Improving English cancer survival rates

Last week, in exploring the new contract between the public and the NHS, I suggested some outcomes that the NHS can and cannot achieve. We should be ready to sign up to the former and, as a society, help with the latter.

As far as I can tell everyone agrees that the NHS should be making a pledge to improve cancer survival rates. English rates are improving, but we still have some way to go to save the number of lives that science tells us we should be saving.

Whilst brilliant work will continue on the science that will help improve treatment, there is agreement that our lower survival rates are caused by two main problems which need solving – late presentation by people who think they have cancer and late diagnosis of those that come forward.

In terms of late presentation, there are cultural issues here – if working class men felt and talked about pain the way in which middle class men did, it would be a breakthrough. But that will take interventions outside of the scope of the NHS to change.

There is however one important policy and practice that the NHS can make in developing early presentation – and that is the general gratitude that so many people feel about the work of their doctors. Obviously that’s a great thing – people should have decent relationships with the staff that help them. But the problem is that too many people feel that they should not bother their NHS with pains and lumps. They don’t go and see the doctors with their pain because “I know you are busy and I don’t want to bother you”. For too many people this attitude results in them not going to see their doctor and ending up being diagnosed at A&E when they go to hospital because it has all become too much. And that is usually very late.

The NHS can play a big role in changing that relationship. It can make it very clear in poorer areas that it wants to be bothered. That bothering us with cancer fears makes it much more likely that we can carry out our core business of saving lives.

This then is down to the thousands of interactions between patients and doctors. If we improve them the NHS can play a role in encouraging early presentation.

So the NHS can do something about late presentation. But we can do a great deal more about late diagnosis. Diagnosis is our core business and something for which these extra NHS resources should be used to improve outcomes for the public.

All diagnosis is a combination of hard science – scans, chemicals, blood tests – and the application of medical judgement about those tests. Over the last 20 years we have organised diagnostic testing to be much swifter and streamlined. We now carry out many many more tests than we did 20 years ago.

The mathematics of diagnostic tests for cancer is that nearly all of the tests come back negative. And in order to have earlier diagnosis we are going to have to have even more that come back negative. So the core practice of getting early diagnosis is to encourage primary care staff to get more people tested. And for everyone involved to be clear that we want to increase the number of negative tests. At first sight that might appear odd In an organisation trying to become efficient, but it is an effective way of increasing the crucial KPI for cancer survival rates.

Over the last 20 years there has been some streamlining of the diagnostic process to move the tissue from the initial patient interaction, through the decision to the report of that decision, and it is this streamlining that we need to make ever more swift.

During the coming decade (the time covered by this pledge) the NHS is very fortunate that science is moving into an area of medicine that was previously carried out exclusively by doctors’ judgements. This is the rapidly developing area of artificial intelligence (AI).

This is not the future but an extension of the present. In 2018 scans are already being assessed by artificial intelligence. As computers carry out more and more of this work they build a better and better database to make better and better judgements. These have proved even more accurate than judgements made by doctors and we can now ensure that we can carry out more diagnoses at a faster rate.

What must happen over the next ten years is the application of AI to the decision making process. This may not be for 100% of judgements, but what we know of AI is that the more it is asked to do the better the database that it uses becomes.

Given the growth in the number of procedures that will be necessary for us to keep