The secondary care diagnostic interval (SCDI) is the period of time, in days, from a patient’s first interaction with secondary care, which can be a referral, appointment or diagnostic test, to their cancer diagnosis. A short SCDI is not necessarily better, as it may represent the presence of clear symptoms or advanced disease, which result in a quicker cancer diagnosis.
In partnership with Cancer Research UK, PHE has published new data showing the SCDIs for cancer patients in England diagnosed in 2014-2015. The data, from PHE’s National Cancer Registration and Analysis Service (NCRAS) was used to calculate this interval for the first time for nearly all patients with 25 different cancers – including breast, colorectal and liver cancer.
SCDI data can be viewed and downloaded from an online tool, which displays different cancer types by demographic factors and Cancer Alliance. Using this tool, we are now able to understand how these intervals vary across different stages of disease, age, sex, ethnicity, comorbidities, levels of deprivation, routes to diagnosis and in different parts of England, to support initiatives to diagnose cancer faster within the NHS.
This blog outlines 7 things we have learnt from the data.
1. Patients diagnosed after a routine GP referral or via outpatients experience the longest intervals
There is considerable interval variation by different routes to diagnosis. Patients diagnosed via routine GP referrals and outpatients have the longest intervals for all cancers, with the exception of acute lymphoblastic leukaemia, for which Two Week Waits and routine GP referrals are the longest. Patients diagnosed via emergency and inpatients have the shortest intervals for all cancers, apart from breast cancer, for which where emergency and screening patients have the shortest intervals.
Patients diagnosed following an emergency are typically more unwell with more pronounced symptoms, leading to quicker diagnoses. This route is associated with worse outcomes.
2. Late stage cancers have shorter intervals
For all cancers, the interval length generally decreases at more advanced stages, except myeloma and leukaemias (where we don’t have the stage of disease). For example, stage 4 cancers have shorter intervals than stage 1 for all cancers, with the exception of melanoma.
Possible reasons for this include the severity of advanced stage symptoms or attending A&E whilst being acutely unwell, leading to a faster diagnosis.
3. Diagnostic pathways differ between cancer sites
Diagnostic intervals differ by cancer. This can, in part, be due to different cancers having varying diagnostic procedures, some of which happen earlier in the pathway. For example, chest X-rays for lung cancer can be requested via a patient’s GP before a secondary care referral, whereas diagnostics for other cancers happen after secondary care referral, such as colonoscopies for colorectal cancer.
Breast cancer has shorter intervals, as there is a well-defined diagnostic pathway with patients usually presenting with distinctive symptoms, such as a lump in the breast, leading to swift referrals and diagnoses.
4. Patients with multiple comorbidities have longer intervals
The Charlson comorbidity index measures the presence and severity of other diseases a patient has prior to their cancer diagnosis. Generally, patients with comorbidities experience longer intervals than those with none. Possible explanations are that cancer symptoms may be confused for existing health conditions, or that these conditions result in patients being less fit for diagnostic procedures.
5. There is some variation across England by Cancer Alliance
Cancer Alliances drive local changes to cancer services to improve cancer outcomes and patient experience. By bringing together clinical leaders and teams in each of Alliance, they aim to transform treatment and diagnosis in their areas – there are 19 Cancer Alliances across England.
Comparing Cancer Alliance intervals with the national average shows geographical variation, with most sites having a number of Alliances with different intervals.
Prostate is the cancer with the most significant differences from the national average interval. A possible reason for differences is a Cancer Alliance's population characteristics could be different to the overall England population. Bladder cancer, acute lymphoblastic leukaemia and chronic myeloid leukaemia are the only cancers where there are no significant differences between England and any Cancer Alliances.
6. Age, sex and ethnicity have inconsistent patterns of variation
We have shown interval length variation by age, with older patients generally having longer intervals and the oldest age group (over 85 years) having shorter intervals. The latter is at least partly due to more people in this age group being diagnosed via an emergency. Some cancers have no variation by age, including breast, cervical and oesophageal cancers.
For most cancers, the interval length is similar for both sexes. However, males have longer intervals for bladder, kidney and liver cancers, and females have slightly longer intervals for chronic myeloid leukaemia and pharyngeal cancers.
There has been no discernible patterns by ethnicity shown to date. Looking at different cancers, there are some variations but no clear pattern of longer or shorter intervals for any particular ethnic group.
7. There are no differences by deprivation levels
There is no variation between deprivation and interval length for any cancer.
The next steps
This work enables further research, both by PHE and by external researchers, to examine the variation in SCDIs in more detail. It is important for us to understand and address the causes of variation, and identify drivers of unnecessarily long intervals to improve outcomes for patients with cancer.
For more details about our methods, we have produced a Standard Operating Procedure (SOP).
Find out more on the NCRAS landing page.