Scientists at Keele University have identified eight symptoms that, when they are presented to a GP, have a more than five per cent chance of turning out to be cancer.
The 'predictive values' of symptoms, or the chance that a symptom means cancer, are useful for GPs to help them decide whether a patient needs further investigations or specialist advice.
Dr Mark Shapley and his colleagues found that eight clinical features had a higher 'predictive value', which were: rectal bleeding, iron deficiency anaemia, rectal examination that gives cause for concern, haematuria (blood in the urine), haemoptysis (coughing up blood), a breast lump, postmenopausal bleeding and dysphagia (difficulty in swallowing).
Researchers identified the eight symptoms after analysing 25 studies from the UK, US, The Netherlands, Belgium, Australia, Denmark and Germany.
Writing in the British Journal of General Practice, they suggest that further research should be carried out to provide GPs with more robust methods to help them diagnose cancers at an early stage.
Dr Shapley said: "We recommend research and development of general practice computer systems to produce effective warning flags when the symptoms, signs or test results with a risk of five per cent or more from unselected primary care populations are entered for patients within the specified sex and age groups.
"GPs should audit their management and reflect upon these cases as part of their appraisal to improve quality of care. There should be more open public debate on the level of risk that triggers a recommendation for referral by a GP."
Writing in an accompanying editorial, GP Dr Kevin Barraclough noted that the relevance of the eight symptoms is affected by an individual patient's age.
For instance, iron deficiency anaemia - one of the eight clinical features - is unlikely to be due to bowel cancer in a 21-year-old female, but more likely to be symptomatic of cancer in a 60-year-old male.
Dr Barraclough said: "More research in primary care is urgently needed to inform cancer referral guidelines. What we really need are numbers from good primary care studies. These are surprisingly few and far between."