GP prescription patterns could help diagnose cancer patients earlier

Researchers are compiling an anonymous dataset of primary care prescription data - totalling 80m prescriptions a month - to identify patterns that could help GPs catch cancer earlier.

Scientists will use GP data to look for trends in medications given to patients before they were diagnosed with cancer – allowing them to flag up any new patients who exhibit similar tell-tale signs and may need referring for further tests.

Only around half of patients with the most common cancers exhibit ‘red-flag’ symptoms, and even fewer in some cancers with poor survival rates such as pancreatic, stomach, ovarian and brain cancer.

The research team at Health Data Insight won Cancer Research UK’s £200,000 Pioneer Award – which encourages innovative ideas for tackling cancer – to take the project forward.

They hope the database can be developed into a tool that will help GPs diagnose cancer earlier in patients with non-specific symptoms - when symptoms are easier to treat - and drive up survival rates.

Cancer diagnosis

RCGP chair Professor Helen Stokes-Lampard said: ‘We know that timely diagnosis of cancer leads to better outcomes for our patients, so any new data that can support GPs refer patients appropriately should be welcomed – but we also need much better access to reliable, diagnostic tools in the community.

‘Some of the most common and most dangerous cancers don’t have "red-flag" symptoms and can have very similar symptoms to more common, non-life-threatening conditions. This makes diagnosis in general practice incredibly difficult, especially within the constraints of the standard 10-minute consultation – and currently we have one of the lowest access rates to diagnostic tests in Europe.

‘Nevertheless, GPs are already doing a good job of appropriately referring our patients that we suspect of having cancer – 75% of patients found to have cancer are referred after only one or two GP consultations, and in the last five years the proportion of cancers diagnosed as an emergency has dropped from 25% to 20%, and a higher proportion of patients are being diagnosed at an earlier stage of the disease.’

Dr Jem Rashbass, medical director at Health Data Insight, said: ‘Large studies like this are only possible because anonymous data on large numbers of cancer patients is available for research through the NHS. Our idea is to use this unparalleled information on prescription data and other information to better identify patients for referrals or follow-up.’

Have you registered with us yet?

Register now to enjoy more articles and free email bulletins

Register

Already registered?

Sign in

Follow Us:

Just published

Subject access requests to GP practices increased by a third under GDPR

Subject access requests to GP practices increased by a third under GDPR

The number of subject access requests (SARs) GP practices receive each month has...

RCGP criticises GP at Hand video showing antibiotics prescribed for sore throat

RCGP criticises GP at Hand video showing antibiotics prescribed for sore throat

A promotional video for Babylon GP at Hand that shows a patient with a sore throat...

Review into overprescribing aims to give GPs power to challenge hospital scrips

Review into overprescribing aims to give GPs power to challenge hospital scrips

A government review of overprescribing in the NHS could see GPs given more power...

Submit your session ideas for the RCGP Annual Conference 2019

Submit your session ideas for the RCGP Annual Conference 2019

GPs can now submit ideas for sessions at the RCGP Annual Conference in Liverpool,...

Scottish GP workforce increases for first time in 10 years, figures show

Scottish GP workforce increases for first time in 10 years, figures show

GP numbers in Scotland have risen slightly for the first time in 10 years despite...

More than 16m GP practice appointments a year lost to DNAs

More than 16m GP practice appointments a year lost to DNAs

More than 16m appointments at GP practices are lost every year because patients fail...