Impact of mode of offer of self-sampling to people overdue cervical screening on screening participation: a randomised controlled trial.

Background: Offering self-sampling to non-attenders increases cervical screening uptake, but the optimal approach for offering kits remains unclear.

Methods: Randomised controlled trial offering self-sampling. 13 GP (general practitioner) practices were randomised (1:1) to flagging women ≥6-months overdue cervical screening so that they could be offered a self-sampling kit if they attended their GP for any reason (N = 6080 women), or no opportunistic offer (N = 6577 women). Additionally, never screened women and those overdue screening by 15- or 27-months were individually randomised (2:1:1) to usual care (no systematic offer), a letter inviting them to order a kit (letter), or being sent a self-sampling kit (kit). The study ran from April 2019 to March 2020. The primary outcome was returning a self-sampling kit, and the secondary outcome was any cervical screening. The International Standard Randomised Controlled Trial Number (ISRCTN) is 23940319.

Findings: In opportunistic offer practices, 342 (5.6%) returned a self-sample compared with 1.9% (123/6577) in practices not randomised to opportunistic offering (adjusted risk difference 4.4% (95% CI: 2.8%-6.0%)). Half (234/449) of women offered self-sampling opportunistically returned a sample. Among 6400 women individually randomised to no systematic offer vs letter vs kit, 1.7% (54/3197), 4.8% (76/1587; difference relative to no systematic offer 3.1% 95% CI: 2.0-4.2%) and 12.3% (198/1616; difference relative to no systematic offer 10.5%, 95% CI: 8.9-12.2%), returned a self-sample (the primary outcome), respectively. These observed differences were maintained in the secondary outcome, any cervical screening. No adverse effects were reported.

Interpretation: In-person offers were most effective, but only a small proportion of non-attenders received such an offer. Postal invitations without a kit were less effective. The secondary outcome suggests those screened by self-sampling would not have been screened otherwise and contribute to increased screening coverage.

Funding: Cancer Research UKC8162/A16892 to PS (consumables), C8162/A29083 to PS (JR, AL), C8162/A25356 to PS (BN); National Institute for Health Research Clinical Research Network (NIHR CRN) Central Portfolio Management System (CPMS) ID: 36156 (AL); Intramural Research Program, Division of Cancer Epidemiology and Genetics, National Cancer Institute (RL).