Only 1 in 10 High-Risk Diabetes Patients Gets Pancreatic Cancer Imaging
New-onset diabetes with weight loss signals pancreatic cancer risk, yet UK primary care massively under-refers eligible patients for abdominal imaging.
Summary
Pancreatic cancer is diagnosed in roughly 3% of adults aged 60 and older who develop new-onset diabetes alongside weight loss — a well-recognized early warning signal. UK guidelines call for abdominal imaging to rule out cancer in this group, yet a large national cohort study found that only about 1 in 4 eligible patients received a referral, and just 1 in 10 received actual imaging. Among the 1,411 qualifying patients identified from 781 general practices, 2.8% were ultimately recorded as having pancreatic cancer — consistent with published risk estimates. The findings expose a major gap between guideline recommendations and real-world clinical practice, and now serve as baseline data for a new NHS England case-finding pilot launched in June 2025 aimed at improving referral rates and electronic health record data quality.
Detailed Summary
Pancreatic cancer carries one of the poorest prognoses of any cancer, largely because most cases are caught at a late stage. Early detection is therefore critically important, and new-onset diabetes combined with unexplained weight loss — particularly in adults aged 60 and over — has been identified as a significant early warning signal. UK clinical guidelines recommend prompt abdominal imaging to rule out pancreatic cancer in this high-risk group, making this an area where primary care has a potentially life-saving role.
This retrospective cohort study used the Royal College of General Practitioners Research and Surveillance Centre database, which covers approximately 8 million patients across 781 general practices in England. Researchers identified 180,815 adults diagnosed with diabetes between 2015 and 2021. Of these, 1,411 met the high-risk criteria of new-onset diabetes with concurrent weight loss, forming the core study population.
Among the 1,411 eligible patients, only 372 (26.4%) received a primary care referral, just 141 (10.0%) underwent abdominal imaging, and 40 (2.8%) had pancreatic cancer recorded in their health records. The 2.8% cancer rate aligns with previously published risk estimates for this subgroup, validating the guideline rationale — but the referral and imaging rates reveal a profound gap between recommended and actual clinical practice.
The study also assessed data quality, finding that 84.2% of the broader diabetes population had at least one body weight measurement and 89.4% had at least one HbA1c value recorded — data completeness that, while reasonable, still leaves room for improvement in identifying eligible patients reliably.
These findings now serve as a pre-implementation baseline for an NHS England case-finding pilot launched in June 2025, designed to improve referral pathways and EHR data quality. For longevity-focused readers, the message is direct: the combination of new-onset diabetes and weight loss after age 60 warrants urgent investigation, and patients should advocate for appropriate imaging if their physician has not already arranged it. Limitations include the retrospective design and reliance on abstract data only.
Key Findings
- Only 10% of high-risk patients (new-onset diabetes + weight loss) received abdominal imaging despite clear guidelines recommending it.
- 26.4% received a primary care referral — meaning nearly 3 in 4 eligible patients were not referred at all.
- 2.8% of the high-risk group had pancreatic cancer recorded, consistent with estimated 3% risk in adults aged 60+ with these features.
- Over 180,000 diabetes cases were analyzed; 51% were aged 60+, the age group at highest pancreatic cancer risk.
- Data gaps in body weight recording (15.8% missing) may contribute to under-identification of eligible patients.
Methodology
Retrospective cohort study using the RCGP Research and Surveillance Centre database of ~8 million patients from 781 English general practices. Adults diagnosed with diabetes between 2015 and 2021 were included, with outcomes assessed over a 12-month window before or after diagnosis. Pancreatic cancer, referral, and imaging events were extracted from electronic health records.
Study Limitations
Summary is based on the abstract only, as the full paper is not open access. The retrospective design limits causal inference, and EHR-based case identification may underestimate both referrals and imaging events if not consistently recorded. Reasons for under-referral — whether clinical, logistical, or systemic — are not characterized in the abstract.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: