Improving cancer survival by understanding how patients are diagnosed

The NHS’s improving data collection over the last few years has revealed a lot about where problems lie in the healthcare system – and potentially how to fix them.

In particular, it’s helped us examine a really important issue in cancer – understanding how the different ways people can be diagnosed – such as by screening, by a GP referral, or as an emergency – affect a patient’s care and chances of survival (we recently wrote about this work here).

And perhaps unsurprisingly, previous analyses have shown that those diagnosed in an emergency often have worse survival.

But proving why has been tricky.

The obvious assumption is that it’s caused, at least in part, by those who are diagnosed in an emergency being more likely to have later stage cancers (diagnoses are divided into four stages – Stage 1 being the earliest, and Stage 4 being the most advanced).

But we didn’t have the data to know whether or not that was true. That is, until today.

New figures we’ve released, in partnership with Public Health England’s National Cancer Intelligence Network, have finally confirmed the long-standing suspicion that people diagnosed in emergencies are indeed much more likely to be diagnosed at a later stage.

It’s a crucial finding. But before we discuss why, let’s take a closer look at the data.

The analyses

For the first time, we’ve been able to look at both how people are diagnosed (e.g. via a GP referral or by screening) and whether they were diagnosed at an early or late stage.

The information we had available was from patients diagnosed with ten cancer types: bladder, breast, bowel, kidney, lung, melanoma, non-Hodgkin lymphoma, ovarian, prostate and uterine cancers, between 2012 and 2013. These were years from which good national data on the stage patients were diagnosed were available.

We’ve also been able to look at the effect a patient’s sex, age, socioeconomic status and ethnicity has on where, and at what stage, they’re diagnosed.

The overall picture

Of the total 574,500 cases analysed, screening picked up the highest proportion of early stage cancers – 63 per cent Stage 1, versus three per cent Stage 4.

And of those patients diagnosed by GPs, either via routine referrals or the ‘two-week wait’ urgent referrals for suspected cancer, just over a third (34 per cent) were stage one compared to just over a fifth (22 per cent) at stage four.

But when it came to cancers diagnosed as an emergency, a tenth (11 per cent) were stage one, compared to more than half (58 per cent) diagnosed at stage four.

And when you put the data together for all cancers, you get this overall picture:

Share a link to this graphic

Of course, the figures showed a wide variation across cancer types.

For some types, such as lung and ovarian cancer, the majority of people diagnosed in emergencies are at a late stage.

Conversely, for others – such as bladder and uterine cancer – people diagnosed in emergencies have more of a chance of being diagnosed early. There’s a caveat here though, in that the data for these contain a large proportion of ‘unknowns’, which might be more likely to be late stage cancers).  

This variation is likely to be due to differences in how these cancers develop, and the way their symptoms appear. So more research will need to be done to understand exactly what’s causing these differences – and what to do about them.

But altogether, our findings confirm that the diagnosis ‘routes’ that are linked to a better chance of surviving the disease – such as by screening and GP referrals – are also the ones where patients are diagnosed earlier.

You can see this particularly clearly in the case of bowel cancer, where the national screening programme has been running for several years, but where a substantial proportion of patients are still diagnosed in an emergency:

Share a link to this graphic

As a reminder of why this is important, here’s how bowel cancer survival varies by stage:

Share a link to this graphic 

Why’s this important?

Diagnosing cancer earlier saves lives, because it increases the options for treatments that can cure. So this information is essential to help understand where improvements can be made, and where the health service should target resources, and help more patients beat their cancer.

That’s why we were pleased to hear Public Health England announce that emergency presentation data and stage at diagnosis data will be presented quarterly to healthcare commissioners (the people who plan and decide where the health care funding goes).

The differences in how cancers, and the symptoms they cause, develop mean that in some cases it may be impossible to avoid an emergency diagnosis. But, for many cancers, there should be ways to make them much less likely to happen.

But there are other ongoing projects that are also looking at the issue of late diagnosis. For example, as part of the National Awareness and Early Diagnosis initiative, Cancer Research UK is funding onging research to help the NHS diagnose cancers earlier and improve survival, which you can read more about here.

There is also some really promising work underway nationally.

• The National Screening Programme recently recommended changes to the bowel screening programme though the introduction of a test called FIT – we looked at some impacts the change in test may have here. The next step is to make sure this brought in swiftly, according to a clear plan.
• The ongoing Accelerate, Coordinate, Evaluate (ACE) programme is looking at more innovative ways for doctors to refer patients for tests, to help ensure that GPs can manage and refer their patients in the optimal way.

• The recent Cancer Strategy for England recommended that all patients diagnosed as an emergency have their cases looked at through a system called . Significant Event Audits – in hopes of learning lessons and making improvements.

• And NICE recently revised its guidelines for GPs on making urgent referrals for potential cancer symptoms and have made it easier for GPs to refer patients to specialist or for tests when they have concerns.

This final point is a positive step, but relies on diagnostic services in the NHS being given the resources they need to meet the extra demand that will create. That’s a really key issue because we know there are services, which are already under significant pressure.

What these examples highlight is that in order to reduce late stage cancer diagnosis we need to tackle a wide range of issues. Each of the insights we gain from them will be vital steps towards understanding how to improve healthcare, and the ways in which cancer is diagnosed.

So what happens now?

The data published today takes us a step towards understanding how to give cancer patients the best possible chance of surviving the disease.

Now, these data need to be used to work out priorities for action, to make real inroads against the issue of late diagnosis. In some cases, that will mean redoubling efforts on existing programmes, such raising awareness of cancer screening programmes, or encouraging people to see their GP if they spot unusual or persistent changes to their bodies.

In others, it will mean more research to understand how cancer develops and what to do about those cancers that are more likely to be diagnosed via an emergency route.

At Cancer Research UK, we’re dedicated to understanding what needs to be done to diagnose cancer earlier, and making sure it happens. These data are a vital step towards that.