ideasIdea
Cancer risk scores running automatically in GP records to prompt urgent referral
Computers can combine minor symptoms, blood tests and age into a cancer risk score in the background. Showing that score to the GP could get more people referred earlier.
QCancer and similar algorithms exist but are rarely used because they need manual entry. Propose EHR-embedded, passively computed risk scores with a threshold-triggered prompt, tested in a randomised trial with stage at diagnosis and emergency presentation as endpoints, following the template of the UK ERICA cluster trial.
Hypothesis
Passive risk-score prompts increase the proportion of cancers diagnosed at stage I-II by at least 5 percentage points and reduce emergency presentations by at least 15% without doubling referral volume.
Rationale
The evidence gap is implementation, not algorithm accuracy.
What would test it
Cluster RCT across 500 practices over two years with registry-linked outcomes.
Maturity
early clinical
Who has to act
clinic
Cost to try
Medium ($1M to $50M)
Years to first evidence
3
Bottlenecks it attacks
- Most lethal cancers are found late · Screening exists for only a few cancers. Pancreatic, ovarian, liver, oesophageal and most lung cancers are found when cure is unlikely.
- Fragmented care and guideline gaps · Patients fall between specialists, wait for referrals and often do not get the treatment guidelines say they should.