(UPDATED 19th June 2024 with more recent Routes to Diagnosis data)

The most common route to receiving a cancer diagnosis begins (as we would hope) with a referral in which a risk of cancer has already been identified, according to the Routes to Diagnosis project. These are the Urgent Suspected Cancer (USC) referrals (previously known as Two Week Wait/TWW referrals), who (as the name suggests) are protected by the statutory cancer waiting time targets in the English NHS Constitution.

The threshold for using the USC referral pathway is set very low at just “three per cent or higher risk of cancer” – in other words, GPs should use USC referrals so frequently that 29 out of 30 USC patients will get the all-clear at diagnosis. Even so, when we look at all patients who eventually receive a cancer diagnosis, less than 40 per cent come through the USC route.

The second most common route in 2019, comprising 22 per cent of cancer diagnoses, is an ordinary GP referral. These are not protected by the cancer waiting time targets, but instead fall under the more general referral-to-treatment (RTT) waiting time target which states that 92 per cent of the RTT waiting list (incomplete pathways, in the jargon) should have waited no longer than 18 weeks. But that target has not been met since 2016, and waiting times are even longer following the pandemic – far too long if you have an undiagnosed cancer.

We can use the Routes to Diagnosis figures to estimate roughly how many patients are on the RTT waiting list, who are not on a USC pathway but will eventually receive a cancer diagnosis, using a four step calculation.

Firstly: we use the Routes to Diagnosis data to estimate how many routes to a cancer diagnosis began with a non-USC GP referral, over the most recent available time period that is unaffected by the covid lockdowns (the calendar year 2019). The methodology paper for the Routes to Diagnosis study says that, as well as the GP Referral category, “it is possible that the majority of cases in the Other Outpatient and Inpatient Elective routes were originally initiated by a GP Referral”. So let us include in our count:

• all patients in the GP Referral category (71,796 in 2019 according to the download data),
• half of those in the Other Outpatient category (half of 31,461), and
• half of those in the Inpatient Elective category (half of 5,013),

which adds up to an estimated 90,033 non-USC GP referrals receiving an eventual cancer diagnosis in 2019.

Secondly: express that as a proportion of all referrals to consultant-led services over the same time period. In 2019 there were 20,797,995 new RTT periods (including estimates for missing data), so non-USC GP referrals who received cancer diagnoses amounted to about 0.433 per cent of all referrals to consultant led services.

There is an argument for dividing by the number of unique patients referred, rather than the number of referrals (there are about 16 per cent fewer unique patients than referrals, because some patients are referred for more than one condition). On the one hand, the Routes to Diagnosis study uses the NHS number to link the various datasets, which relates to unique patients; however I am not clear from the description whether the tracking of outpatient episodes avoids this issue by using pathway identifiers. On the other hand it could be argued that multiple referrals signify higher risk than a single referral per patient; this is clearest in cases where a patient has one referral for a symptom that is unlikely to be caused by cancer (cataracts, for example), and another referral for a symptom that is more likely to be caused by cancer (coughing, for example). Because of those indications that referrals may be the better indicator, I will use referrals in these estimates.

Thirdly: convert the proportion of referrals into a proportion of the pre-diagnosis waiting list. This will turn out to be an easy calculation, but we have to know what we are doing because the logic is important. On the grounds that these are non-USC referrals for whom cancer is not yet suspected, let us assume that their management is no different to all the other referrals. They therefore leave the pre-diagnosis waiting list at the same rate as any other patient, throughout their wait. So they will comprise 0.433 per cent of every pre-diagnosis waiting time cohort, and therefore comprise 0.433 per cent of the total pre-diagnosis waiting list.

Fourthly: multiply that proportion by the latest estimated number of pre-diagnosis patients on the waiting list. At the time of writing, the most recent available figures come from the Incomplete Commissioner data tables for April 2024. The total RTT incomplete pathways are on the ‘National’ worksheet (7,572,563 patient pathways), and the ‘with Decision to Admit’ (DTA) incomplete pathways are on the ‘National with DTA’ worksheet (1,168,803). The difference is those without a DTA, which is our best estimate of those without a diagnosis: 6,403,760 patient pathways. We multiply this by 0.433 per cent to get our answer.

The result (27,721 at the end of April 2024) is the estimated number of patients on the latest waiting list who will eventually receive a cancer diagnosis, but whose cancer is not yet suspected. They are not protected by the cancer waiting time targets, and will therefore wait as long as everyone else for a diagnosis and decision, currently 42.6 weeks nationally (using decison to admit as our best proxy for diagnosis and decision). Because this is far too long to wait for a cancer diagnosis, I argue that waiting times to diagnosis and decison should be reduced first.

To repeat the calculation when future data is published, the sum is:

( (Total RTT incomplete pathways) – (‘with Decision to Admit’ incomplete pathways) ) * (90,033 / 20,797,995)

The latter two numbers are from 2019 and will change when the Routes to Diagnosis study is next updated.

(I am indebted to Carl Baker at the House of Commons Library for pointing out that the Other Outpatient and Inpatient Elective categories should be partially included in the non-USC GP Referrals, for raising the issue of unique patients, and for signposting the updated Routes to Diagnosis data.)