Key causes and long-term trends related to emergency department and inpatient hospital admissions of homeless persons in England

Background It is estimated that approximately 300,000 people are experiencing homelessness in England. The aim of this study was to evaluate key causes and long-term trends of emergency departments (EDs) and in hospital inpatient admissions of persons experiencing homelessness in England. Methods ED and hospital inpatient admissions data were obtained from Hospital Episode Statistics (HES) covering all National Health Service (NHS) England hospitals. Anyone identified or declared to be experiencing homelessness during the service usage are recorded in HES datasets. Data were extracted for the 10-year study period and compared to the general population, which includes all patients attending the ED or admitted to inpatient care in England. Results Drug- and alcohol-related causes contribute to the most frequent reasons for attendance and admissions of persons experiencing homelessness in the ED and inpatient respectively. A total of 30,406 ED attendances were recorded for persons experiencing homelessness in the year 2018/2019 (+ 44.9% rise vs 2009/10) of which injuries and poisoning respectively represented 21.8% and 17.9% of all persons experiencing homelessness presentations to the ED. Poisoning (including drug overdose) represented only 1.9% of all attendances by the general population during the same study year (rate ratio vs general populations 9.2 95% CI 9.0–9.4). High mortality rates were observed in relation to presentations attributed to drug- and alcohol-related causes. A total of 14,858 persons experiencing homelessness inpatient admissions were recorded in 2018/2019 (+ 68.6% vs 2009/2010). Psychoactive substance use constituted 12.7% of all admissions in 2018/2019 compared to 0.4% of in the general populations (rate ratio: 33.3, 95% CI: 31.9–34.7). There was a 44.3% rise in the number of admissions related to poisoning in the study period amongst persons experiencing homelessness in England (vs 14.2% in general population). Conclusion Marked disparities around primary causes of ED and inpatient admissions were identified between persons experiencing homelessness and the general population. There is a continued need for prevention measures to reduce the prevalence of drug and alcohol, injury and poisoning-related admissions to the ED, enhanced service provision at the community level, and multisector collaborations. These initiatives should maximise opportunities for early interventions and improve outcomes for persons experiencing homelessness, including increased accessibility of healthcare and mental health services, particularly in areas that demonstrate increasing ED and inpatient attendance rates over time.


Background
Over 300,000 people are known to be currently experiencing homelessness in England [1].Homelessness includes rooflessness (without a shelter of any kind, sleeping rough), houselessness (with a place to sleep but temporary in institutions or shelter), living in insecure housing (threatened with severe exclusion due to insecure tenancies, eviction, domestic violence), or living in inadequate housing (in caravans on illegal campsites, in unfit housing, in extreme overcrowding) [2].
While early deaths and mortality causes in persons experiencing homelessness are well reported in the literature, long-term data trends, with regard to disease epidemiology and factors that require hospitalisations and urgent care needs, have been less well documented.Systematic reviews of international literature suggest that the health status of persons experiencing homelessness is lower than the rest of the population.Persons experiencing homelessness have 12 times higher mortality rate compared with the general population, mainly owing to opioid overdose, psychoactive substance use and related heart failure [3,4].As per recent estimates, persons experiencing homelessness die at an average age of 46 (male) and 43 (female) years, with drug overdose and accidents contributing to the excess mortality [1].Health status worsens with the increasing length of time spent homeless [5].In an attempt to mitigate the negative impacts of homelessness, policy initiatives such as the Homelessness Reduction Act 2017 and the National Health Service (NHS) Long-Term Plan in England have aimed to improve outcomes for persons experiencing homelessness.These include providing support to those who are homeless or at risk of being homeless and increasing access to integrated, tailored services for rough sleepers [6,7].
Persons experiencing homelessness present more often to emergency departments (EDs) than the general population.There is a lack of high-quality studies in England exploring the health conditions, demography, management and discharge outcomes in relation to such presentations.Many use hospitals as their only source of healthcare as they find primary care 'complex' to navigate, experience barriers when accessing services and have negative experiences of service use [8].To our knowledge, long-term data trends around utilisation of ED and inpatient admissions by homeless populations have not been previously explored.Obtaining such data is imperative to identify, strengthen and evaluate appropriate primary care, community and outreachbased prevention programmes and health policies.
Due to overlapping prevalence of substance misuse and severe mental health problems associated with homelessness, much of the current healthcare focus remains on presentations related to these conditions.However, previous studies have demonstrated that important longterm health conditions, such as cardiovascular diseases amongst homeless populations, are often underdiagnosed and undertreated [9].This study analysed ED and inpatient utilisation by persons experiencing homelessness in England and compared this with persons experiencing homelessness datasets relating to the general population.Previous research has shown that drug-and alcohol-related conditions are amongst the most frequent reasons for presentation to the ED amongst persons experiencing homelessness [10]; therefore, specific analysis explored drug-and alcohol-related presentations and admissions in detail.Using in-depth analysis of hospital episode statistics datasets from England, this study compares causes of ED and inpatient admissions by homeless populations between 2009/2010 and 2018/2019.

Study design
This study used retrospectively collected routine data from government sources pertaining to the EDs and inpatient service utilisation and outcomes relating to persons experiencing homelessness in England.General population data relating to all other patients attending the ED or admitted to inpatient care in England, available from NHS Digital [11,12], was used as the comparator.

Setting
Data were extracted from the Hospital Episode Statistics, a routinely collected data source relating to EDs and inpatient admissions.Data specific to persons experiencing homelessness was requested through a data procurement agreement between University of Birmingham and NHS Digital.Persons experiencing homelessness-specific data were identified by searching patient records for specific postcode fields that are used to record homelessness when patients with no fixed abode present to EDs and inpatient services.Datasets from 10 years (2009/2010 to 2018/2019) were extracted and analysed.ED attendance and inpatient admissions data relevant to all English general populations over the study period were used as a comparator population.

Data extraction ED admissions data
Counts of all attendance over the study period and specific to the two-character level primary diagnosis codes were extracted for both persons experiencing homelessness and the general populations.In addition, data on age category, sex, ethnicity (White, non-White and unknown), referral methods (e.g.emergency services, general practitioner (GP), local authority, police), arrival modes (ambulance, other, unknown), admissions methods (elective, emergency and other) and attendance disposal methods (to hospital; died, other and unknown) were extracted.

Inpatient admissions data
A count of all consultant episodes over the study period and specific to the three-character level primary diagnosis codes as per the International Classification of Diseases (ICD) codes were extracted for both persons experiencing homelessness and the general populations.In addition, data on age category, ethnicity (White, non-White and unknown), admissions methods (elective, emergency, other), discharge methods (e.g. through clinical consent, self-discharged, still in hospital) and disposal methods (i.e.dead, alive and not known) were extracted.

Data collection and management
Data search was run by NHS Digital, the national information and technology partner to the health and social care system.All data were anonymised, and small numbers suppressed to protect the anonymity and confidentiality prior to release to the research team.All evaluation materials were stored, processed and destroyed in accordance with University of Birmingham research governance policies.This study was reviewed and approved by Data Access Request Service (DARS) Review Committee of NHS Digital (approval reference number NIC-341255-H2F7H).

Analysis
Histograms presenting the numbers of persons with each diagnosis were plotted for all included persons.For all included persons, the number and percentage of all key causes of ED presentations and inpatient admissions were descriptively presented.In depth analyses of poisoning, drug-and alcohol-related causes were separately presented.All persons experiencing homelessness data from England were compared to the general population ED attendance and inpatient admission datasets.A time trend plot was constructed to illustrate the changes in ED presentations and inpatient admissions pattern over the 10-year study period.Comparisons between years were performed using ratios of proportions and 95% confidence intervals.2, Fig. 1).The total consultation for drug and alcohol problems in people experiencing homelessness saw an 44.9% increase from 3865 in 2009/2010 to 8948 attendance in 2018/2019.Whereas in the general populations, the increase was 132.7% with numbers increasing from 156,860 in 2009/2010 to 364,948 in 2018/2019.Other conditions that saw an increase in admissions amongst persons experiencing homelessness related to infectious disease (174%), septicaemia (683%), other vascular conditions (151%), respiratory conditions (104%), diabetes and other endocrinological conditions (101%) and psychiatric conditions (144%).In the general population, septicaemia saw the highest increase in ED admissions of 1135% during the study period.

Emergency department presentations
Amongst the persons experiencing homelessness presenting with drug-and alcohol-related causes in 2018/2019, approximately 83% were male (77% males for all causes), 70.4% were of white ethnicity (68.0% for all causes), and nearly twice as many (73.5%)arrived by ambulance compared to all causes (43.2%).
Approximately, 0.12% of all persons experiencing homelessness who presented to ED during 2018/2019 died in the ED compared to 0.10% of deaths in general population (rate ratio 1.23, 95% CI: 0.94 to 1.60).Nearly 1 in 5 who presented with drug-and alcohol-related problems in the ED died compared to 0.1% of recorded deaths for all diagnoses in 2018/19.However, there was a reduction in mortality in the ED due to drug-and alcohol-related problems during the study period (31.0% in 2009/2010 vs 19.9% in 2018/2019).
Nearly half of all arrivals of persons experiencing homelessness in the ED were via ambulance (44.7%) compared to 22.7% in the general population (rate ratio 1.97, 95% CI: 1.95-1.99).A total of 5.7% (n = 2518) of all persons experiencing homelessness ED attendance were related to referrals by the police.This was the second most common mode of referral to the ED after self-referral (n = 27,652, 62.7%) in England.Seasonal variations in presentations were low.Presentations during admissions quarter 1 (April-June), quarter 2 (July-September), quarter 3 (October-December) and quarter 4 (January-March) were 25.4%, 27.0%, 24.3%, and 23.3% respectively in 2018/19.

Adult inpatient admissions
There were a total of 14,858 (2018/19)  Injuries were recorded as the most prevalent cause of persons experiencing homelessness inpatient admissions in 2018/2019 (Table 3), which saw a total of 2155 admissions.This was followed by mental and behavioural disorders due to psychoactive substance use (n = 1881) and poisoning (n = 1664).Psychoactive substance use constituted 12.7% of all admissions of persons experiencing homelessness to inpatient units compared to 0.4% of admissions in the general populations (rate ratio: 33.27, 95% CI: 31.87-34.72].Similarly, 11.2% of all admissions to inpatient units in 2018/2019 were contributed by poisoning compared to 0.8% contribution in the general population (rate ratio: 14.55, 95% CI: 13.90-15.23).
Mental health conditions such as schizophrenia, schizotypal and delusional disorders, mood (affective) disorders and cellulitis are also featured as important causes of admissions.For example, in 2018/2019, a total of 569 and 274 inpatient admissions were noted for schizophrenia, schizotypal and delusional disorders and mood (affective)   disorders respectively amongst people experiencing homelessness.Another 1881 admissions were related to mental and behavioural disorders due to psychoactive substance use.In the English general population, cancers, injuries, influenza and pneumonia (3.0%), diseases of oesophagus, stomach and duodenum and anaemias and ischaemic heart diseases were featured amongst the most important causes of admissions in 2018/2019 (Table 3).Amongst these, sepsis (800%) saw the biggest increase.In the general population in England, sepsis (728.5%),influenza and pneumonia (130.1%)saw a significant increase in the number of admissions during the study period.A total of 24.3% (3612) of all persons experiencing homelessness inpatient admissions in 2018/2019 were attributed to drug-and alcohol-related causes in England (Tables 4 and 5).Of these, a total of 1531 admissions were related to 'mental and behavioural disorder due to the use of alcohol' (Table 5).There was a 6.8% increase (compared to 13.7% increase in the general population in England) in the number of admissions categorised as 'mental and behavioural disorders due to psychoactive substance use' during the 10-year study period.This included admissions related to the use of alcohol, opioids, cannabinoids, cocaine and polysubstance use.Mental and behavioural disorders due to use of cocaine (+ 480%) and 'other stimulants, incl.caffeine' (360.0%),saw the highest increase in admission number amongst persons experiencing homelessness during the study period.In comparison, admissions related to mental and behavioural disorders due to use of tobacco, cocaine and cannabinoids saw the largest increase amongst the general populations in England (Table 5).
There was a 44.3% rise in the number of admissions related to poisoning amongst persons experiencing homelessness in England during the study period (compared to 14.2% rise in the general population) (Tables 4  and 5).Poisoning included toxic effects of alcohol, drugs    of abuse, prescribed drugs with no abuse potential and corrosive substances and other known poisons.

Self-discharge without clinical consent
A total of 3415 out of 14,858 (23.0%) persons are experiencing homelessness self-discharged (i.e.without clinical consent) from the inpatient units in 2018/2019.The data was not available to compare this figure with the general population.

Summary of key findings
The findings of this study provide an overview of the secondary healthcare utilisation by persons experiencing homelessness in England over the 10-year study period.
Comparisons were drawn with data from the general population covering all ED attendance and inpatient admissions.
Cases of poisoning contributed to over 1 in 5 presentations to the ED by persons experiencing homelessness.The observed rate was over 10 times higher compared to the rates in the general population.Of note, during the 10-year study period, poisoning-related attendances amongst persons experiencing homelessness increased by over 160% in England.The rate of increase was higher than those observed in the English general population.
This study shows that drug-and alcohol-related admissions contributed to high mortality in the ED (approximately 20%) compared to 0.1% for all diagnoses.
However, over the study period, mortality rates in the ED due to drug-and alcohol-related causes saw significant reductions.
Psychoactive substance use contributed to the most frequent cause of inpatient admissions in persons experiencing homelessness.Key causes of admissions were markedly different when compared with the general population.In the latter, cancers, respiratory and cardiovascular causes dominated the key causes of admissions.Causes such as sepsis, leg ulcers and diabetes saw the largest increase in the number of persons experiencing homelessness inpatient admissions during the study period.Furthermore, for mental health, drug-and alcohol-related diagnoses such as schizophrenia, personality disorders and poisoning, the rate of increase was markedly higher in persons experiencing populations.These findings corroborate with persons experiencing homelessness mortality data as reported by the Office of the National Statistics in England which demonstrate that drug-and alcohol-related conditions are amongst the most significant causes of mortality [5].

Study strengths and limitations
This is the first study evaluating the causes of ED presentations and inpatient admissions by persons experiencing homelessness using nationally collected datasets from England.Consideration of 10 years of data allowed time trends to be analysed.It is important to note that recording of homelessness is often not complete or accurate in emergency departments and hospital inpatient settings.We relied on the postcode fields used to record the domicile of anyone with no fixed abode when they present for services.Therefore, the data presented in this report may not represent all attendance and outcomes of persons experiencing homelessness in the two study settings.Changes in homelessness numbers, recording practices and evolution of services, particularly those aimed at persons experiencing homelessness, need to be taken into consideration when interpreting the time trend data.For example, lately, there is an increasing emphasis on the identification of persons experiencing homelessness in the hospital and primary care settings since the introduction of the Homelessness Reduction Act [6].In addition, there is a lack of accurate data in regards to the number of persons experiencing homelessness in England as official sources often refer to the 'households' [13].We were unable to estimate the incident rates of persons experiencing homelessness utilising ED and inpatient due to the aforementioned factors.It is important to note that homelessness has increased in England.Given the lack of accurate figures on the number of experiencing homelessness as described above and by instead using measures of households (not individuals) in temporary accommodation, there has been a 65.7% increase in homelessness in England during the study period 2010 (n = 51,310) v 2019 (n = 85,040) [13].Therefore, the time trend estimates in relation to persons experiencing homelessness presentations should be interpreted accordingly.The link between homelessness and substance and/or alcohol dependence is well understood in the literature; however, this study allowed all key health conditions to be systematically examined in relation to their contribution.Although the presentations and admissions due to nondrug-and alcohol-related causes were lower amongst persons experiencing homelessness, literature demonstrates that they experience higher burden of multimorbidity and mortality rates attributed to these conditions compared to the general population [5,14].

Implications for practice
This study reinforces the need to improve the provision of mental health and substance misuse-related support to persons experiencing homelessness in the community.As is well rehearsed, prevention measures should be further strengthened to address the health inequalities faced by this population.
Previous studies show that persons experiencing homelessness are underrepresented in the mainstream practices [15].While specialist primary healthcare centres for homeless persons have been established, mainstream services need to be further inclusive of persons experiencing homelessness.In addition, outreach-based and community services are best able to serve this population [16][17][18][19][20]. Training and education of frontline staff at mainstream general practices are required to reinforce the registration guidelines [21].Entry criteria to primary care, mental health and substance misuse services for persons experiencing homelessness need to be reviewed in order to increase accessibility.As indicated in the Public Health England Guidance [22] providers of alcohol and drug, mental health and other services need to adopt an open-door policy for individuals with co-occurring conditions and should ensure every contact opportunity counts.Treatment for any of the co-occurring conditions should be available through every contact point.Services such as needle exchange and naloxone provision need to be readily available in the community, including through community pharmacies.It has been reported that the COVID-19 pandemic impacted heavily on provision of drug and alcohol services as well as those serving the homeless populations [23,24].It is important that lessons and good practices from the pandemic are considered for future service provision.These include the use of remote-and technology-assisted services to serve the affected populations which were reported to have been well received by the clients [23,24].
There is also a requirement for compliance with the Homelessness Reduction Act 2017 [6] to ensure healthcare settings, particularly hospitals, proactively identify persons experiencing homelessness and work collaboratively with social services to offer support to them or those threatened with homelessness.Furthermore, this study has reinforced the need to adhere to the goals set out in the NHS Long-Term Plan in order to provide outreach services and invest in mental health support for persons experiencing homelessness.A comprehensive health needs assessment tool, for use by primary care, mental health, substance misuse and public health practitioners, is needed in order to support practitioners and increase their confidence when addressing complex issues.
Future studies should consider accessing individual medical notes and health-related data from multiple sources to triangulate the findings.In this study, it was not possible to investigate repeat ED attendance by a person due to lack of identifiable personal data and postcodes.There is a need for research investigating repeat attendance and associated reasons and in addition to develop integrated models of care to address multiple morbidities including overlapping substance misuse and mental health [25][26][27].Published literature report that persons experiencing homelessness are likely to attend the ED far more than the general population, with wide variations in the figures reported [28].For example, persons experiencing homelessness have been reported to use the ED 5 [29], 20 [30] or 60 [14] times more compared to the general population.

Conclusion
Poisoning and drug-and alcohol-related causes contribute to the most frequent reasons for attendance and admissions of persons experiencing homelessness to the ED and inpatient settings.The key causes of healthcare utilisation differ markedly from the general population data.There is a need to increase outreach-based services and support and review entry criteria to primary care mental health and substance misuse services for persons experiencing homelessness.There is a continued need for prevention measures, development of outreach-based support and enhanced service provision at the community level.Multisector collaborations are needed to maximise opportunities for early interventions.The data presented here relates to pre-COVID-19 pandemic timelines, and further research is required to assess the wider impacts of the pandemic on homelessness in England.A longitudinal evaluation will enable the identification of how the above data trends have changed as a result of the pandemic and support provisions made by the government and local authorities to alleviate the impact of the pandemic on persons experiencing homelessness.

Table 1
All causes of ED presentations by persons experiencing homelessness (PEH)

Table 2
Drug-and alcohol-related ED presentations by persons experiencing homelessness (PEH) (England)

Table 3
Causes of inpatient admissions persons experiencing homelessness (PEH)

Table 4
Drug-and alcohol-related inpatient admissions amongst persons experiencing homelessness (PEH) in England

Table 5
Drug-and alcohol-related inpatient admissions amongst persons experiencing homelessness (PEH) in England-detailed causes