The NHS is entering a defining moment in cancer care. Rising demand, workforce pressures, diagnostic backlogs and widening inequalities are forcing healthcare systems to rethink how cancer is identified and managed.
Sponsored and written by
At the same time, advances in artificial intelligence (AI), data infrastructure and clinical decision support are creating an opportunity to shift healthcare from reactive diagnosis to proactive detection.
The urgency of this shift is illustrated by the story behind Jess’s Rule. Jessica Brady visited her GP practice 20 times in six months before her cancer was diagnosed. By then, it was too late. Her family transformed that tragedy into action, leading to the creation of Jess’s Rule, now formally recognised by the Department of Health and Social Care and NHS England. Its purpose is simple: ensure repeated presentations are recognised as a potential warning sign, and opportunities for earlier diagnosis are not missed.
Repeated contact with healthcare services can itself be an important signal of risk. Historically, identifying these patterns has relied on clinician memory or manual review of records, an increasingly difficult task within overstretched healthcare systems. This is precisely where AI can make a meaningful difference.
C the Signs has operationalised Jess’s Rule within its platform, automatically flagging patients with three or more relevant consultations within a 90-day period and pre-populating a cancer risk assessment using existing electronic health record data. Working seamlessly in the background, it provides clinicians with contextual intelligence at the point of care, supporting earlier identification of risk and timely intervention.
As Andrea Brady, Jess’s mother, explains:
“We are so pleased to have the support and backing of C the Signs. Jess’s Rule will be flagged on the platform, enabling you all to action it with even greater ease.”
Jess’s Rule represents something much bigger than a single feature. It demonstrates how healthcare is evolving beyond isolated episodes of care towards continuous, proactive surveillance, where patterns and risks can be identified before patients reach a crisis point.
For decades, cancer diagnosis has largely relied on a reactive model. Patients present with symptoms, clinicians assess risk, and investigations follow once concern becomes sufficiently clear. Yet cancer rarely behaves predictably. Symptoms can be vague, intermittent or seemingly unrelated. In primary care, where cancer accounts for fewer than 2 per cent of consultations, clinicians are expected to identify one of more than 200 cancer types against a vast background of everyday illness.
The challenge is not a lack of expertise. It is the volume and complexity of information clinicians must process every day. Healthcare systems can no longer rely solely on episodic care. The future lies in intelligent systems capable of continuously identifying risk, surfacing hidden patterns and enabling earlier intervention.
Integrated directly into electronic health records, C the Signs uses AI-driven clinical decision support and population-level case finding to identify patients at risk of cancer earlier and more accurately. The platform analyses structured and unstructured clinical data in real time, including symptoms, blood results, demographics, risk factors, consultation history and free-text clinical notes.
Importantly, this is not about replacing clinicians. The future of AI in healthcare will be defined by augmentation, not automation. AI provides an additional layer of intelligence embedded within existing workflows, helping clinicians make more informed and consistent decisions.
Much of the discussion around AI in cancer care has focused on imaging or emerging technologies such as liquid biopsies. While these innovations hold promise, they can only be used once a patient has entered a diagnostic pathway. Clinical AI operating within primary care offers a different advantage: it works with data the NHS already holds today.
This matters because only around 5 per cent of cancers in the UK are diagnosed through national screening programmes. The vast majority are detected through symptomatic presentation in primary care. Many patients require multiple appointments before cancer risk is recognised, while around one in five are still diagnosed through the accident and emergency department, where the one-year survival rate is less than 36 per cent.
The opportunity, therefore, lies not simply in creating more tests but in using existing data more intelligently to identify risk earlier.
C the Signs has demonstrated that AI-driven clinical decision support can achieve 99 per cent sensitivity for identifying patients with cancer, while predicting tumour origin with more than 94 per cent accuracy. The technology is now deployed across more than 1,500 GP practices and is detecting a new cancer every 22 minutes. To date, it has helped identify more than 100,000 patients with cancer across the NHS.
The impact is not simply about increasing referrals. It is about improving referral quality and directing diagnostic resources more effectively. In a nationwide NHS evaluation involving more than 235,000 risk assessments, C the Signs safely redirected 21.7 per cent of patients away from unnecessary cancer investigation pathways while maintaining strong cancer conversion rates.
Across Somerset Integrated Care Board, implementation of C the Signs resulted in a 36.8 per cent increase in stage I and II cancer diagnoses through urgent suspected cancer referrals, alongside a conversion rate of 10 per cent, nearly double the national average.
This distinction is critical. The future of AI-enabled healthcare is not about generating more activity or overwhelming secondary care. It is about enabling smarter activity: directing finite diagnostic capacity towards the patients who need it most, reducing unnecessary investigations and accelerating diagnosis where risk is highest.
The future of cancer diagnosis is unlikely to depend on a single breakthrough test. Instead, it will come from intelligently combining data, AI and clinical expertise to identify risk earlier across entire populations.
The question is no longer whether AI belongs in primary care. The question is whether health systems can afford not to adopt technologies capable of identifying cancer earlier, reducing variation in care and helping ensure that missed opportunities become earlier interventions and better outcomes.
