By simply running this test on their medical records, GPs may be able to clarify the patient’s bowel cancer risk, helping them decide on a referral for a cancer screen or colonoscopy.
- Dr Tim Holt, University of Oxford
Published today in Cancer Medicine, University of Oxford researchers show how an advanced computer algorithm technology can successfully indicate levels of risk by analysing results in patients' existing medical records from routinely collected blood tests and blood markers.
Advancements in algorithm technology – a mathematical analysis of data – have led to the ability to ultra-rapidly search through existing NHS records to determine whether a patient is at risk of colorectal cancer. Colorectal cancer is the fourth most common cancer in the UK, with around 41,000 people diagnosed annually. The new screen has the potential to save lives through early detection as well as reducing the need for unnecessary, costly and invasive tests.
The algorithm, established originally in Tel Aviv, has now been demonstrated for the first time as effective within a UK population following a two-year research study funded by the NIHR Oxford Biomedical Research Centre.
The study was led by Professor Julietta Patnick, a Visiting Professor in Oxford University’s Cancer Epidemiology Unit, Nuffield Department of Population Health, and previously Director of NHS Cancer Screening. Commenting on the results, Professor Patnick said: “The beauty of this algorithm is that it allows healthcare professionals to screen an individual with their particular history – rather than mass screening as at the moment. And simply by the push of a button, it is a system which can be applied to GP surgery computer systems, running quickly and efficiently to give a more accurate level of risk.
“Our initial research is very encouraging and we now plan to test this with GPs as quickly as possible as it has the ability to provide more information earlier, and so raise the level of suspicion for GPs about particular patients. Earlier detection of bowel cancer saves lives, there is no question about it.”
A team from Oxford University’s Nuffield Department of Primary Care Health Sciences were involved in the study. Dr Tim Holt, Senior Clinical Research Fellow in the department and a GP in Earls Court, said: “This has the potential to be a really valuable tool for GPs, especially if they are concerned about a patient’s symptoms. By simply running this test on their medical records, GPs may be able to clarify the patient’s bowel cancer risk, helping them decide on a referral for a cancer screen or colonoscopy.”
Dr Holt, corresponding author on the study, worked with Dr Jackie Birks of the Centre for Statistics in Medicine, and Dr Clare Bankhead and Dr Alice Fuller in the Nuffield Department of Primary Care Health Sciences’ Medical Statistics Group to analyse 2.5 million NHS records for the study, obtained through the Clinical Practice Research Database. The project was supported by, but conducted independently of the company Medial EarlySign, who produced the risk scores.
Dr Matthew Hallsworth from the NIHR said: “There is a real opportunity to improve our ability to assess the risk of colorectal cancer. This has the potential not only to improve patient outcomes but also to save money for the NHS.”
Health Minister Lord O’Shaughnessy said: “This promising study is further evidence that using data to aid research can save lives. The benefits of using data are clear – more advanced and effective treatments and diagnoses for NHS patients both now and in the future."
Father of three, Richard, from Surrey was recently treated for bowel cancer. He says: “At age 54 I was diagnosed with stage 3 bowel cancer. My diagnosis came about following a chance visit to hospital on an unrelated matter that lead to a blood test showing I was mildly anaemic. In the few weeks before the diagnosis the only symptom I had noticed was being slightly more breathless than normal but only when exercising heavily, nothing else. This lack of symptoms is apparently quite common in cancers located far back in the colon.
“Any test that identifies a higher risk and enables early diagnosis is welcome and could have saved me from chemotherapy and maybe even surgery .The scariest thing is how subtle the symptoms were and how well you can feel while still being seriously ill. I am still being monitored but hopefully this is the end of my treatment. I feel both very lucky and unlucky!”
Evaluation of a prediction model for colorectal cancer: retrospective analysis of 2.5 million patient records. Birks J, Bankhead C, Holt TA, Fuller A, Patnick J. Cancer Medicine 2017. DOI: 10.1002/cam4.1183