Impact of changes to recording of Same Day Emergency Care Activity on Hospital Episode Statistics (HES) data

We are implementing a consistent standardised way in which we record activity of patients accessing Same Day Emergency Care (SDEC) activity within our national data sets. This will see changes in certain types of activity that may have previously been recorded within HES data set for a number of care organisations. Doing this will help with identifying opportunities to promote and enhance this key care activity within the NHS.

Understanding how this change may impact the HES data product going forward will help users interpret what this may mean for their current uses of the HES data product and inform what, if any, approaches they may need to take to factor this into their own analyses in the future.

Same Day Emergency Care (SDEC) is the provision of same day care for emergency patients who would otherwise be admitted to hospital. It aims to benefit both patients and the healthcare system by reducing waiting times and hospital admissions, where appropriate.

Under this care model, patients presenting at hospital with relevant conditions can be rapidly assessed, diagnosed, and treated without being admitted to a ward, and if clinically safe to do so, will go home on the same day that their care is provided.

We need to be able to analyse SDEC related activity to make effective and informed data driven decisions within the NHS. This requires that we capture all SDEC activity in a consistent format via submission through one single data source to create the data assets and products that are needed. At present SDEC activity is being recorded either in the Commissioning Data Sets which are used to create HES in the Admitted Patient Care (APC) or Outpatient (OP) data flows by organisations, or not at all.

To address this, the following information standards notice has been published that will see the recording and submission of SDEC activity occur within the Emergency Care Data Set (ECDS) data submission. This requires all organisations to record SDEC activity as part of their submission by 1 July 2024 at the latest.

However, a number of organisations are already migrating to the new data source. Depending on their previous SDEC reporting approaches, such changes may have implications for how certain users of HES data use and interpret the existing data as well as understanding the impact on future analytical products and services.

The purpose of this methodological change is to support users of HES data products by making them aware of the anticipated changes within the data submitted through this transition period. This is supported with accompanying analyses seeking to assess the potential impact of the changes to the data products along with any documented limitations associated in the methods adopted to do so.

Further details about NHS England's Same Day Emergency Care (SDEC) are available. 

Whilst there is a requirement for all SDEC to be recorded within ECDS by 1 July 2024, a number of providers have already moved to ECDS based recording of SDEC. More are due to follow, from the implementation date of ECDS version 4.0 from 1 July 2023. 

An extended transition window means it is not possible to isolate the impacts caused by this shift in recording of SDEC to a specific date in the data, either nationally or regionally. Users of HES data should consider this in their own analyses and interpretations of data particularly where this spans the transition period.

Currently we do not have a universal picture of how current providers report their SDEC activity. However, in 2021, NHS England conducted the NHS Benchmarking Audit with responses from 128 providers. Of these responses, 119 providers reported that they recorded SDEC (either partially or fully) within the APC dataset. 

A user survey is planned for Summer of 2023 to understand the current reporting behaviour and to establish a clear overall picture to support monitoring of the implementation of ECDS version 4.0 reporting. It is planned that the results of this will be published and updated over the transition period to help analysts to understand changes they may be observing in reporting from HES data particularly at organisational and sub national geographies during this period.

Some organisations who previously reported SDEC activity have already made this transition.

A recent example is Manchester University NHS Foundation Trust who moved from reporting SDEC activity in APC to ECDS in September 2022. This notice includes analyses specifically relating to this example to help HES users anticipate the types of changes that may be experienced in their analyses before and after the transition of SDEC reporting occurs.

Methodology in projecting impacts in HES data 

There is no clear way to accurately identify SDEC activity within existing HES data products such as APC or OP data. This is part of the reason why the reporting of SDEC activity is being consolidated solely into the existing submission of ECDS data. Due to this it is not possible to say with certainty what the impact of these reporting changes of SDEC activity will have on existing HES data products.

Analysts within NHS England have developed a proxy mechanism which has been developed to try to identify SDEC activity, which is used in the associated analyses articulated later in this notice.   

This methodology considers activity that may be amenable to SDEC where:

• Finished Admission Episode (FAE) has one of six Treatment Function Codes and
• the patient is discharged within one day, and
• the patient was admitted from an A&E attendance

This has been used to inform the basis of the following findings within this notice and in the associated excel data tables published with this report.

From analyses on activity relating to Finished Admission Episodes (this being the first recorded consultant episode in a hospital provider spell) in APC, we have projected a possible 7% reduction in the national reported activity post full implementation of reporting SDEC activity within the ECDS from existing levels observed in 2021-2022, as shown in Table 3a in the accompanying data file.

From reviewing local experiences, anecdotal evidence suggests that the actual reduction may not be this high.  

Analysis of possible SDEC outpatient activity was negligible, so no significant impact is expected to the outpatient data overall. However, we need to recognise there are individual providers that record SDEC within their outpatient data and this will reduce over the transition period. 

From projected analyses we are expecting to see larger decreases for activity relating to patients under the age of 20. This illustrates an 11% reduction in nationally reported FAE activity, as shown in Table 3a in the accompanying data file.

Out of a total of 500 providers, 29 have a rate of more than 10% but less than 20% of activity that would be considered SDEC, with no providers over 20%. There are 62 providers showing between 5% and 10%, as shown in Chart 1 below and Table 3c in the accompanying data file.

Using a projection with the proxy SDEC mechanism described above for attendance activity moving out of APC to ECDS SDEC, looking at 2021-2022, emergency admissions could reduce by as much as 18% and emergency admissions with zero bed days could decrease by 50%, as shown in Table 1c in the accompanying data file.

The ten most affected diagnoses, taking the highest percentage decrease into account, can be seen in Chart 2 below. The underlying data is also available in Table 1c.

The top ten most affected diagnosis groups, taking the highest percentage decrease into account, and can be seen in Chart 3 below. [ann-ann] donates the ICD-10 diagnosis chapter.

We have not sought to map the change on all the classifications and breakdowns possible in HES data reporting here and would advise that analysts consider and review local impacts on specific breakdown compositions that are relevant to their own analytical outputs and data products.

As described earlier there is no clear and accurate way to identify SDEC activity within our current data flows and these counts and percentages should be considered as estimates only.

It is important to recognise that activity levels can be influenced by many factors in HES. These include seasonality, increases in hospital provision following the backlog created from necessary service changes for the COVID-19 pandemic, flu, winter pressures and ongoing industrial action. As such, accurately isolating and identifying reductions purely due to SDEC implementation is not possible and indeed may not be visible if offset by a surge in demand responding to one or more of the other factors listed.

In this section we seek to explore what the impact of transition to ECDS based SDEC reporting may mean for specific organisational level analyses using HES data. We do this by analysing existing ECDS SDEC activity compared with the corresponding APC data of organisations who have been early adopters of SDEC reporting in ECDS moving from APC

Possible impact on the HES APC data from looking at 'Early adopters'

We have been instructed to stop their existing method of recording SDEC once they have moved SDEC activity to ECDS, so as not to duplicate their submissions, including during the pilot period.

As such, we have taken the early adopters' monthly activity in ECDS SDEC and run a comparison with their APC data to help understand how their APC activity has been impacted for these provider sites.

We have also run detailed analysis of key fields for Manchester University NHS Foundation Trust, as this provider has a clear cut off point at which it moved the reporting of SDEC activity from APC to ECDS SDEC, to gain an understanding of possible changes to APC record demographics.

17 early adopter provider sites have already moved to this model, by submitting SDEC activity to Emergency Care Data Set (ECDS) department type 5. This accounts for 165,263 records since April 2021, of which 88,665 are since April 2022. Providers have until 1 July 2024 to move to reporting SDEC activity in this way, as per the published Information Standard for ECDS version 4.0. 

Table 1: Provider Sites submitting SDEC to Emergency Care, April 2021 – Oct 2022

From internal pilot data, we compared the monthly counts of SDEC ECDS provider site attendances with their corresponding APC activity. From this analysis, we observed a clean switch over for the top three sites in Manchester. For other sites, although APC activity decreased in general, there was no clear switchover. Also, a number of providers will not have been submitting SDEC activity in APC in the first instance. This being the case, a reduction in APC recorded activity is not necessarily expected or indicative of our projected SDEC.

For this reason we have chosen Manchester University NHS Foundation Trust to observe changes in HES activity pre and post their move to submitting SDEC to ECDS.

Using Manchester University NHS Foundation Trust as our example, we have compared admission activity prior and post moving SDEC activity out of APC to ECDS (for September to December 2021 and September to December 2022).

Emergency admissions have dropped from 41,120 to 28,250 and emergency admissions with zero bed days have dropped from 15,585 to 5,575, as shown in Tables 2a and 2b in the accompanying data file.

As shown in Table 2c, diagnoses that observe a decrease in finished consultant episodes (FCEs) aligning with the proxy national SDEC projections are: Viral infection of unspecified site (B34) of 665; Pain in throat and chest (R07) of 2,030 and Other soft tissue disorders, not elsewhere classified (M79) of 830. However, Manchester also observes the second largest decrease for emergency use of U07 which is COVID-19 related of 1,595, which is not represented in the top ten in the proxy methodology.

A prediction that does not align with the proxy national SDEC projection is that Manchester University NHS Foundation Trust does not observe the predicted larger percentage decrease for patients under age 20. Instead, the percentage impact is consistent across all age groups.

Caveats and uncertainties 

As described earlier there is no clear, consistent way that provider sites switch over their SDEC activity from APC to ECDS. Analysis of individual sites cannot therefore be assumed for other sites and is intended as a guide only.

We welcome any feedback on the methodological change paper, particularly with respect to how these changes will affect users in any ways that have not been communicated as part of this document.

Feedback can be submitted via [email protected] with the subject HES SDEC Methodological Changes 2023.