Completed Digestion, Kidneys & Other Organs Cancer

CANcer DIagnosis Decision rules (CANDID)

Funder NIHR (part of DHSC)
Recipient organisation Southampton
Funding £2.1M
Period Jan 2012 — Sept 2015

In plain English

AI plain-English summary

Doctors in UK primary care will collect standardised clinical information from patients with breast or bowel symptoms to build and test prediction rules that could flag who is most likely to have cancer. Current guidelines for referring patients with possible cancer symptoms rely largely on age and a small number of “red flag” signs, but many cancers are diagnosed late because early symptoms are vague and common. This study addresses the lack of large, prospective primary care studies powerful enough to develop and validate clinical prediction rules for breast and colon cancer. The researchers will combine symptom patterns, signs, and results from routine blood tests (full blood count, CRP, and ferritin) with lifestyle questionnaires and genetic samples. If successful, the resulting prediction rules could give GPs a more precise, evidence-based tool for deciding which patients need urgent referral and which can be safely monitored. That could reduce both missed cancers and unnecessary referrals, improving diagnostic efficiency in primary care without requiring new equipment or specialist input. The rules would be tested in separate validation cohorts before any clinical use.

View original technical description
Expert Summary: There is suggestive evidence that clinical prediction rules (CPRs) for diagnosing both breast and colon cancer can be developed in primary care. However there have been no sufficiently powered prospective primary care cohort studies to develop CPRs, nor to test such rules in validation cohorts in primary care. This collaborative group has a successful track record of using very similar methods to assemble large prospective cohorts in multicentre primary care studies to develop CPRs. We will select the most likely candidate symptoms and signs to predict breast cancer or colon cancer based on the prior literature, patient interviews, and a ‘Delphi’ exercise. 10,000 patients presenting with breast symptoms, and 10,000 patients presenting with lower bowel symptoms will have clinical information collected using standardised internet based clinical proformas. Willing patients will also complete validated lifestyle questionnaires and provide blood or saliva samples for GWAS (Genome Wide Association Study – to be funded separately), and blood samples (particularly for full blood count, CRP and ferritin). Patients will be followed up using the National Cancer registries for the diagnosis of breast and colon cancer. CPRs developed from the derivation cohorts will then be tested in validation cohorts for a further 2000 patients for each condition

View the original record at the funder ↗

Researchers

Paul Little (Principal Investigator)

Related Research

Grants with similar aims, by meaning.

Developing a decision rule using existing blood test data to select primary care patients with non-specific symptoms for urgent cancer investigation.
Assessing the potential for earlier detection of colorectal cancer in primary care, using the GP research database (supported by the Bobby Moore Fund)
ColoRectal Early Diagnosis: Information Based Local Evaluation (CREDIBLE)
Towards earlier diagnosis of colorectal and ovarian cancers in Primary Care.
ICEBERG: Improving Cancer Early detection using Blood tEst Results in General practice

Original classification

NIHR School for Primary Care Research - Primary Care

Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.