Why are delayed discharges from hospital increasing? Seeing the bigger picture

Background

In December 2022, more than 13,000 of a total of around 100,000 hospital beds in England were occupied by patients who were medically fit for discharge. This number has increased by 57% over the last 2 years, representing an additional 6,000 people waiting in hospital who do not need to be there. This reduces the availability of hospital beds for newly admitted patients, contributing to longer A&E and ambulance waiting times. Delays can also have a significant impact on patients: longer hospital stays increase the risk of hospital-acquired infections and patients may lose mobility and cognitive function, making it harder to regain their independence after leaving hospital and increasing their need for care. Delays can also be stressful for the family and friends of people stuck in hospital.

The policy and media narrative about delayed discharges often focuses on limited capacity in social care as the root cause of delays. In January 2023, the government announced an additional £200m for extra beds in care homes and other settings to speed up hospital discharge. This followed the announcement of a £500m adult social care discharge fund in September 2022. But there are a mix of other factors that can cause problems discharging patients from hospital, including challenges accessing other NHS services in the community and poor hospital processes.

Discussion about what is driving the recent increases in delayed patient discharges is based on a limited set of data collected from hospitals. These data were recently described by Sir David Nicholson, chair of Worcestershire Acute Hospitals NHS Trust and Sandwell and West Birmingham NHS Trust, and a former chief executive of NHS England, as ‘useless’ and ‘designed to show how bad social care is’. So what can we learn from the data about delays in discharging people from hospital?

Trends in delayed discharge from hospital

It is not possible to reliably compare delays in hospital discharge now with pre-pandemic delays, because the way these data are collected changed during the COVID-19 pandemic (Table 1). However, we can still compare trends separately before and after 2020.

Figure 1 shows these trends using the two different data sources. Under the old methodology, the number of delayed transfers of care (DTOCs) remained stable between 2010 and 2014. It then increased, peaking in 2016/17 before falling back again. Numbers were on the rise in early 2020, but were still some way from their previous peak.

When data collection resumed using the new method, the average number of patients no longer meeting the criteria to reside in hospital was around 8,400 in December 2020. A year later it had risen to around 10,000 and reached an average of above 13,000 patients in late 2022. The scale of this increase from December 2020 to December 2022 is unprecedented.

How have discharge wait times changed over time?

Despite some limitations, the data can help us understand how changes in patient flow and discharge wait times have contributed to the recent increase in delayed discharges. As set out above, the number of patients who are medically fit for discharge but remain in hospital at the end of each day reached more than 13,000 in December 2022. This number is determined by both the rate at which patients are discharged (around 10,000 per day), and how long patients wait to be discharged (from being declared fit for discharge until they are actually discharged), as shown below (Figure 2).

Figure 2

We compare data from the first 3 weeks of December 2021 (we use data from December 2021 because earlier data are not publicly available) and December 2022 to look at the relative contribution of these (Table 2).

Between December 2021 and December 2022 the number of patients discharged increased by around 300 per day: this indicates that, far from hospitals becoming less efficient, they have increased their daily patient discharge flow in the past year. However, the number of new patients fit for discharge grew by an average of 650 per day. The inflow of patients ready for discharge was greater than the outflow of patients discharged by around 300 people in December 2022 (Figure 2). This explains the increase in the number of patients experiencing delayed discharge, from 10,619 to 13,399 in the past year: the average discharge wait also increased as a result, from 1.07 to 1.31 days (Table 2). This average includes some patients with no delays (ie they are discharged on the same day as the decision to discharge); most patients who experience delays wait much longer.

The average discharge wait varied widely across regions (Figure 3). In the north west, the region with the longest average discharge wait, the average delay was 2.07 days – more than double that in the east of England and London. There is significant variation between integrated care systems (ICSs) too, with average delays exceeding 3 days in Birmingham and Solihull, Gloucestershire, and Cheshire and Merseyside ICSs, and less than 1 day in nine ICSs, including Mid and South Essex (0.51) and Leicester, Leicestershire and Rutland (0.68).

Where can delays occur along care pathways?

In Figure 4 we consider two possible care pathways involving patients needing support after leaving hospital, when they are discharged either back to their own home or to a community setting (such as a care home). The first pathway is a traditional model, where a full needs assessment takes place in hospital and patients are discharged directly to their longer term place of care. The second is a ‘discharge to assess’ model, under which a limited needs assessment is carried out in hospital and patients are then discharged home or to other community settings including care homes (see Box 1), where a more detailed assessment of their needs takes place to inform their longer term care.

Figure 4: Discharge to assess models aim to bypass some sources of delays which exist in direct discharge pathways for patients with non-acute NHS, social or community care needs

Delays in discharging patients can occur at several points in these pathways.

1. Delays in obtaining assessments of post-acute care needs: these may be due to insufficient NHS or social care capacity to coordinate and carry out these assessments. These delays should be reduced under the discharge to assess model. If a patient is discharged to social care, more extensive assessments need to take place.
2. Delays in organising and coordinating any care and support needed, or in organising a discharge to assess pathway: these are primarily due to a lack of social care beds or care packages, non-acute NHS beds, or community care packages (eg for rehabilitation). Delays can be further compounded if NHS staff are overstretched. Other factors include disputes over funding for social care (following a needs assessment), patients’ needs for post-discharge care changing during this time, and agreement with the family on the discharge plan.
3. Delays in organising discharge itself: this may happen due to limited NHS capacity to prepare discharge papers, medication and medical devices, limited transport availability, or if a service has been secured but will only be available after a certain amount of time.

Delays at each point also cause difficulties to patients and families, including stress related to uncertainty and concern for the patient’s wellbeing, and difficult decisions about loved ones. These delays result from pressures on and capacity in social care and the NHS. We now look at these in turn to understand how each contributes to the increase in delays.

What is driving these increases in discharge delays?

Pressures on social care

Policymakers and the media have often focused on social care as a driver of these increased delays. Pressures on social care services have increased in recent years because of underinvestment and staff shortages. There has been a 3% decrease in filled posts in social care in England since 2020/21 and vacancies have increased by 52% over the same period – now standing at 165,000 vacant posts. These pressures are undoubtedly a significant factor in delays in pre-discharge assessments and in securing placements for patients in need – particularly in London, the east of England and the south east, regions with the greatest unmet need for social care staff.

But available data suggest that waiting for social care services accounts for less than half of discharge delays (Figure 5). The percentage of patients experiencing delays who were waiting for a residential or nursing home placement, or domiciliary care, increased from around a third to just under half during the decade to 2020, but never reached 50%.

Reasons for delays are not directly comparable pre- and post-2020 due to changes in data collection (Table 1). Data from February to December 2022 show that at most 40% of patients staying 1 week or more in hospital and experiencing delays were waiting for a permanent bed in a nursing or residential home, or home care (Figure 6). (Lack of publicly available data prevented us from assessing whether it increased before February 2022.) It is important to note that ‘home care’ includes care provided by the NHS as well as social care, so this percentage is an overestimate of delays due to waiting for social care services. Nonetheless, if social care was singlehandedly driving the rise in delayed discharges, we would expect the percentage of patients waiting for a permanent care home bed or home care to have increased in 2022, alongside discharge delays. Instead, this percentage has decreased since its peak at 40% in February 2022, fluctuating around 37% in September to December 2022. More patients were waiting for short-term rehabilitation (eg in a community hospital) than for a permanent bed in a care home throughout July to December 2022.

Lastly, not all delays among patients waiting for social care services are caused by social care. As well as the reasons for delays, until early 2020 information was collected on which organisation – NHS, social care, or both – was responsible for each delay. Although 45% of delayed patients were waiting for social care services in February 2020, social care was recorded as solely or jointly responsible for only 40% of all delays (the NHS was solely responsible for the remaining 60%). Unfortunately, information on the organisation responsible for delays is no longer collected (since March 2020).

Although social care is a significant contributing factor to delayed discharges, most delays are caused by a combination of other reasons (as was the case pre-COVID).

Pressures on the NHS

Another explanation for recent increases in delayed discharge could be increased pressures on the NHS and its staff, leading to low capacity to perform pre-discharge assessments, set up discharge plans, and coordinate post-discharge care, among other things.

One symptom of these pressures is increased bed occupancy in hospitals. Bed occupancy increased steadily in the decade before COVID-19 (Figure 7a). From its COVID-related dip in mid-2020, it reached 92% among all general and acute beds in the NHS in October–December 2022. Bed occupancy is particularly high for adult general and acute beds, rising from 88% in December 2020 to above 95% in September 2022–January 2023 (Figure 7b).

These bed occupancy levels exceed the maximum threshold of 92% set by the NHS to maintain patient flow this winter, and are far above the 85% threshold recommended by the Royal Colleges of Emergency Medicine and Surgeons. Compared with other countries, the UK has very high levels of hospital bed occupancy. For example, Germany has more than three times the number of beds per inhabitant than the UK, and its bed occupancy was only 68% in 2021. High bed occupancy indicates that wards are functioning at their capacity limit, with high numbers of patients per staff, meaning staff are less able to perform essential assessments and organise post-discharge care for patients needing support. High bed occupancy is known to have adverse effects on patient safety, including higher rates of hospital-acquired infection and inpatient mortality.

High bed occupancy is partly caused by a long-term reduction in the number of hospital beds in England, which decreased by more than 50% between 1987/88 and 2015/16. Although this reduction partly reflects the shift towards more community-based care (eg mental health services) and improvements in productivity (eg increases in day-case surgery), current levels of bed occupancy indicate that bed reduction was implemented without considering changing population needs. Unoccupied beds are essential for helping manage patient flow; too few hospital beds mean wards are vulnerable to spikes in the number of patients admitted. Indeed, delayed discharges would not be in the limelight as an issue of public concern if bed occupancy was at a more manageable level.

Community health and other non-acute NHS services are increasingly stretched too, including from increased use of step-down care as part of discharge efforts, with one in five waiting for short-term rehabilitation (eg in a community hospital) in December 2022.

Lastly, although the total number of hospital admissions is lower than before COVID-19, patients may have more complex needs because of treatment backlogs during the pandemic. This may lead to greater need for complex post-discharge care than pre-pandemic.

Average bed occupancy levels above 92% in hospitals and stretched non-acute NHS services reflect severe pressures on hospitals and community settings. This can delay the provision of patient discharge care – including pre-discharge assessments, communication with family, and coordination with other NHS or social care bodies. Necessary steps often happen one after the other, rather than in parallel, and community services are less able to accept patients, leading to patients with complex discharge needs waiting longer to be discharged.

Conclusion

Social care is often seen as the main cause of delays discharging patients from hospital. Pressures on social care are extreme and many people go without the care they need. But it’s not the whole story: available data suggest that patients waiting for social care services account for less than half of delayed discharges in England. And the considerable pressures on the NHS – including high bed occupancy in hospitals and stretched community services – appear to have contributed to the substantial increase in delays in the last 2 years. These pressures limit capacity to provide timely discharge care, as well as the ability of the NHS and social care services to collaborate smoothly. Factors outside of social care and the NHS may also contribute to delays.

Discharge delays are self-reinforcing: patients kept in hospital longer than necessary are at risk of hospital-acquired infections and increased frailty, which may lead to more complex needs and longer hospital stays. There are also risks to patients if discharge is rushed or premature – including compromising high-quality assessments and patient choice, and avoidable hospital readmissions.

Policies to improve discharge are often short term and focus on the ‘back door’ of hospitals, such as emergency winter funding for social care placements to help patients leave hospital more quickly. There is little published evidence that these short-term funds have made a difference to hospital discharges so far, as the most recent data show there were more than 13,500 patients experiencing delayed discharge in mid-February. But our analysis suggests a broader approach is needed that tackles the mix of causes of delays throughout the system. This includes increasing NHS capacity in hospitals to improve patient flow, identifying and tackling process issues within hospitals that may contribute to delays, and expanding the range of NHS and social care services available to support patients outside hospitals.

The NHS delivery plan for urgent and emergency care services commits to an additional 5,000 staffed hospital beds in 2023/24, and recognises the need to improve joint discharge processes and expand capacity in intermediate care and social care to improve discharge. But the plan includes few new measures and making it happen will be difficult without additional investment and a comprehensive plan for recruiting and retaining staff. The plan also includes no specific measures beyond 2023/24. A longer term approach is needed. REAL Centre analysis suggests that the NHS in England may need an additional 23,000–39,000 hospital beds by 2030/31 to simply deliver 2018/19 rates of care.

Better data are also needed to target action to reduce delayed discharges. As David Nicholson highlights, the quality of SitRep data is doubtful, and efforts should be made to improve it. Publicly available quality-assured data on reasons for delays would enable national policymakers, service planners and managers to make informed decisions to alleviate hospital pressures and allow patients to leave the hospital as soon as they safely can.