What do the data on hospital discharge delays really tell us?
The limitations of existing data and suggestions to improve data quality
Authored by Ellen Coughlan, Caroline MacNish Porter & Francesca Cavallaro
In December 2023, almost 13,000 patients in England were in hospital awaiting discharge despite being medically fit, an increase of 8,000 compared with December 2020. Tackling these delays has been an important focus for policymakers and NHS managers, particularly given the knock-on effects on waiting times for A&E and ambulances. It is also distressing for patients having to be kept in hospital longer than necessary and increases the risk of hospital-acquired infections and increased frailty, which may lead to more complex needs and longer hospital stays.
We published an analysis of delayed hospital discharges in March 2023. In this we used Urgent and Emergency Care Situation Reports (SitRep) data to explore changes in the number of delayed patients and reasons for these delays since the start of the COVID-19 pandemic.
SitRep delayed discharge data are reported by trusts, often by discharge teams. Each day, they report the total number of patients who are medically fit for discharge (‘no longer meeting the criteria to reside’), and of these, the number of patients discharged (including discharge pathways) or remaining in hospital. Reasons for discharge delays among patients staying in hospital for at least 7 and 21 days are reported weekly. Data collection is time-consuming and may take several hours per day in trusts where data systems aren’t set up to collect this information, time that could be better spent coordinating discharges.
These SitRep data are the only available data for England on delayed discharges since October 2020. Despite this, the data quality is poor. Since we published our initial analysis, conversations with analysts and experts across health and care have improved our understanding of its limitations. In this article, we outline these limitations, and present recommendations for improving the quality and utility of data on delayed discharges.
Key limitations of SitRep data
1. Lack of standardised definitions and practices
Differences between trusts
The number of delayed patients is not recorded consistently across trusts. The ‘criteria to reside in hospital’ — used to determine which patients have a delayed discharge — have been questioned by some clinicians. This has led to local adaptations in how these definitions are applied.
The list of reasons for delays is subjective, not exhaustive, and mixes different causes together. For example, ‘awaiting […] rehabilitation bed’ includes both patients waiting for temporary NHS and social care beds, making it impossible to distinguish between the two. Additionally, only one reason for discharge delay can be recorded even if many apply, meaning that we do not have an accurate overall picture of local bottlenecks.
Changes over time
It’s likely that the recording of delayed patients has changed over time. Trusts may have improved their methods for counting patients, or been incentivised to undercount patients due to policy updates and the introduction of different key performance indicators (KPIs). This makes it difficult to know how much of the increase in discharge delays over the past 2 years is real, and how much is due to changes in reporting.
Similarly, there have been changes in how reasons for delays are recorded. For example, increased political and media attention may have encouraged trusts to report a greater number of delayed patients as waiting for social care. Data reported by Caja Consultancy and obtained through FOI, seen by us, show wide fluctuations in reasons recorded for delayed discharges for each integrated care board (ICB). Abrupt changes within ICBs are likely to reflect arbitrary changes to the way data are recorded, making it difficult to know how the causes of delays change over time.
2. Incomplete data
On any given day between November 2022 and March 2023, up to 6 of 121 NHS England trusts did not report data on discharge delays. Therefore, the daily total number of delayed patients is usually underestimated, and some of the day-to-day variation might be due to the different number of trusts reporting, rather than actual changes in delays.
3. Problems using and interpreting discharge delay data
Poor indicator of trust performance
More broadly, SitRep delayed discharge data misrepresent trust performance. Patients ‘no longer meeting the criteria to reside’ are not necessarily ‘ready’ for discharge, in the sense that practical steps for discharge may not have yet been put in place. Therefore, some patients recorded as ‘delayed’ in fact fall within an expected timeframe for discharge care to be organised. For example, a new social care placement typically takes a minimum of a week to organise after a patient is considered medically fit to be discharged. Additionally, some delays may be caused by factors beyond a trust’s control (for example, availability of rehabilitation beds in community hospitals or social care).
Lack of documentation
There is limited documentation describing what is captured in the SitRep delayed discharge dataset. This makes it difficult for analysts to use and interpret data, and for leaders to make informed decisions about how to improve patient flow.
Ways to improve delayed discharge SitRep data
One of the main challenges facing the NHS today is delayed hospital discharges. As the only data available to understand the extent and causes of this issue in England, the poor quality of SitRep data is a problem. Given the burden of data collection and the importance of federated data systems, as highlighted in Data Saves Lives, we need to ensure that data is both robust and routinely released in the public domain to help understand local and national trends.
A positive step that has recently been taken is the publishing of data on delayed discharges on a routine, ongoing basis which allows for monitoring of local and national trends. Work is ongoing within NHS England to improve the data on reasons for delay and to improve the accuracy of data on length of delays in line with commitments made in the Delivery Plan for Urgent and Emergency Care Services.
Based on our work, and the conversations with analysts and managers, we have identified several suggestions to help improve the quality of SitRep data and reduce the data collection burden on hospital staff, whose time could otherwise be spent discharging patients:
· Revert back to weekly (rather than daily) reporting given limited quality of data, as was the case pre-COVID-19 with delayed transfer of care (DTOC) data.
· Improve definitions of the ‘criteria to reside’, discharge pathways and reasons for delays, through consultation with clinicians and managers.
· Ensure trusts only report data not reported elsewhere (such as total number of discharges), supporting aim described in Data Saves Lives to reduce the data collection burden on frontline NHS staff.
· Build information requirements, for example, for discharge pathway and reason for delays, into routine patient data systems, making it easier for staff to collate.
We believe these changes can empower trusts to use good data well, provide a better understanding of delayed hospital discharges, both at a national and local level, and improve how bottlenecks are addressed across health and care.
If you are working on solving this question too, we would love to hear from you. Join the conversation on X (formerly Twitter) or get in touch by emailing us.
