NHS Digital Data Release Register - reformatted

The Clatterbridge Cancer Centre NHS Foundation Trust

Project 1 — DARS-NIC-14170-X2G3L

Opt outs honoured: N, Y

Sensitive: Sensitive, and Non Sensitive

When: 2017/03 — 2017/05.

Repeats: One-Off

Legal basis: Section 251 approval is in place for the flow of identifiable data, Health and Social Care Act 2012, Section 42(4) of the Statistics and Registration Service Act (2007) as amended by section 287 of the Health and Social Care Act (2012)

Categories: Anonymised - ICO code compliant, Identifiable

Datasets:

  • Hospital Episode Statistics Admitted Patient Care
  • Hospital Episode Statistics Accident and Emergency
  • Hospital Episode Statistics Outpatients
  • Office for National Statistics Mortality Data

Benefits:

Objective One • Policy Development NATCANSAT supports DH expert working groups and committees in their development of policy. Analysis of the HES (and linked) data benefits patients by facilitating evidence based decision making by these groups, where the policy being developed needs to include: o Numbers of patients and outcomes for specific patient cohorts o Volume of cases affected by a proposed change o Impact of the change in the service (cost/resources/travel times/critical mass etc) Example: DH Cardiovascular Strategy Supported by the analysis carried out by NATCANSAT, the committee developing the strategy made a number of recommendations regarding the improvement of diagnosis and treatment of Cardiovascular Disease, which will save lives, and improve the quality of life of survivors. Objective Two • Performance Management NATCANSAT provides support to NHS and DH organisations in identifying progress in the implementation of new policy, or identifying trends in specific measurables which may indicate progress towards a particular target or goal. Performance management benefit patients by facilitating service improvement in the following ways: o Development of a measure or measures which can be used to report on progress (including a clear definition of patient cohort and outcome) o Baseline reporting of performance against the measure(s) o Regular reporting of change in the measure. • As a result cost reductions and improved care can be provided. Example: Performance Management – 23 hour breast model The monthly performance management data provided to NHS Improvement and participating providers on their progress in implementing shortened lengths of stay for patients undergoing mastectomies supported NHS Improvement to work with Providers to implement the new policy resulting in reduced costs for the service and improved outcomes for patients. Objective Three • Benchmarking Benchmarking analysis work carried out by NATCANSAT benefits patients by informing the service of variations in practice and outcome, identifying areas of strength and weakness in the service which may form examples of good practice, or require additional support. The variation identified may lead to changes in policy. Example: Benchmarking – Transforming In-patient Care The National Cancer Director used a breakdown of trends in bed days for cancer patients by year/disease type/patients classification/speciality/admission method in order to identify the reasons for variation in length of stay for cancer patients and to highlight areas where reductions might be possible. This facilitated a reduction in the cancer bed days, resulting in cost savings for the service and improvement in quality of life for cancer patients. Objective Four • Service Planning/Reconfiguration NATCANSAT provides to NHS and DH organisations a range of analyses associated with service reconfiguration on a local or national basis. Analyses range from calculation of the resources required at a new facility and the potential reduction in activity at those facilities currently providing a service, to assessment of the most appropriate locations for a new national service, including the projected activity at each location. Example: Service Reconfiguration – Cardiac Revascularisation Service NATCANSATs analysis facilitated the provision of a cardiac revascularisation service at optimal locations, and an accurate estimate of the expected demand in each location. As a result the service was able to locate these services optimally and provide the appropriate resources at each site. The outcome of this change will be to save lives by ensuring that patients have access to this life-saving treatment in a timely fashion, and to save money by ensuring that the correct resources are installed in each facility.

Outputs:

Outputs are in the form of anonymised, aggregate tables and interactive tools, and also maps and narrative. Interactive tools normally consist of a ‘microsite’ using SQL server reporting services and SQL server cubes placed on a website and accessed securely. Identifiable or sensitive items are NEVER included in the source data used to populate interactive tools. Only the minimal dataset needed is included. Record level extracts are NEVER shared with third parties. Output which is shared with third parties (beyond sub-licenses) does not include aggregations which are able to produce ‘small numbers’ as defined in the HES Analysis Guide (eg: data is not aggregated below Strategic Health Authority level), or small numbers are suppressed, including totals/sub-totals, in line with the HES Analysis Guide. Analysis has been used: • Analysis of resection rates for prostate cancer using linked cancer registration and HES data for 2009 and 2010 for a cancer network This work used identifiable data for linkage purposes. The complete Cancer Registration database was needed in order to include those patients with prostate cancer who had not been admitted to hospital (and had no HES record) in the denominator. • Analysis of the NHS footprint for long term survivors of cancer using HES data with updated death data for the National Survivorship Initiative. Identifiable data was needed for linkage to the demographics batch tracing service, to carry out up-to-date survival analysis for these patients. • DH Cardiovascular Strategy where it supported important decision making around the placement of the service for acute cardiac revascularisation services, thus ensuring access to this life saving treatment for patients in all geographical areas. Identifiable data was needed for linkage to the demographics batch tracing service, to carry out up-to-date survival analysis for these patients. • DH Cardiac Rehabilitation Document which outlines analysis demonstrating the positive impact of rehabilitation on re-admission rates after cardiac events, and supports the development of business cases to provide this important service to improve patient care. No identifiable data used for this analysis. • DH Enhanced Recovery Program where it was used to support the implementation of a program to improve patient safety and reduce costs by shortening length of stay and minimizing side effects before and after surgery. No identifiable data used for this analysis. • In breast cancer surgery work, there was a significant reduction in length of stay for breast cancer surgery. This was not caused by the production of data, but the progress reports produced facilitated the intervention of the NHS Improvement team to ensure that the targets were met. No identifiable data used for this analysis. • Liverpool CCG microsite- monthly monitoring of a variety of reports on a monthly schedule completing each month within 6 weeks of receipt of the data from MMES. No identifiable data used for this analysis. • Detailed analysis of Adrenalectomies for the DH Expert working group of the British Association of Endocrine Surgeons to inform the development of national guidance on the organisation of adrenal surgery. Produced in September 2014, for publication in a peer-reviewed journal in 2015. No identifiable data used for this analysis • Emergency Admissions Catchment Map (based on all Emergency Episodes) for CQC produced in September 2014. No identifiable data used for this analysis • In-patient Mental Health catchment area/population/proportional activity analysis was due February 2015, but delayed as a result of our extended application process for data. No identifiable data used for this analysis • Analysis of hospital care received by people dying as a result of suicide during the last months of life had been requested in late 2014, NATCANSAT has been unable to deliver this work due to the extended application process for data. It is expected that this work will still be of value once the approval process is complete. No identifiable data used for this analysis. • Cancer Commissioning Toolkit which is used by commissioners to identify and address variation in cancer service provision No identifiable data used for this analysis • These data have been used previously to respond rapidly to the needs of NHS Improvement (now NHSIQ) to measure progress of its activities in shortening length of stay for breast cancer surgery, and implementation of Enhanced Recovery after orthopaedic and colorectal procedures. The breast cancer surgery work was the subject of an HSCIC case study on the use of HES data http://www.hscic.gov.uk/casestudy/breastcancersurgery No identifiable data used for this analysis Some examples of specific outputs relating to the Cancer and Cardiac work are available on the following links - CANCER: http://www.natcansat.nhs.uk/data/hescancer.aspx CARDIAC: http://www.natcansat.nhs.uk/data/hescardiac.aspx

Processing:

Extracts are processed to mark relevant records using a range of groupings which allow a rapid response to specific queries (eg: identifying episodes which involve particular diagnoses and procedure codes in isolation, or as a group, identifying episodes for patients who have relevant features in earlier or later episodes, linking episodes together to identify groups of procedures carried out over a period of time). The processed tables are then readily available to respond to queries as they arrive from government and the service. The majority of the analysis completed is done using non-identifiable/non-sensitive data alone. Only the more complex and sophisticated analysis including data linkage to sources without HES ID require identifiable data, and only analysis specifically around the patient’s GP and Consultant require sensitive data. However, for completeness, the whole processing activity undertaken using the data is described below. Activities including the sensitive or identifiable data are clearly identified. 1. Non-identifiable data is received from the HSCIC and imported in to a SQL database. 2. A number of fields are added to the data, which are populated with data derived from other items in the record, or other records in the data for the same patient (all from the non-identifiable extracts). Eg: Activity records are flagged if they contain a cancer diagnosis Activity records are flagged if there is another record in the same extract which contains a cancer diagnosis ‘current provider’ are populated with the current provider code for the record – this changes if the provider has merged or changed identity since the activity took place. 3. Identifiable/Sensitive data is received from the HSCIC and imported in to a SQL database. This table is only accessed by one analyst responsible for creating extracts (extract analyst) 4. The identifiers are used by the extract analyst to link to other data sources (cancer registrations, radiotherapy) and are used to augment the annual extract data tables generated in steps 1&2. Eg: Activity records are flagged if the patient has received radiotherapy Activity records are populated with the tumour site and histology, and date of diagnosis from the cancer registrations database 5. The SQL tables are processed to generate OLAP cubes which facilitate rapid analysis. An OLAP cube is a multidimensional data array, which ‘slices and dices’ a large dataset, allowing subsections to be rapidly summated. 6. An ‘output analyst’ (who does not have access to the identifiable data), will then use the table or cubes to carry out analysis and generate outputs. 7. When analysis is requested which is not supported by the existing tables and cubes, it may be possible to add new fields to the table/cubes which will support the analysis, or it may be necessary for the ‘extract analyst’ to return to the raw non-identifiable or identifiable/sensitive data in order to create a new data source for the analysis. An ‘output analyst’ will then use the new data source to generate the outputs needed. Cubes/Tables developed for a specific task will contain the minimum possible information needed. This may include sensitive data, and/or record level data. 8. HES data is also used to augment other datasets (eg: to add detailed surgical data to the cancer registration database). In these cases the ‘extract’ analyst uses the identifiable data to carry out linkage, and then generates a pseudonymised version of the extract table in line with the HES analysis guide, to be used by the ‘output analyst’. 9. NHS Number is used for linkage to other data sources (Cancer Registrations taken from PHE and/or ONS, Radiotherapy Dataset, Cancer Audit Databases, in order to augment the HES data using data items from these sources, or to augment the other data sources with data items from HES.(carried out in step 4,7 or 8) 10. Postcode is used for geographical analysis, where it is used to: a. correctly ascribe the patient’s residence to a range of geographical areas (eg: CCG/LAT, or to a catchment area calculated by NATCANSAT using travel times, patients flows or projections), b. plot the patient on a map using a grid reference c. identify socioeconomic measures for the patient’s residence d. carry out travel time analysis from the patients residence to service locations or proposed service locations Example: Objective 1 - Policy Development – DH Cardiovascular Strategy DH in development of a cardiovascular strategy needed information on the patient pathway for patients suffering cardiac events, undergoing cardiac revascularisation procedures, with or without subsequent cardiac surgery. This analysis required the extract analyst to build a specific data table made up of patient pathway data extracted from a series of hospital admissions and outpatient attendances. The ‘extract analyst’ requested up-to-date death data using the NHS number, Date of birth and Postcode for relevant patients from the Demographics Batch Tracing service, and linked this (without the identifiers) to the data table using HESID. This data table was then passed to an ‘output analyst’ who was able to generate the outputs needed. Example: Objective 2 - Performance Management – 23-hour breast model As part of the ‘Transforming Inpatient Care’ initiative aimed at reducing patient bed days in hospital, a new model of care was developed for patients with breast cancer undergoing mastectomy. Outputs were needed in the form of tables which indicated the length of stay, and the readmission rate for patients admitted to hospital for a mastectomy. The ‘extract analyst’ was able to add new fields to the data table created in step 2 above, which flagged the relevant records, and identified those with readmissions to hospital. The output analyst was then able to use the table to generate the outputs needed. The process was repeated each month upon receipt of the MMES data in order to provide timely progress reports. Example: Objective 3 - Benchmarking – Transforming In-patient Care The National Cancer Director wanted to monitor the number of hospital bed days being used for patients with cancer. Using the tables generated in step 2 above, an extract analyst was able to generate monthly updates on the number of bed days used broken down by a number of key parameters.(eg: provider unit, cancer diagnosis, speciality) Example: Objective 4 - Service Reconfiguration – Cardiac Revascularisation Service New policy had indicated that people suffering a myocardial infarction should be able to receive revascularisation treatment promptly in order to improve outcomes. In order to implement the policy, it was needed to identify locations where the service should be offered so that all members of the population would be within easy reach of a facility. NATCANSAT undertook a population based geographical analysis to identify the key locations for the service to ensure that this was possible. New fields were added to the table by the extract analyst to identify patients who had received this treatment, and also those who might have benefitted from it had it been available to them. The output analyst used these data to estimate the demand at each of the locations, which could be used to calculate the resources needed. More complex processing activities are undertaken on a one-off basis, in order to support the development of policy or measure progress in specific areas. Generally, a specific extract is generated to be used by the NATCANSAT analyst for all but the most straightforward analysis, in order to minimize sharing of irrelevant data within the team, and to facilitate efficient working. Data Minimisation As part of this application, the data required has been rationalised and HES data currently held by NATCANSAT no longer covered by this agreement will be securely destroyed. Data supplied in 2012 from the PROMS database has been destroyed. Data supplied in 2012 from the SUS/PbR database has been destroyed. National Data is requested as projects include all of England. Years from 1997 onwards are requested to facilitate benchmarking over the maximum period possible. (earlier years do not include HESID, and use ICD9 diagnosis codes so analysis is complex) Long period trend analysis is important when identifying variations in rare conditions or procedures. Minimal identifiers, sensitive fields and non-identifiable fields are requested to carry out the analysis set out in this application.

Objectives:

National Clinical Analysis and Specialised Applications Team (NATCANSAT) is hosted by The Clatterbridge Cancer Centre NHSFT. NATCANSAT supports a range of DH and NHS organisations involved with National Cancer and Cardiac Programs by providing detailed analysis as required with reference to specific diagnoses and/or procedures. Only individuals substantively employed by the Trust, working for NATCANSAT, have access to the data. Pseudonymised Data Wherever possible, analyses are undertaken using pseudonymised data but identifiers are necessary for different reasons. Unique Identifiers are required only for the linkage process, in order to generate pseudonymised datasets for further analysis. Careful consideration is always given to the use of identifiable data, and whether it is appropriate or necessary for the purpose. This approach alongside the wider availability of linked data sources has allowed NATCANSAT to limit its use of identifiable data very considerably in recent years. Pseudonymised and sensitive extracts are used to carry out analysis of hospital activity for a range of diagnosis and procedure codes, in order to support the objectives identified below. The extract contains all records to support work which compares affected patients with more general populations, or to provide denominators for complex analysis. Identifiable Data for Cancer/Cardiac Identifiable extracts are for patients identifiers for patients identified as having diagnoses or procedures which are relevant to cancer or cardiovascular disease. These will be linked to the full HES dataset, to the radiotherapy dataset (RTDS), and other data sources as set out in the CAG approvals granted. They will also be used with geographical information systems (GIS) to carry out spatial analysis. Date of Birth is preferred by cancer registry linkage algorithms where it is used to validate the NHS number in the absence of a valid NHS Number status indicator in the HES data. NHS number is used as the key field for linkages carried out by NATCANSAT, it is requested from all of the data sources NATCANSAT link to. Postcode of home address from the HES database is requested for geographical location purposes. Full Postcode is used in conjunction with geographical information systems to establish a grid reference for the patient’s home. This allows work to establish the geographical spread of patients, and to identify catchment areas and population and measure variations in provision of service in relation to other factors. NATCANSAT have carried out large scale analyses of cancer service provision in the UK since 1998, these analyses have been used to allocate resources from central capital resource programs and to inform the processes for policy making and business case development across the DH and NHS. PIAG originally granted section 60 support for this use of data in 2003, and this was expanded to also facilitate linkage to cancer registration data and sharing of data between NATCANSAT and the English cancer registries in 2003 and 2004. In 2013 NATCANSAT reviewed it’s use of Cancer Data and made a new application to CAG replacing a number of previous approvals, this approval (CAG 1-06(FT2)/2013) provides the legal basis to hold identifiable data for patients with a diagnosis of cancer, or undergoing procedures relevant to cancer, and to link these data to a range of related datasets and has been reviewed and upheld by CAG each year since. NATCANSAT analyse data to demonstrate variation in coronary heart disease (CHD) service provision in England for preparation of guidance, planning service configuration and identifying areas for action. PIAG granted section 60 support in 2003 for NATCANSAT to hold identifiable data for patients with a diagnosis of cardiovascular disease, or undergoing procedures related to cardiovascular disease, and to link these data to other specific datasets. This approval (PIAG 4-09(g)/2003) has been reviewed and upheld each year. The objectives are described in detail below, when NATCANSAT is tasked to carry out a specific piece of analysis work to meet these objectives, a governance review is undertaken by the Senior Analyst specifying the work alongside the Head of NATCANSAT, or the Senior Analyst responsible for Information Governance. The purpose of the review is to establish whether it is appropriate to carry out the work in terms of • Will the output from the work meet the needs of the request? • Is the output already available elsewhere? • Does NATCANSAT have permission to use the required data for the purpose specified? • Does NATCANSAT have a legal basis to process the data needed for the purpose specified? The purpose of analysis is to fulfil the objectives below; Objective One • Policy Development (eg: provision of specific tabulations for expert clinical groups, or for inclusion in DH policy publications) NATCANSAT supports DH expert working groups and committees in their development of policy. In order to provide analysis which: o supports clinical trial or other evidence regarding the new policy being proposed (eg: variation in outcome for patients undergoing a particular procedure who have a particular diagnosis) o helps to identify the volume of cases which might be impacted by a change (eg: the number of patients requiring readmission to hospital after undergoing a particular procedure) o assesses the effect that a new policy might have on the service (eg: identify the additional resources which might be required if the criteria for undertaking a particular procedure are changed) Objective Two • Performance Management (eg: provision of tabulations or interactive tools which identify variations in service between providers and/or identify trends in response to change) NATCANSAT provides support to DH and the NHS as identified in section 4 above in identifying progress in the implementation of new policy, or identifying trends in specific measurables which may indicate progress towards a particular target or goal. Performance management objectives fall into the following categories: o Development of a measure or measures which can be used to report on progress (including a clear definition of patient cohort and outcome) o Baseline reporting of performance against the measure(s) o Regular reporting of change in the measure. (eg: NATCANSATs support of the implementation of the 23-hour model for breast cancer surgery) Objective Three • Benchmarking (eg: provision of tabulations or interactive tools which identify variations in activity or outcome between providers) Benchmarking analysis forms an important part of NATCANSAT’s output in order to inform the service of variations in practice and outcome, to identify areas of strength and weakness in the service which may form examples of good practice, or require additional support. The variation identified may lead to changes in policy. Objective Four • Service Planning/Reconfiguration (eg: geographical analysis to identify likely patient flows associated with a change in service, or to predict the impact on travel times) NATCANSAT provides to NHS and DH organisations (Restricted to the organisations included in the SRSA 2007, s42(4) ) a range of analyses associated with service reconfiguration on a local or national basis. Analyses range from calculation of the resources required at a new facility and the potential reduction in activity at those facilities currently providing a service, to assessment of the most appropriate locations for a new national service, including the projected activity at each location.