Official Statistics

Chemotherapy, Radiotherapy and Surgical Tumour Resections in England

Published 8 April 2020

1. Main points

For cancers diagnosed between 2013 and 2016 in England, of those receiving at least one of the main treatment types, 28% were treated with chemotherapy, 27% with radiotherapy and 45% with surgery, with some cancers receiving a combination. A third of diagnoses had no record of receiving any of these most common treatment types.

For cancers diagnosed between 2013 and 2016 in England, 39% were treated with one of chemotherapy, radiotherapy or surgery, 22% were treated with 2 of these, and 7% with a combination of all 3.

For cancers diagnosed at a later stage, chemotherapy was used more commonly than for early-stage diagnoses; less surgery was used for later-stage diagnoses.

Cancers diagnosed in younger patients are more likely to be treated than those diagnosed in older patients. For example, 76% of cancers in patients aged under 50 are treated with surgery, compared to 23% of cancers in patients aged 80 or over.

2. Treatment proportions and their variation

2.1 Patients may have multiple types of treatment

• the proportion of cancers treated with surgery only is 22%
• the proportion of cancers treated with radiotherapy only is 10%
• the proportion of cancers treated with chemotherapy only is 7%
• the proportion of cancers treated with surgery and radiotherapy is 9%
• the proportion of cancers treated with surgery and chemotherapy is 8%
• the proportion of cancers treated with radiotherapy and chemotherapy is 5%
• the proportion of cancers treated with surgery, radiotherapy and chemotherapy is 7%
• the proportion of cancers with no record of being treated with these most common treatment types is 33% (reported as ‘other care’)

Examples of ‘other care’ include hormonal therapy or management of symptoms.

2.2 Cancers diagnosed in older patients are treated with surgery less often

As age at cancer diagnosis increases, the proportion of patients receiving each treatment decreases. For example, 49% of cancers diagnosed in those aged under 50 are treated with chemotherapy, compared to only 9% of cancers diagnosed in those aged 80 or over. Similarly, the proportion of cancers treated with radiotherapy decreased from 35% to 14% between these age groups, and the proportion of surgeries dropped from 76% to 23%.

2.3 Different cancer sites have different treatment proportions

Treatment proportions for specific cancer sites will differ due to differing typical practice between cancer types. For chemotherapy, small cell lung cancer has the highest proportion (68%) and prostate has the lowest (4%). Cancers of the head and neck (specifically the oropharynx, base of tongue, and tonsil) are treated with the highest proportion of radiotherapy (83%) and ovary, fallopian tube and primary peritoneal carcinomas have the lowest proportion (2%). The highest proportion of cancers treated with surgery is those of the uterus (84%), compared to only 2% of small cell lung cancers.

2.4 Cancers diagnosed at later stages are treated with less surgery and more chemotherapy

Stage at diagnosis affects the proportion of some treatments, with proportions of surgery decreasing from stage 1 to stage 4, while proportions of chemotherapy increase. Comparing all invasive cancers, 69% of those diagnosed at stage 1 are treated with surgery, compared to 13% of stage 4 diagnoses. Proportions of radiotherapy are similar across stages, with 27% of those diagnosed at stage 1 treated with radiotherapy, compared to 25% of those diagnosed at stage 4. Chemotherapy proportions increase from 12% of stage 1 diagnoses to 40% of stage 4 diagnoses.

2.5 Cancers that are not sex-specific tend to see similar treatment proportions between females and males

In general, cancers occurring in both sexes see similar treatment proportions, for example, pancreatic cancer diagnoses were treated with similar proportions of surgery between males (10%) and females (9%). However, proportions varied by sex when considering all cancers together, with 59% of all female cancers treated with surgery compared to 31% of all male cancers. This is likely to be mostly driven by differing typical practice between 2 of the most common cancer types, as 81% of breast cancers were treated with surgery compared to only 16% of prostate cancers.

2.6 Apparent differences between the most and least income deprived may be driven by different cancer types in these groups

Proportions of chemotherapy and radiotherapy show little variation by level of income deprivation. For surgery, there is more variation, as proportions reduce from 49% to 40% between the least and most deprived groups. Some of this variation is likely explained by the different mix of cancers in more deprived groups, as discussed by assistant researcher and statistician Barclay, M in The European Journal of Public Health.

2.7 Apparent differences between White and non-White groups may be driven by different case-mix and demographic characteristics in these groups

When considering all cancers, treatment proportions differ between White and non-White populations. However, these differences can be largely explained by accounting for variation in cancer type and differences in the age at cancer diagnosis and the sex of patients. This is discussed in a paper by Toral Gathani, senior clinical research fellow in clinical epidemiology and consultant surgeon at the Nuffield Department of Population Health (currently awaiting publication).

2.8 Cancers diagnosed in patients of worse health are less likely to be treated

Treatment proportions decrease as a patient’s Charlson comorbidity index (CCI) increases. The CCI is a quantitative measure of a patient’s other diseases, where a higher index means a patient is in worse health. It uses the number and seriousness of a patient’s other diseases to quantify comorbidities. Comparing patients with no reported comorbidities to those with an index of 3 or more, the proportion of chemotherapy drops from 31% to 13%. Similarly, the proportion of radiotherapy reduces from 29% to 15% and the proportion of surgery reduces from 48% to 25%.

3. Things you need to know

You can explore these statistics in an interactive tool.

These statistics are the proportion of cancers being treated with the relevant treatment or treatment combination, out of the total number of cancers that could be treated with that treatment. For more information, see Quality and Methodology.

Treatment is considered related to the diagnosis of cancer if it occurred within either 6, 9, 12, 15 or 18 months of the diagnosis, depending on the cancer site. For full details, please see Appendix 2 of Linking treatment tables – chemotherapy, tumour resections and radiotherapy.

For some cancers, there is no record of chemotherapy, radiotherapy, or surgery in the national administrative datasets. This may be because they were treated with forms of care other than the 3 most common types, or a patient may have received treatment in a private healthcare setting. Alternatively, there may be data missing, for example, collection of certain chemotherapy data, such as Systemic Anti-Cancer Therapy (SACT), only became mandatory in April 2014 and so data from before this point may be of lower completeness.

The proportions presented here are not adjusted for any factors that may affect treatment. These include stage at cancer diagnosis, age at cancer diagnosis, sex, income deprivation, broad ethnic group and the presence of other illnesses.

The proportions are available as treatment combinations and as total proportions for each treatment type. The demographics tab allows users to explore variation by cancer type and a choice of:

• age at cancer diagnosis
• broad ethnic group
• Charlson comorbidity index (CCI)
• income deprivation
• sex
• stage at cancer diagnosis
• year of cancer diagnosis

The geography tab allows users to explore variation by cancer type and Cancer Alliance (CA), alongside those for the whole of England. These variations could be because the demographics of populations vary between CAs. For example, a patient’s age affects the treatment they receive, so if the CA has a different age distribution the treatment proportions may differ. Variations could also be because there are different levels of missing data both between CAs and compared to England.

4. Links to other related statistics

This work is used in several Cancer Research UK outputs.

Other statistics, including incidence, prevalence, survival and mortality, are available on the wider CancerData site.

The Data Resource Profile of cancer registration data describes the data which these flags are created from.

5. Quality and methodology

For each cancer case, a flag is created for chemotherapy, radiotherapy or surgery if there is evidence of that treatment. Flags are created regardless of the order of treatments, or if treatments were delivered together. If a cancer case has no evidence of any of these treatments it is classed as ‘other care’.

Percentages are taken from the total number of cancers that could be treated with that treatment. For chemotherapy and radiotherapy, the total number of cancers includes all malignant cancers excluding non-melanoma skin cancer. For surgery, the total is all cancers for which we have groups of clinically relevant surgical codes defined. All sites for which we haven’t defined surgical code groups are classed as ‘other malignant neoplasms’. To calculate surgery proportions, all cancers with site code ‘other malignant neoplasms’ should be excluded. We are working to extend these flags to cover more sites and reduce the size of the current ‘other malignant neoplasms’ category.

Statistics were calculated using: cancer registration data, the Systemic Anti-Cancer Therapy dataset (SACT), RadioTherapy DataSet (RTDS) and inpatient Hospital Episode Statistics (HES).

If a patient has only one cancer, then all the above datasets are used to identify treatment data. If the patient has 2 or more cancers, all datasets are used only if the cancers are more than 18 months apart. If any 2 cancers are within 18 months of each other, only cancer registration data that has been quality assured and linked at tumour level is used. Other datasets are linked at patient level, so where a patient has multiple cancers these datasets are excluded to avoid the risk that treatment delivered for the incorrect tumour is used.

6. Authors

Responsible statistician: Emma Hope.

Production team: Jess Fraser, Emma Hope, Jack Anderson, Wouter Verstraete, Ravneet Sandhu and Sean McPhail.

For queries relating to this bulletin contact [email protected].

7. Acknowledgements

This work uses data that has been provided by patients and collected by the NHS as part of their care and support. The data is collated, maintained and quality assured by the National Cancer Registration and Analysis Service, which is part of Public Health England (PHE). Initial work was undertaken as part of the Cancer Research UK – Public Health England partnership.