Delayed Discharges in NHSScotland
Summary of definitional changes and their impact on figures reported by Information Services Division
July 2016 data onwards
Introduction
The Information Services Division (ISD) has published information on NHSScotland delayed dischargesmonthly since June 2015. Prior to then, information was published quarterly from 2001.
Since the move to a monthly publishing, a number of data issues relating to the consistency of recording and interpretation of the recording guidelines have been identified.ISD has taken steps to address these issuesby conducting a review and consultation of the existing data definitions manual. This paper describes the approach taken, the outcome of that work and the impact of these changes.
Delayed discharges data definitions manual
Adata definitions manual (effective May 2012) wasavailable to all NHS and local authority partnerships who supplied delayed discharge information. This manual supported staff in these organisations to record and capture of delayed discharge data in a consistent way.
Quality assurance exercise
The Cabinet Secretary for Health and Wellbeing asked ISD and the Scottish Government to conduct a quality assurance exercise in late 2015 to ensure national consistency of data recording. This was in response to ongoing discrepancies between the monthly census figures and the average number of delays per day derived from the bed days occupied submission.
Public Bodies (Joint Working) Regulations
The Scottish Government developed acore suite of indicators in support of the Public Bodies (Joint Working) (National Health and Wellbeing Outcomes)(Scotland) Regulations 2014. One of these indicatorswill measure the percentage of people who are discharged from hospital within72 hours of being ready for discharge.This is a new indicator not previously reported by ISD.
ISD consultation
ISDrecognised that the national publication of delayed discharge information and the associated data definitions manual needed to bere-assessed and modified to reduce any discrepancies in the data and ensure that the new indicator could be measured.
ISD undertooka consultation exerciseon reporting and delayed discharge information. The responsesto this consultationare published on the ISD website.
As part of the consultation, the following factors were considered to ensure that data are robust:
- known data quality issues
- data extraction
- core minimum data set
- local recording practice
- local training and advice.
This work has resulted in a revised data definitions manual and national data requirements, both effective from 1July 2016.
Summary of changes
The changes detailed below indicate the main changes in the collection and reporting of delayed discharge information following the implementation of the revised data definitions manual and national data requirements effective from 1 July 2016 which are published for the first time on 13 September 2016.
Change / Up to 30 June 2016 / From 1July 2016Data source / Two files submitted to ISD:
- Census
- Bed days occupied
Codes 42 / 42X (Healthcare delays) / Included / Not included
Non hospital locations – patients in care home locations (including those awaiting choice of care home or completion of guardianship) / Included / Not included
Patients receiving Intermediate Care in a dedicated Intermediate Care facility / Not included / Not included
3 day discharge rule (Delayed patients discharged within 3 working days of census date) / Not includedin census snapshot
Includedin monthly bed day figures / Included
Census date / Day 15 of each month / Last Thursday of the calendar month
Reasons for delay
(see Appendix D in the national data requirements) / Separate codes and categories / Amended groupings to reflect health and social care integration
Impact of changes
The single data download for each calendar month supports reporting of delayed discharge information by ISD. It also offers flexibility to develop new outputs in the future to support Integrated Joint Boards such as the 72 hour indicator.
These changes will ensure the outputs (census and bed days) are more aligned as all the delays and associated bed days are provided for the same time period, and the census figure more in line with the average daily number of bed days occupied.
Impact on delayed discharge figures
- The removal of healthcare delays will result in a reduction (<1%) in both census and bed day figures from July 2016 onwards.
- The removal of patients in non hospital locations will result in a reduction (4-5%) in both census and bed day figures from July 2016 onwards.
- The inclusion of patients delayed at the census but discharged within three days of the census date will result in an increase (15-20%) in the number of delays reported at the census from July 2016 onwards. It will not have an impact on the bed days occupied as bed days occupied by these patients are already included.
- Changing the census date to the last Thursday of the month will ensure consistency month on month when analysing the figures by length of delay.
- The revision of the reason for delay groupings will have no impact on the total number of delayed discharges.
- There may be a further impact on the figures due to the change in recording and definitions. It should be noted that there may also be further/additional data validation being undertaken locally in the recording of delayed discharge patients. It is not possible to estimate what the impact of this would be on the reported figures.
Impact on trend figures
- A trend of census information is possible in two ways which reflect numbers using both the new and previous data definitions and guidance.
- It is not possible to provide comparable Scotland level bed days occupied trend information due to the fact that the actual number of bed days occupied by patients delayed due to healthcare reasons and those in non hospital locations cannot be explicitly identified and excluded from data up to June 2016. However it is possible to provide an estimate of the impact (approx 4% of bed days should be removed from historic data in order to compare to figures post July 2016) and this note will be annotated on the published tables.
Change / Summary of impact
Data source / The requirement to provide a single, validated data download may result in a change to figures reported by partnerships
Healthcare delays (reason codes 42 / 42X) / Reduction in census (<1%) and bed day figures (estimate <1%) from July 2016 onwards
Non hospital locations – including patients in interim beds in care homes (awaiting choice of care home or completion of guardianship) / Reduction in census (5%) and bed day figures (estimate ~4%) from July 2016 onwards
Patients receiving Intermediate Care in a dedicated Intermediate Care facility / No impact as not currently included
3 day discharge rule (census patients discharged within 3 working days of census date) / Increase in reported census figure (15-20%) from July 2016 onwards
No change to bed day figures
Census date / Minimal impact on high level figures but will allow consistency month on month when looking at length of delay
Reasons for delay / No impact on high level figures – data presented in revised groupings
Overall impact / Reported census figure will increase by ~200 from July 2016 onwards
Bed days occupied will decrease by ~4% from July 2016 onwards
What data will beavailable?
Census figuresAll changes reflected in the revised data definitions manual and national reporting requirements will be applied to census figures from July 2016 onwards. The result of this is:
- all patients discharged within three working days of the census date will now be included in the census figure
- patients delayed for healthcare reasons or in non hospital locations will no longer be included.
The overall impact of these two changes will show an increase in the reported figure of ~200 delays.
Information published in previous reports cannot be used in direct comparison to figures published from July 2016 onwards. To provide comparable census information the trend data reflects, where possible, both the revised and previous definitions.
The census figure trend available from October 2015 will:
- present high level numbers for all delays by NHS Board
- include the number of patients discharged within three working days of the census date from October 2015 – June 2015 (as submitted by NHS Boards from Oct 2015)
- exclude patients delayed for healthcare reasons or in non hospital locations from historic data up to June 2016.
The census figure trend available from July 2012 will:
- present more detailed information by high level delay reason group, NHS Board and Local Authority
- exclude all patients discharged on or within three working days of the census date from July 2016 data onwards
- exclude patients delayed for healthcare reasons or in non hospital locations from historic data up to June 2016
- the overall impact of this will show a lower figure to previously published data but is comparable across each time period.
Bed days occupied
All changes reflected in the revised data definitions manual will be applied to data from July 2016 onwards. This will:
- exclude patients delayed for healthcare reasons or in non hospital locations (these patients were historically included)
- the overall impact of this will show a decrease of ~4%.
Historic data will not be directly comparable with data from July 2016 onwards and relevant notes will be added to explain this.
The monthly published delayed discharge information is available on the ISD website, where the latest information on delayed discharges, as well as trend data, can be found.