Six months after turning 24, women across the UK receive their first cervical screening invitation – and they’ll be invited at regular intervals until their mid-60s.

Taking part offers vital protection against cervical cancer – but it can be an intimate, uncomfortable clinical procedure. Despite this, most people turn up – and as a result, the UK’s national screening programmes have prevented tens of thousands of premature deaths since the late 1980s.

Yet many don’t take up their invitation, for a whole range of reasons.

Whatever the individual reason, the number of people who don’t attend is growing rapidly. After hitting the NHS target of 80% for many years, rates have now dropped well below it, with the latest data showing that just two-thirds attend their appointments. That translates into around 1.3 million people foregoing cervical screening every year, putting them at greater risk of developing cancer.

This is a particular issue among certain groups – deprived communities, ethnic minorities, and those with disabilities – who consequently experience a greater impact from the disease.

“Cervical screening rates have been falling fast and it’s not unfair to say the system is in crisis,” says Professor Peter Sasieni, co-lead of the Centre for Cancer Screening, Prevention, and Early Diagnosis, based at Queen Mary, University of London’s Faculty of Medicine and Dentistry, and whose career has been supported by Cancer Research UK since the late 1980s.

Most recently, their work – culminating in two large studies, YouScreen and HPValidate – has proven pivotal in the recent Government decision to introduce the latest innovation in cervical screening: self-sampling. This allows people to take samples at home using a vaginal swab, rather than through attending appointments with health professionals. And it could help give screening rates a big boost, particularly in underserved communities.

The path to self-sampling

Cervical screening was first introduced in the UK as a patchwork of local endeavours in the mid-1960s before being mandated nationally in the late 80s – and it proved to be an overwhelming success. After cervical screening began to be coordinated nationally, cervical cancer death rates began to plummet, ultimately dropping by 75%. Today, cervical cancer rates are at historic lows, with a 2016 estimate suggesting that the UK’s national programmes had prevented around 2,000 premature deaths each year.

Early screening relied on a technique developed in the 1940s, the ‘Pap smear’, which directly fixed cells onto a microscope slide to look for abnormalities. But the UK’s programmes continually evolved, and since the early 2000’s cells have instead been analysed with liquid-based cytology. As well as improved accuracy, the switch to liquid-based samples also opened the door to another innovation: these samples can be easily tested for HPV.

The NHS programme introduced HPV testing in 2010, initially as a ‘backstop’ to determine whether borderline cytology results needed further action. But as evidence began to show HPV testing was more accurate, the NHS began to switch over to HPV testing, and now uses this as its primary screening method, with cytology relegated to the backstop role.

“It was clear to us, back then, that it was likely to be a better way to do screening than cytology,” Sasieni recalls, “because with cytology, we knew you had to take the sample precisely. Abnormal cells are only found in a very small region of the cervix, whereas we suspected an HPV infection would be detectable pretty much everywhere, meaning you didn’t need to be so precise”.

The growing promise of HPV testing also prompted Sasieni’s colleagues, including Professor Sir Jack Cuzick and the late Dr Anne Szarewski (who had also led the UK’s participation in the global HPV vaccination trials), to investigate whether it might allow people to take samples themselves, and so overcome many of the barriers around screening. For example, self-testing offers the potential for reducing both the ‘admin burden’ of booking and attending an appointment, while also avoiding negative aspects like discomfort, pain or lack of dignity that some associate with a speculum examination.

In 2007, Szarewski and her team published results of a small study of self-sampling, run in collaboration with researchers at UCL. The results showed it would be reasonable to offer HPV self-testing to women who are reluctant to attend for cervical smears.

They called for more research to directly evaluate self-testing among this ‘non-attender’ population.

Among the study team was health psychologist Professor Jo Waller, then a PhD student, now Professor of Cancer Behavioural Science at QMUL. She remembers how, by the mid-2000s, there was growing interest in self-sampling’s potential to widen access to cervical screening.

However, this would need to be done carefully, based on robust evidence and careful evaluation, to ensure it didn’t introduce unintended barriers.

To move forward, says Sasieni, there were several distinct questions that needed to be answered. The first – which Sasieni’s team would address – was practical: “We needed to find out whether it was feasible – logistically and economically – to offer self-sampling to non-attenders within the current screening programme, and do so in a way that would enable them to take part in screening.”

The other questions, which Rebolj and Waller would tackle, were around technical aspects, such as test accuracy and ease of handling in NHS laboratories, and how acceptable different devices were to those using them.

You screen if you want to

“We’d already discovered that non-attenders do still tend to visit their GPs for other reasons,” says Sasieni, “it’s not just a group who don’t engage with primary care. And that got us wondering, could you put a little flag on non-attenders’ GP records when they were six months overdue, and this would pop up during an appointment to prompt their GP to offer them a self-sampling kit for screening.”

His team piloted this concept in diverse areas of South and East London with reasonable success, publishing findings in 2017. “But uptake was really variable by GP practice,” Sasieni recalls. “And when we looked into this, it seemed that although people who’d been offered a kit were consistently likely to return it, in some practices GPs just weren’t having those conversations.” This suggested that GP behaviour and incentives might be just as important as anything to do with kit acceptability.

“We then ran a second study in West London, ALOHA, funded by Cancer Research UK, where we randomised GP practices to either do what we’d done before – have the flag come up if people were 6-month overdue – or not,” says Sasieni. But then on top of that, to get around the fact that some GPs may not be offering kits, when people were 15 months overdue, they were also randomised to receive either a letter offering self-sampling; or a kit sent to them directly; or nothing.

“In terms of efficiency, the opportunistic personal approach – a conversation with a GP – was the most successful, and we also found that sending kits directly was much better than sending letters,” he says.

But ALOHA’s results, drawn from one corner of London, were still very much provisional. Fully understanding how best to invite people for self-testing would need an even bigger study.

“Around that time, we were contacted by the NHS’s North Central & East London Cancer Alliance, who were concerned about their poor cervical screening rates, and wanted to know what to do to improve them,” says Sasieni.

The Alliance – which has now morphed into two separate entities, North East London, and North Central London – agreed to fund a large self-sampling study, YouScreen, to build on the findings of ALOHA.

“Essentially, in YouScreen, we did what we did in ALOHA but with a key difference,” says Sasieni. “We persuaded England’s national screening programme that if study participants returned a sample, it formally counted as having been screened.”

This was important for two reasons. First, if GPs knew participating would count towards their screening reimbursement, they’d be much more likely to offer a test kit. And second, it benefited participants too. “During ALOHA, participants would occasionally get annoyed – they go ‘hang on a minute, four months ago, I got a letter to say my test result was negative and everything’s fine. And now I’ve got a letter saying I missed my last appointment, so I need to come for screening’. So, it was vital that we made the trial screening count as part of the national programme.”

YouScreen launched in 2021, and its findings were published three years later: nearly two-thirds of around 9,000 non-attenders returned a sample after being offered a kit by their GP, along with 12% of some 17,500 who received one in the post. Overall, this represented a significant increase in coverage among non-attenders: self-sampling was indeed feasible and could help boost rates.

And, even more positively, 6 in 10 people who took part were from deprived backgrounds, and around 6 in 10 were from ethnic minority groups. This showed self-sampling had huge potential to minimise inequities in cervical screening.

Sasieni’s team also led a parallel economic analysis, in collaboration with colleagues in Australia, based on the YouScreen data – which rigorously modelled how cost-effective self-sampling might be under a range of different conditions.

Together, these were vital pieces of the puzzle. But YouScreen’s primary aim was to determine whether participants would, in practice, return their kits. It was less concerned with how accurate the tests were, nor whether they’d work on standard NHS equipment.

For self-sampling to be scaled up across the UK’s diverse patchwork of healthcare systems, these questions needed answers, and Matejka Rebolj’s team, working closely with Jo Waller and her colleagues, set out to find them.

Validation

A HPV test has two components.

On the one hand, there are the kits used to take vaginal samples – and there now are scores of such devices on the market.

On the other, there are the laboratory machines used to analyse these samples. And by 2019, with cervical screening across the NHS now reliant on HPV testing rather than cytology, two separate, equally effective laboratory technologies had been approved: one that relied on testing for viral DNA, the other analysed levels of viral mRNA. That means, says Rebolj, “half of the country’s labs now use DNA and the other half use mRNA.”

With attention turning to self-sampling, that meant that any home test kits the NHS procured would need to be compatible with both platforms. QMUL experts Cuzick and the late Louise Cadman had begun to investigate this issue in the late 2010s, testing a range of self-sampling devices in the Predictors studies.

But with momentum building, and recognising the need to pin this down before self-sampling could be rolled out, Public Health England commissioned a large study to test five combinations of three different kits and two analysis machines, within existing NHS HPV testing laboratories. The study, called HPValidate, launched in 2021 and ran across five regions of England.

“After we finished HPValidate, we presented the data at so many different meetings,” Rebolj recalls. “There were so many people who were looking at these data within the NHS, within the Department of Health and Social Care, cross-institutionally and so on.”

“And with the YouScreen data – both the findings and the cost-effectiveness analysis – we had all the ingredients to allow policymakers to move forward.”

And in March 2025, based on YouScreen and HPValidate – and following a three-month public consultation – the UK’s National Screening Committee recommended that self-sampling be offered to non-attenders within all four nations’ screening programmes.

Roll-out is now underway across England, where people who are overdue for six months or more will be offered a self-sampling kit to complete at home and return by post. Scotland and Wales are soon to follow suit, with Scotland recently launching a formal pilot. (As of April 2026, Northern Ireland has yet to formally announce plans).

The evidence assembled by the QMUL team is clear: offering self-sampling to people who hadn’t recently attended screening was feasible, effective, and has enormous potential to break down barriers and boost screening rates – particularly among underserved communities.