Login to Pancreatic Cancer UK

news hero image

New NICE Referral Guidelines for Suspected Cancer

Posted by: Policy and campaigning 23 June 2015

Today the National Institute for Health and Care Excellence (NICE) has published new Referral Guidelines for Suspected Cancer. (New Scottish Referral Guidelines were introduced last year).

View our short statement or read on to find out in more detail what this means for pancreatic cancer patients.


Late last year the National Institute for Health and Care Excellence (NICE) launched a consultation on when and how GPs should refer patients if they show symptoms that could be cancer. You can see our December blog post here which gave our initial reaction to the consultation document.

 We have been calling for an updated version of the guidelines for some time – the current ones in use are ten years old - and the All Party Parliamentary Group (APPG) on pancreatic cancer also made a specific recommendation for the Guidelines to be updated. So we were pleased when it was announced last year that new Guidelines would be published in 2015. In our view a beefed up version of the referral guidelines was long overdue and could mean more GPs making more, and earlier, referrals for pancreatic cancer.

 In January, we submitted a response to the consultation jointly with Pancreatic Cancer Action. We argued that the Draft Guidelines, as they stood, would not lead to the larger number of earlier referrals from GPs that we believe are necessary. You can read our responses on different issues - and see NICE's comments back - via this link. (Scroll to page 240).

 The new guidelines:

 We are extremely disappointed to report that the final version of the NICE Guideline, published today, has not taken on the board the comments we made. Whilst NICE are adamant that there will be an increase in the number of referrals for suspected pancreatic cancer, we feel the Guidelines do not go far enough towards addressing the problem of so many pancreatic cancer patients being diagnosed when the disease has spread beyond the stage for curative surgery, or indeed effective life-extending treatments, to be an option.  We are concerned that the new Guidelines will not address the current unacceptable situation where half of all diagnoses are made as a result of an emergency admission route.

We do support some measures in the new Guidelines, for example:

  • A separate section on pancreatic cancer for the first time.
  • A symptom specific section, allowing GPs to refer to symptoms, or clusters of symptoms, rather than having to check different sections relating to different cancer types.
  • An additional symptom for GPs to consider that could be due to pancreatic cancer, namely new onset diabetes.
  • GPs being encouraged to consider ‘an urgent direct access CT scan’ for patients presenting with the symptoms chosen by NICE. This is because a CT scan can image the whole of the pancreas whilst ultrasound can only image the head, and ultrasound can produce both positive and false negative results. (You can read more about CT scans on our information pages here)
  • Recognition that more CT scans will mean the NHS needs to build capacity in imaging services across the country.
  • Recognition of the need for more research for symptoms of pancreatic cancer within a primary care setting.

 However, we have great concerns that the Guidelines do not go far enough in addressing the problem of delays in diagnosing pancreatic cancer.

In summary, our concerns are:

  • Exclusion of some common symptoms of pancreatic cancer

NICE have only accepted one particular measure for evidence of symptoms that should trigger a referral. This is called  Positive Predictive Value (PPV). We feel this approach is not flexible enough, in part due to a lack of research into pancreatic cancer symptoms that resulted in PPV scores being published.  We wanted NICE to take into account other peer-reviewed research that used different measures of relevance instead of PPV to be considered. We felt this would have led to more symptoms (e.g. non-responsive dyspepsia) or clusters of symptoms being included as warning signs).  Given that pancreatic has the worst survival outcome of any of the most common cancers we feel this exceptional approach would have been justified.

  • Inclusion of age-thresholds

The fact that age-thresholds have been introduced is of great concern: again due to the use of strict PPV research data, the Guidelines state that only patients aged 60 or over who present with weight loss combined with one or more symptom of diarrhoea, back pain, abdominal pain, nausea/vomiting, constipation, or new onset diabetes warrant a referral for a CT scan. Likewise an age-threshold of 40 has been set for patients showing jaundice to be referred down a cancer pathway. There were an average of 1,236 newly diagnosed cases of pancreatic cancer each year between 2009-2011 across the UK. And our own online Symptoms and Diagnosis Survey responses suggested, around nearly 60% of patients would not have been referred for a CT scan under the cluster of symptoms route, as they either did not have weight loss AND at least one of the other symptoms chosen by NICE, or they did have those symptoms but were under the age threshold of 60 years old selected by NICE. (You can read more on this issue in our Symptoms and Diagnosis Survey document, pages 10 and 11).

  • The PPV threshold is too high

If PPV has to be used by NICE, we wanted the threshold of a 3% score for all cancer types in adults to be lowered for those cancers with the poorest survival outcomes, like pancreatic cancer, to perhaps as low as 1%. This would have resulted in a more comprehensive list of symptoms being included and the possible reduction in age thresholds.

 As such, we do not feel that the Guidelines go far enough and believe that this represents a missed opportunity.

Of course, guidelines are just guidelines and GPs may still refer patients on for scans or other investigations if they see fit, based on their own clinical judgement, experience and intuition. However, too many patients are already returning to their GPs on a too frequent basis before finally being diagnosed - as the answers to our Symptoms and Diagnosis Survey shows, together with the fact around half of all pancreatic cancer diagnoses are made via an emergency admission route. We need to see stronger referral guidelines than those included in these new guidelines.

So, what is next?

 There appears to be no other opportunity for consultation on these referral guidelines. So, the best we can hope for is that there will be an early review of the pancreatic cancer section of the guideline in response to new evidence, in isolation from the rest of the Guidelines documents if necessary. We will certainly be lobbying for this as we do not want to have to wait another 10 years for new guidance to be issued.

And if the Guidelines are to be reviewed soon, we need to make sure that new evidence exists in the format NICE want. So if PPV is the only measure NICE will accept, we need to see more research carried out into symptoms of pancreatic cancer in primary care that reports its findings in that format. (We believe there needs to be more research into pancreatic cancer symptoms and early diagnosis more broadly as well).

 In the meantime, we need to do more to ensure that GPs have the appropriate level of training and support – for instance through the use of computerised decision aid tools – to help them make more and earlier referrals which will in turn lead to more cases of pancreatic cancer being diagnosed at a less advanced stage.