Suspected cancer pathway: quality report

Data description

Data are published showing patient pathways closed – either because a patient started their first definitive treatment or because they were informed they did not have cancer – and for the proportion of pathways where treatment started within 62 days from cancer first being suspected, the Welsh Government’s performance target.

Until January 2021, cancer services statistics were reported on three bases: the single Suspected Cancer Pathway, the urgent suspected pathway and not via the urgent suspected pathway.

Targets for the urgent and non-urgent pathways ceased in December 2020 (February 2021 publication) and no new data has been collected or published since. Historical data remain available on the StatsWales website.

From February 2021, data are only published for the Suspected Cancer Pathway (SCP), with a two month delay.

The SCP starts at the point of suspicion (for example when a GP makes a referral) and this is when recorded waiting times start. Pathways are closed, and waiting times end, when patients start their first definitive treatment or are told they do not have cancer (downgraded). Pathways where patients die or choose not to have treatment are also closed, but are not included in the closed pathways data. This is because the statistics are intended to capture NHS ‘activity’.

Data are published on StatsWales with breakdowns for age group, sex and tumour site.

Statistical concepts and definitions

The suspected cancer pathway statistics cover three headline measures:

• number of patient pathways informed they do not have cancer
• number of patient pathways that started their first definitive treatment
• number and percentage of patient pathways that started their first definitive treatment within the 62 day target time

Data are also published for the number of patient pathways opened on the suspected cancer pathway. All new patient pathways are included in open pathway measures regardless of their source of suspicion. This includes patient pathways that were referred to secondary care in Wales but may receive treatment outside of NHS Wales (either in a different country or in private hospitals) but does not include patient pathways with a recurrence of the original primary cancer.

The data used to derive the 62 day target measure do not include some cases, known as suspensions. A patient may be suspended from the waiting list for medical reasons, for example because they are unfit to undergo treatment, or for social reasons such as going overseas for a period of time.

Statistical unit

Suspected cancer pathway data reflect patient pathways rather than unique patients. This is because a single patient can have multiple cancer pathways opened if cancer is suspected in more than one tumour site.

Raw data suggest the number of pathways opened is approximately 2 to 3% higher than the equivalent number of individual patients suspected of having cancer.

Statistical population

The data cover all recorded cancer services in Wales, with information collected by local health boards. Powys Teaching health board does not provide acute cancer services but returns information on patients who have entered the pathway with a suspicion of cancer. When patients start treatment, they are included against the health board they are referred to.

The suspected cancer pathway data collection includes some patients who were referred to secondary care in Wales but who receive treatment outside of NHS Wales (either in a different country or at private hospitals – including those who initially started their pathway privately and later transferred to NHS care).

Time coverage

Data are published on a monthly basis. Published data for closed SCP pathways start in June 2019 and data for open pathways start in December 2021.

Source data

The suspected cancer pathway data are collected from local health boards through the National Data Resource (NDR), hosted by Digital Health and Care Wales (DHCW). DHCW provide an aggregate of administrative patient level data, submitted by each health board, to Welsh Government.

Frequency of data collection

An aggregated data extract is provided to the Welsh Government between the fourth and fifth working day of each month, with an approximately six week lag from the reference month. For example, data for January data would be provided in early March.

Data collection

Local health boards collect live administrative patient cancer services data locally on a daily basis. Local health board data are collated by DHCW through a central data collection process each month.

Most health boards use a cancer tracker system on the Welsh Patient Administration System (WPAS) which is updated daily. Others use local systems to capture administrative data and provide extracts to DHCW at month end, using SQL.

Once validated, DHCW provide an aggregated extract to the Welsh Government.

Data validation

Data validation occurs at multiple stages before publication:

Local Health Boards undertake their own validations including at the point of entry and weekly or monthly checks for missing, duplicate or outlier values. There are some differences in the types of validations carried out by LHBs, largely because of underlying differences in administrative data systems.

LHBs submit patient level data to DHCW and DHCW undertakes further validations, namely checks that data entries match values outlined in the latest SCP Data Standards Change Notice (DSCN) (DCHW) and data set structure (NHS Wales Data Dictionary). DHCW compiles a (non-public) dashboard showing summary analyses at LHB level. LHBs sign off their own data based on the summary analyses in the dashboard.

DHCW provides aggregate data to Welsh Government. Welsh Government compiles analyses in the format required for publication and conducts some final validations in the form of checks for outliers as well as sense checking aggregate numbers and trends. Any anomalies identified are queried with DHCW who then investigate and verify with LHBs.

Data compilation

DHCW compiles aggregate analyses from the patient level data supplied by LHBs. Patient identifiers are removed to anonymise the data and the aggregated data are validated against the original extract. Aggregates are provided to Welsh Government, where breakdowns in the specific form required for publication are produced.

Adjustment

Ordinarily, adjustments are not necessary. However, in the event of a LHB being unable to provide data for a particular month, solutions such as entering an average value or rolling forward the previous month’s data might be considered in order to enable a full publication. Any action taken would be noted in the NHS activity and performance summary statistical release and StatsWales tables.

Quality assurance

At Welsh Government quality assurance is undertaken in line with the following quality strategy and the Quality pillar of the Code of Practice for statistics.

Quality assessment

The SCP cancer services statistics were labelled as experimental from June 2019 to November 2022. Prior to this, cancer data were collected on the basis of urgent and non-urgent cancer pathways, and were designated as National Statistics. The SCP-based cancer statistics were designated as experimental because the method was new and subject to testing and ability to meet user needs.

A review in 2022 concluded that the methods were appropriate and the statistics were robust and reliable. They meet a range of key user needs and users regard them as credible, trustworthy and critical to their work. They were judged to be compliant with the key principles of the Code of Practice for official statistics and the experimental label was removed in November 2022. In February 2023 the Office for Statistics Regulation (OSR) agreed that the statistics meet the standards required for designation as National Statistics.

Information on the development of the suspected cancer pathway collection is available on DHCWs data dictionary pages.

User needs

An understanding of current waiting times for cancer services, and historical trends, is crucial for those involved in planning and decision making at national and local levels. The key users of the statistics are:

• ministers and their advisors
• members of the Welsh Parliament (Senedd)
• members Research Service in the Welsh Parliament (Senedd)
• officials within the Department for Health and Social Services at Welsh Government
• NHS Wales
• Local Health Boards
• hospitals delivering cancer services
• students, academics and universities
• other areas of the Welsh Government
• other government departments
• media
• citizen users

The statistics are used in a variety of ways. Some examples of these include:

• advice to ministers
• to assess, manage and monitor NHS Wales performance against targets/standards
• to inform service improvement projects for areas of focus and opportunities for quailty improvement
• as a source of information to inform GP referral decisions and to advise patients when referring them for diagnostic or therapy services
• to contribute to news articles on waiting times in Wales
• by NHS Local Health Boards, to benchmark performance against national targets and other Local Health Boards
• to understand the service the public receives from NHS Wales

If you are a user and do not feel the above list adequately reflects your needs, please let us know via stats.healthinfo@gov.wales.

User satisfaction

Engagement with key users across the Welsh Government, DHCW and LHBs is frequent. A small user engagement exercise in 2022 found that the statistics are critical to understanding activity and performance in cancer services in Wales and that the data are regarded as credible and reliable. There are some further breakdowns that would be helpful but aren’t currently collected in the data, for example data below the main types of tumour (tumour sub-site). Some issues around accessibility, largely related to StatsWales functionality, were identified and it may be possible to address these as part of a review and re-build of StatsWales. The statistics are intended to provide a relatively high level picture of cancer services activity and performance in Wales. More detailed analyses based on the same underlying data are published in a DHCW dashboard, and statistics on cancer incidence, survival and mortality are published by Public Health Wales (PHW).

Feedback relating to the suspected cancer pathway statistics is welcome and can be submitted to stats.healthinfo@gov.wales.

Completeness

The statistics reflect data from all patients who are suspected of having cancer. Data are extracted from administrative systems and are a complete record of known and recorded cancer activity. The pathway measures the time from the point of suspicion to the start of treatment.

All patients who receive their first definitive treatment within Wales are included in these figures. Patients who receive their first definitive treatment outside of Wales are excluded from the official statistics, however, local health boards are still expected to manage and report informally on these patients to Welsh Government where possible.

LHBs where patients are initially referred are responsible for reporting the cancer waiting times for those patients, regardless of where consultations or treatment are provided. LHBs where patients are referred are always responsible for monitoring patient’s progress and reporting on it.

Overall accuracy

The statistics are based on administrative data. Consequently, data entry and processing errors (non-sampling error) are possible. These could occur when data are entered incorrectly into administrative systems or through measurement errors resulting from inconsistent interpretations of definitions.

To reduce non-sampling error, standards and guidance are provided about the data collections (DHCW). Standards relating to this data collection have been reviewed and passed by the Welsh Information Standards Board (WISB). Where non-sampling error affects the data, and is identified, we provide full information for users to allow them to make informed judgements on the quality of the statistics.

Coverage errors are also possible. Under-coverage could result from cases not being recorded, or errors in data entry. Over-coverage could result from duplicate records not being identified and removed. There is no known reason to suspect the extent of over and under coverage in the data is on a meaningful scale, and the quality assurance processes in place mitigates the risk.

The risk of any material impact on the aggregate statistics from coverage errors and non-sampling error is judged to be very low.

Data revision

Revisions to data can occur for a number of reasons including late returned data, incorrect estimates or revised back data. Monthly data are subject to revision up to five months after publication. This is to allow local health boards sufficient time to evaluate any revisions prior to submission. For example: in the statistical release published in August 2022, where the latest data for cancer services was June 2022, the data for the fourth financial quarter 2021-22 (January, February & March 2022) was revised on StatsWales.

Any further revisions to the data are noted in the statistical release and in the information accompanying the StatsWales datasets. Large revisions would be noted on the summary headline page.

Notes at the end of the release inform the users whether the outputs have been revised or not (denoted r). We will also give an indication of the size of the revision between the latest and previous release.

Users are informed about revisions in line with the Welsh Government’s revisions, errors and postponements policy.

Timeliness

Statistics are published on a monthly basis around 7 weeks after the end of the reference period. Outputs adhere to the Code of Practice for official statistics. Publications are pre-announced on the statistics and research pages of the Welsh Government website. Postponed outputs are handled in line with Welsh Government’s revisions, errors and postponements arrangements.

Punctuality

Postponed, cancelled or late publications are extremely rare. Since the SCP-based cancer services statistics were introduced in 2019 there has not been an instance where the statistics were not published on time at 9:30 on the pre-announced release day.

Geographical comparability

Agreed standards and definitions within Wales (DHCW) provide assurance that the data is consistent across Local Health Boards. These are published on the DHCW website.

The four nations of the UK publish statistics on cancer waiting times separately. The outputs differ, reflecting the different policy priorities of the different nations. A four nations group has been established to gather information about concepts and definitions in NHS statistics, including cancer statistics. Summary information can be requested from gss.health@ons.gov.uk and more detailed work is ongoing to fully understand and describe the extent to which the nations’ statistics are comparable.

In England, the statistics are published on a monthly basis by NHS England - Cancer Waiting Times. Guidance on operational standards for cancer waiting times commitments can be found in the National Cancer Waiting Times Monitoring Data Set (NHS Digital).

In Scotland, the statistics are collected and published on a quarterly basis by Information Services Division (ISD) Scotland - ISD Scotland - Cancer Waiting Times.

In Northern Ireland, the statistics are published on a quarterly basis by the Department of Health - Northern Ireland - Cancer Waiting Times.

Comparability over time

The SCP cancer services statistics are not directly comparable with the previous published statistics because previously data were reported separately on the basis of urgent and non-urgent pathways. The SCP statistics have been compiled on a consistent and comparable basis since December 2020.

While suspensions were applied under the previous method, they are no longer applied for performance monitoring with the SCP statistics. Therefore, current closed pathway data for patients starting their first definitive treatment are comparable to the pre February 2021 SCP data without suspensions (StatsWales), dating back to June 2019.

Publication

The statistics are pre-announced on the Welsh Government website and published at 9:30am on the penultimate Thursday of every month. In shorter months where publishing on the penultimate Thursday would not allow for sufficient processing and validation, the release is moved to the final Thursday of the month. Notice of this change will be given in the previous month’s publication. All publication dates for the next year are outlined in the upcoming publications calendar.

Headlines, narrative commentary and charts illustrating trends are provided in the monthly statistical release, published in HTML format on our website. The full back series of data can be viewed or downloaded as open data on the StatsWales website. There are some known issues affecting presentation and functionality of all data published on StatsWales. These may affect some users’ ability to access and use the statistics and these will be taken into account as part of a review of StatsWales which is intended to lead to an improved platform.

We also publicise the outputs on Twitter. All outputs and releases are available to download.

The more detailed data available on the StatsWales website can be manipulated online or downloaded into spreadsheets for use offline.

We aim to use plain English in our outputs and all outputs adhere to the Welsh Government accessibility policy. Our statistical releases are published in Welsh and English.

Online databases

The full historical series of statistics are published in the following StatsWales tables:

• Suspected cancer pathway – closed pathways (StatsWales)
• Suspected cancer pathway – open pathways (StatsWales)

Documentation on methodology

The following sources of information are relevant to the data collection and the published statistics on cancer services:

• Information about the Suspected Cancer Pathway – NHS Wales Health Collaborative
• Suspected Cancer Pathway section of the online NHS Wales dictionary

Quality documentation

The following sources of information are relevant to quality of the data and published statistics on NHS services in Wales, including cancer services:

• The Information Quality Improvement (IQI) (NHS Wales) initiative
• NHS Wales Information Design and Standards Development
• Information about data quality, standards and quality assurance and data quality tools in NHS Wales
• UK Statistics Authority: Quality Assurance of Administrative Data

The SCP data are derived from the Welsh Patient Administrative Data System (WPAS) via a ‘cancer tracker’ module. Two health boards do not use WPAS and therefore provide manual extracts to DHCW from their own systems. The data provided to DHCW are loaded into a national centralised database of cancer patient care records across Wales. Only data that are necessary for the statistics are included.

The administrative systems underpinning the data exist for operational purposes, and no dedicated data collection is necessary to produce these statistics. Local health boards conduct validation exercises before providing the data each month, which leads to some resource burden. However, it is not possible to measure the costs associated with the validation and aggregation of the data specifically for the purpose of producing the official statistics.

Confidentiality: policy

The Welsh Government statistics and research statement on confidentiality and data access describes our approach to data confidentiality and conforming with the data governance principle of the Trustworthiness pillar in the Code of Practice for Statistics.

Confidentiality: data treatment

Though patient level data is collected by DHCW, only aggregate data is contained in the extract provided to Welsh Government to ensure only the necessary data is held centrally by Welsh Government.

While the published statistics include some small numbers, the information is not considered to be sensitive in nature and no personal identifying information is presented.

Following an internal disclosure risk assessment we assess the disclosure risk based on the published statistics to be very low. Therefore, we do not apply suppression to small values. This is in-line with practice in the other nations of the UK with cancer waiting times statistics.

Statistician: Ryan Pike
Email: stats.healthinfo@gov.wales

Media: 0300 025 8099