Research review 049
Pragmatic implementation study with randomised start dates
Short answer
YouScreen showed that self-sampling could be offered to people overdue for cervical screening within routine English primary care. Many who accepted a personal offer submitted a sample. However, the results cannot be read as a direct randomised comparison between personal offers and kits mailed to people's homes.
What did the researchers investigate?
General practices offered self-sampling when a person overdue for screening attended a consultation for another reason. An alert in the electronic medical record was intended to remind staff to make the offer. At the same time, kits were sent directly to other eligible recipients through the screening programme.
The start dates for the practice-based intervention were randomly assigned. Researchers compared screening activity before and after implementation and also used practices without the intervention to account for general changes over time. This provides a stronger basis than a simple before-and-after count, but individuals were not randomly assigned to personal contact versus direct mailing.
What did they find?
Personal offers and postal invitations had different starting points
With direct mailing, 12.9% of kits were returned. After a personal offer, 65.5% of those who had accepted it returned a sample. However, when calculated among everyone who received the personal offer, the proportion was lower, approximately 55.9%. The denominator therefore changes the figure.
In the analysis based on assigned start date, the monthly proportion of previously non-attending people screened increased by approximately 12% in relative terms. This does not mean twelve additional percentage points of the entire population each month. Relative change describes an increase compared with the original level.
The study also showed that self-sampling could reach people from different backgrounds. Nevertheless, a positive HPV result required further follow-up. Submitting a sample was the first stage of the pathway.
How certain can we be?
People attending a practice who accepted an offer are not necessarily comparable to everyone who received an unsolicited kit by post. The difference between 65.5% and 12.9% therefore cannot be interpreted as the pure effect of the mode of invitation.
Completeness of data and follow-up varied, and adverse events were not formally collected. The study took place during the COVID-19 pandemic, with changes to services and heavy workloads. This may have affected both screening by healthcare professionals and interest in self-sampling. Projections of future national coverage are estimates, not outcomes already observed.
Who do the findings apply to?
A total of 133 general practices across five London boroughs participated, with 62 other practices providing a comparison. The target group was people at least six months overdue for screening. The areas had substantial ethnic diversity, social inequality and low screening coverage.
A total of 8,338 people submitted self-samples. This is the number who submitted, not the entire invited population. Approximately 17,604 kits were mailed directly. Personal offers were made to an estimated 10,849 people. The approaches could therefore reach different sections of the population.
What might this mean in everyday life?
The study shows that self-sampling can be integrated into existing consultations and screening routines. At the same time, services must also reach people who seldom visit their GP. Planning requires a distinction between how many people are offered testing, how many accept, how many submit a sample and how many complete subsequent examinations.
What do we still need to know?
We need to investigate the best combination of personal offers and postal mailing when services operate without the constraints of the pandemic. Researchers should follow both people offered the service and those it fails to reach.
Costs and users' experiences must be considered alongside sample numbers. It is also important to determine whether people with positive results complete follow-up and whether some groups face greater barriers than others. A sample is only the first stage of a screening pathway.
Sources
The publication covered by this review
Lim AWW, Deats K, Gambell J, Lawrence A, Lei J, Lyons M, North B, Parmar D, Patel H, Waller J, Warwick J, Sasieni PD, YouScreen Joint Steering Group. Opportunistic offering of self-sampling to non-attenders within the English cervical screening programme: a pragmatic, multicentre, implementation feasibility trial with randomly allocated cluster intervention start dates (YouScreen). EClinicalMedicine, 2024.
Other studies and professional sources mentioned in the text
Background source: Impact of mode of offer of self-sampling to people overdue cervical screening on screening participation: a randomised controlled trial. 2025. Read the publication.
Previous reference on the card
HPV self-testing and choice in Estonia. Randomised pilot study. Previous source.
Last reviewed: 30 October 2026.
Source basis: The publication's full text.
Editorial update, 1 November 2026: Language, explanations and references were reviewed. The source-review date is stated above.