Offering cervico-vaginal self-sampling for high-risk human papillomavirus testing (hrHPV self-sampling) to non-participants in a cervical cancer screening program may increase uptake depending on the delivery mode of the self-sampling offer. We compared the effectiveness of different approaches for delivering a self-sampling offer to non-participants in an organized program in terms of screening up-take, and analyzed the proportion of self-samplers that received appropriate follow-up.
The study included 9,791 women aged 30-64 from the Central Denmark Region who have not participated in cervical cancer screening despite invitation and one reminder. They were randomized 1:1:1 to either: 1) direct mailing of a HPV self-sampling kit (directly mailed group); 2) mailing of an offer to order a self-sampling kit by either e-mail, text message, phone, or webpage (opt-in group); or 3) mailing a second reminder to contact a general practitioner (GP) for usual care, viz. cytology (control group). Women offered self-sampling were informed that they could also receive usual care if wanted. The self-sampling kit comprised a brush device (Evelyn Brush) for hrHPV testing using Roche Cobas® 4800. Performing an intention-to-treat analysis, we estimated the up-take 180 days post intervention, including self-samples taken at home and cytologies taken at a GP. Self-samplers’ compliance with GP follow-up was measured 90 days after a hrHPV-positive test result.
The up-take was significantly higher in the directly mailed group (37.0%) than in the opt-in group (29.9%) (absolute participation difference (PD): 7.1%, 95% CI: 3.1-11.1%) and the control group (24.1%) (PD: 13.0%, 95% CI: 8.8-17.0%). Of 118 hrHPV-positive self-samplers, 91.0% (107) attended follow-up. Self-samplers were significantly less likely than controls to have been screened in the previous screening round (30.8% vs. 12.8%, RR: 0.42, 95% CI: 0.34-0.51). We estimated an overall participation rate of 71% in the directly mailed group, 68% in the opt-in group, and 65% in the control group. The direct mailing strategy increased the absolute overall participation among invited women by 6.0%, (95% CI: 5.8-6.2%).
Direct mailing of self-sampling kits to non-participants proved to be the most effective strategy for increasing screening participation. Using timely opt-in procedures yielded an only limited participation gain compared with a second reminder to attend regular screening. Most hrHPV-positive women had appropriate follow-up. Implementing self-sampling in the Danish screening program may increase overall up-take and help recruit under-screened women, thereby increasing the program’s effectiveness.