HPV self-sampling among long-term non-attenders to cervical cancer screening in Norway: A pragmatic randomized controlled trial

Self-sampling at home can improve compliance with cervical screening recommendations. But will it work for women who rarely or never screen? And what is the best way to deliver this service to women?
HPV self-sampling among long-term non-attenders to cervical cancer screening in Norway: A pragmatic randomized controlled trial

May offering HPV self-sampling improve cervical cancer prevention?
Screening has substantially reduced the incidence and deaths due to cervical cancer. However, the coverage of cervical cancer screening programmes is suboptimal. Women who have not undergone screening are at higher risk of cervical cancer. Thus, interventions that improve screening participation would benefit women’s health. Self-sampling for human papillomavirus (HPV) testing identifies women with cervical precancer and cancer with similar accuracy as clinician-collected samples. Vaginal self-sampling is shown to be more acceptable and preferred by women over clinician sampling. Self-sampling may reinforce the importance of screening, empower women, and mitigate some of the barriers associated with attending a gynaecological screening exam.

The study aimed to compare cervical cancer screening participation among women in Norway who have not attended screening for at least ten years, who received either a standard reminder letter to attend screening in a clinic (control arm), an offer to order an HPV self-sampling kit by mail, e-mail or through a webpage (opt-in arm), or were sent an unsolicited HPV self-sampling kit (send-to-all arm). As secondary endpoints, we estimated the corresponding high-risk HPV (hrHPV) positivity rates and occurrence of cervical pre-cancer and cancer among the long-term non-attenders who participated in the screening, as well as their attendance to follow-up if the screening test turned out positive.

The study setting
The Cancer Registry of Norway (CRN) is responsible for the national cervical cancer screening programme (CervicalScreen Norway). CRN invites women aged 25 to 69 years old to undergo screening. CervicalScreen Norway uses a centralised invitation procedure and issues standard open reminder letters to women who have not been registered with a screening test during the recommended interval. If after one year a woman is still not registered with a test result, a second reminder is issued to her. These reminders encourage women to schedule an appointment with their GP or a gynaecologist for a screening test. All screening and triage results are registered in the CervicalScreen Norway database. Thus, these registries ensure a precise definition of the study cohort and complete information on participation and diagnostics.

In March 2019, we initiated a pragmatic randomised controlled trial with participation as primary outcome in collaboration with CervicalScreen Norway. A random sample of 6000 women, aged 35-69 years old residing in four different counties, who had not attended screening for at least ten years were randomised 1:1:1 to the three intervention arms. 

With focus on user-friendliness and safety, the research team prepared an HPV self-sampling kit. This kit included the self-sampling device with user instructions, a sealable plastic bag, a prepaid return envelope addressed to the laboratory, and an information leaflet. The team closely collaborated with the national HPV reference laboratory to ensure safety in the conduct of the trail. In addition, during the launch of the trial, we sent out press releases which resulted in media coverage that communicated the trustworthiness of the study.

Women who tested negative to HPV were encouraged to attend the regular cervical cancer screening programme at the recommended interval. Women who tested positive to HPV received a pre-booked triage appointment with a physician. To address feasibility and performance of alternative triage strategies that could be implemented in a screening programme, women were sequentially allocated to triage at their GP or a gynaecologist practicing in the largest city in their county. 

Offering HPV self-sampling increased participation considerably
Because of the opt-in and send-to-all self-sampling strategies, screening participation among long-term non-attenders to cervical screening increased. Total participation was 4.8%, 17.0% and 27.7% among control, opt-in, and send-to-all, respectively. Self-sampling increased the total participation of women in all ages, including those who rarely and never have been screened. Compared to previous studies, these are high numbers, particularly for opt-in. 

High-risk HPV was detected in 11.5% of self-samples and 9.2% of clinician-collected samples. Management of hrHPV-positive self-samples by GP or gynaecologist give similar results in terms of triage attendance. Triage attendance was similar in the opt-in and send-to-all arms at 92.6% and 92.5%, respectively, for any attendance within six months. Of the 933 women screened, 33 (3.5%) had high-grade cervical lesions (CIN2+), and 11 (1.2%) had cervical cancer. 

The hrHPV positivity rate of the total study population of long-term non-attenders was nearly twice as high as the corresponding rate among women of the same age in the ordinary screening population. Long-term non-attenders have an elevated risk for hrHPV infection and high-grade cervical lesions. This highlights the large potential to improve cervical cancer prevention by increasing the screening participation of this population.

Future implementation in the national cervical cancer screening programme
The Department of Health and Care recently granted funding to CervicalScreen Norway to implement HPV self-sampling. Results from the current study informed this decision. CervicalScreen Norway is now assessing how the self-sampling will be implemented. Maintaining an offer for clinician-collected screening will still be important since some long-term non-attenders who were offered self-sampling preferred this option. Building on the results presented here, we are now working on a cost-effectiveness analysis of the interventions addressed in this trial. Self-sampling may relieve burden on public health care provision. Through the COVID-19 pandemic such consideration may have gained increased importance in designing future healthcare services. 

Please sign in or register for FREE

If you are a registered user on Nature Portfolio Cancer Community, please sign in