Hospital Episode Statistics data changes in 2021

Extensive work has been undertaken to change the way in which we process the Hospital Episode Statistics (HES) data set and facilitate a wide range of benefits including but not limited to

• improve matching of patients across all national patient-level datasets through the Master Person Service
• exploit new analytical and data science capabilities
• increase efficiency through automation
• reduce cost by retiring legacy systems

a small number of issues have been identified within our new processes that we are actively looking to address as quickly as possible. These may impact upon the either the data quality or availability of the specific data fields concerned following our move to the new processing systems until these are resolved. Details of the data fields impacted and what this means in terms of the data are available in the consolidated technical output specification (ToS) document, under the ’Known Data Item Issues’ tab. 

Specifically, Critical Care has a number of DQ issues, particularly for neonatal and paediatric, and particularly around the how records are being grouped into a critical care period. The issues are understood and documented within the ToS. 

Consideration should be taken when using these data as it could distort any time series analysis significantly for fields listed with a known data item issue.

Data users will see changes to HES in the following ways:

• the introduction of the MPS Person ID will enable direct linkage of patient records across data sets, reducing the time and complexity in the analysis of HES data
• the HES_ID field will be retired along with a list of other fields that are no longer valid

We are aligning our handling and updating of reference data to other data sets. These changes will help to increase the amount of linkable data available to support research and planning.

These changes will help us

• improve matching of patients across all national patient-level datasets through the Master Person Service
• exploit new analytical and data science capabilities
• increase efficiency through automation
• reduce cost by retiring legacy systems

All users of HES data will be affected by these changes but please note that the data submitted to SUS+ is not affected.

For activity occurring from  April 2021, the patient identifier in HES will change from HESID to the Master Person Service (MPS) person Identifier (MPS Person ID). This change will take effect from the April to July 2021 year to date provisional HES data release in September 2021 and will be applied to all previous years of HES data.The MPS Person ID will be the biggest change to HES and is part of a wider strategy to move to a common patient identifier across all national patient level data sets.

MPS Person ID will support data users to

• measure counts of patients rather than attendances, appointments, episodes
• link together activity in the data relating to the same patient such as continuous inpatient spells
• Identify and track specific patient cohorts across time

MPS Person ID confirms the individuals identity using PDS data on Spine and assigns a common ID across all NHS Digital data sets.

A detailed analysis of the changes to HES and how this impacts data and statistics can be found in the methodological change paper and associated data tables accompanying this.  

We are aligning our handling and updating of reference data to other data sets such as  Mental Health Services Data set (MHSDS).

The areas of change are

• the criteria for accepting values in some fields as valid
• the frequency with which derivations will be rerun with latest reference data

All HES products, from the April to July 2021 year to date provisional HES data release in September 2021, will contain TOKEN_PERSON_ID only. 

Any HES data received from September 2021 will contain TOKEN_PERSON_ID instead of HES_ID.

The overall structure of HES data will not change. For a period of time, your HES data will still contain HES_ID field, but this will not be populated for activities recorded past March 2021 for the April to July 2021 year to date provisional HES data release in September. Instead, data referring to activities recorded from April 2021 onwards, will contain TOKEN_PERSON_ID. 

DARS customers will not need to make any changes outside their usual renewal schedule. However, if you are currently working on your affected Data Sharing Agreement (DSA) you may be able to update it. Please speak to your case officer for further information.

TOKEN_PERSON_ID  is available to select on DARS Online from November 2020.

Should you need to, HES_ID will be available to select on DARS Online until February 2021.

To receive your first drop off date with TOKEN_PERSON_ID your DSA will need to be signed off and active by August 2021.  

If you would like to make changes to your fields when you are renewing your DSA please speak to your case officer so they can arrange these changes for you, as you will not be able to make changes to your fields online. 

There will be no additional charge for adding TOKEN_PERSON_ID to your application.  

We have reviewed the list of variables within HES data and identified fields that are no longer populated. Whilst those retired fields are still available to select on DARS Online as they are valid for historical data they are not populated. Therefore, we strongly advise you to review the list of retired fields before selecting your HES variables on DARS online.         

If you have received HES data (as extracts) before June 2021 and are requesting further HES data, you will receive a HES Person ID bridge file to enable you to attribute TOKEN_PERSON_ID to any historic data you hold. There is no intention to re-supply any HES data provided prior to June 2021.

If you received provisional HES data prior to the April to July 2021 year to date provisional HES data release in September 2021 then you will not receive a bridge file for this activity as this activity will be superseded with a year to date extract containing the new person identifier covering the the April to July 2021 year to date provisional HES data release in September 2021 onwards.

If you did not previously receive HES_ID the assumption will be that you also do not require TOKEN_PERSON_ID, unless you tell us otherwise during your application process in which case you will need to make an amendment to your application. You will continue to receive your local identifier or Study_ID to track patients as before.

HES admitted patient care bridge files are linkable to historical HES admitted patient care data using the EPIKEY field.

HES outpatient bridge files are linkable to historical HES outpatient data using the ATTENDKEY field.   HES accident and emergency bridge files are linkable to historical HES accident and emergency data using the AEKEY field.

This will be supplied on a case by case basis, allowing you to assign TOKEN_PERSON_ID to your historical data, in place of HES_ID. From 2021/22 onwards data will be supplied with TOKEN_PERSON_ID so no further bridge files will be supplied.

If you receive new copies of historical HES data frequently, you will not need the bridge file as these will be supplied with TOKEN_PERSON_ID in future. You will therefore have a consistent person identifier (TOKEN_PERSON_ID) across all data years.