collection devices • Difference in response between practices offering self-sampling collection at the GP's office vs at home.

net difference in response rate between practices offering self-sampling collection at the GP's office vs at home.

Inclusion Criteria: women who have not been screened for cervical cancer since the last three years visiting their GP for whatever reason Exclusion Criteria: pregnant women, women under active follow-up because of previous cervical abnormality, women who had a total hysterectomy, women who had a history of cervical cancer or a treatment for cervical precancer less than three years ago, non-consenting women, women who are not able to understand and sign the informed consent