Exclusive GPs are set to have a new set of risk calculators embedded in their computer systems to help them more easily identify which patients are most likely to have a range of common cancers, Pulse can reveal.
The calculators – developed by researchers at the University of Nottingham and ClinRisk Ltd. – have were described as a potentially ‘defining moment’ in cancer diagnosis by the RCGP last year and have now been extended to include four new cancers.
The QCancer tool is already available online (www.qcancer.org) and researchers told Pulse they have reached an agreement with EMIS and are in talks with the providers of TPP SystmOne to make it available directly to GPs via their IT system provider.
Results from the QCancer tool were published last year for lung and gastro-oesophageal cancers, and were called a ‘vitally important’ development for earlier GP diagnosis of cancer by RCGP chair Dr Clare Gerada.
This month the results of new scores for ovarian, colorectal and pancreatic cancers have been published, and results for a new renal cancer score are expected to be published soon.
Data from the QResearch database in patients aged 30-84 years looked at those who did not have the cancer or major symptoms at baseline, but who were diagnosed within the following two years.
Published in the British Journal of General Practice, the data show a family history of gastrointestinal cancer, anaemia, rectal bleeding, abdominal pain, appetite loss, and weight loss were independent predictors for colorectal cancer.
In men the list also included alcohol consumption and recent change in bowel habit.
Exactly the same analysis done for pancreatic cancer – a disease for which there are no NICE guidelines on referral of patients with suspected cancer – found age, smoking, type 2 diabetes, chronic pancreatitis, abdominal pain, appetite loss and weight loss were independent predictors.
In women, the list also included abdominal distension and in men dysphagia and constipation.
Results for ovarian cancer, published today in the British Medical Journal, found that age, family history of ovarian cancer, anaemia, abdominal pain, abdominal distension, rectal bleeding, postmenopausal bleeding, appetite loss, and weight loss, predicted a person’s risk of the disease.
All three calculators predicted the 10% of the population with the highest risks who accounted for roughly two-thirds of the specific cancers over the two-year study period.
Study leader Professor Julia Hippisley-Cox, professor of clinical epidemiology and general practice at the University of Nottingham and a GP in the city, said the tools could be used to evaluate an individual’s risk during a consultation but also to ‘batch’ assess a practice population to flag up high risk patients for investigation much in the same way as GPs already use QRISK for cardiovascular risk.
Using the tools to assess the importance of a patient’s symptoms can ‘either be useful for reassurance or it can help the GP decide on further investigation and referral and the degree of urgency,’ she said.
Professor Hippisley-Cox said EMIS have already agreed to implement the cancer calculators, which are available to all software suppliers, if there is demand from GPs. The providers of TPP SystmOne told Pulse they were also in talks to provide the risk scores.
Professor Willie Hamilton, professor of primary care diagnostics at Peninsula College of Medicine and Dentistry and a part-time GP, said the calculators were based on very solid pieces of work but questioned if GPs would have time to run them routinely.
He said: ‘Personally, I think GPs should use the symptoms these studies have found and then consider testing, and wait for the algorithms to be tested out further.
He added: ‘NICE hasn’t really worked in selecting patients for the two week wait clinics sothis and other similar work does encourage us GPs to test a bit earlier, and is a good thing.’
A spokesperson from TTP said: ‘Our teams here at TPP are aware of the new algorithms and our clinical director, Dr John Parry is currently in talks with Nottingham University regarding the incorporation of their recently developed tools into SystmOne software.
How risk profiles compare
A 60-year old man who is a non-drinker with a positive family history of gastrointestinal cancer, anaemia, and a recent change in bowel habit has an estimated risk of colorectal cancer of 1.5%. If he also has loss of appetite, the estimated risk is 3.1%. If he also has rectal bleeding in addition to these symptoms, his estimated risk of colorectal cancer is 48.6%.
A 75-year old woman, who is a moderate smoker and has abdominal pain, abdominal distension and loss of appetite has an 11.3%. If she also has weight loss, the estimated risk would be 44.1%.