Table 1.
Reviewed literature on homelessness in the ED
| Author(s), year | Study year | Country | Study aim | Study design and study duration | Setting and study population | Number and/or proportion of unique patients who are homeless | Number and/or proportion of ED visits made by homeless persons | Key reasons for presentation to the ED | Mean number of ED visits per person, per year | |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Tadros et al. 2016 [15] | 2016 | USA | To analyse changes in ED utilisation of homeless patients and compare that with non-homeless visits | Comparative analysis of the 2005 and 2010 NHAMCS dataset | Patients presenting to non-federal hospital ED and outpatient departments | In 2010, 679,854 (0.55%) out of 124,043,357 presentations to the ED were made by homeless persons | In 2010, homeless persons made an average of 5.8 visits per person per year compared with 1.7 by non-homeless persons | ||
| 2 | Oates et al. 2009 [16] | 2009 | USA | To analyse the national utilisation of the ED by the homeless population | Cross-sectional, secondary analysis of data using NHAMCS dataset | Patients presenting to non-federal hospital ED and outpatient departments | In 2005, 472,922 (0.41%) out of 115,322,815 presentations to the ED were made by homeless persons | |||
| 3 | Holtyn et al. 2017 [17] | 2017 | USA | To examine the relationship between ED utilisation and alcohol use in homeless alcohol-dependent adults | Analysis of self-report assessments of alcohol and emergency department use alongside random breath collections | Homeless, alcohol-dependent (met DSM-IV criteria) adults from an inpatient detoxification unit and homeless community agencies | Out of 86 recorded ED visits, 29.1% presented due to alcohol intoxication, 4.7% for alcohol withdrawal, 2.3% for drug/medication overdose, 11.6% for psychological problems and 18.6% for medical problems | Average of 4.4 ED visits per person per year | ||
| 4 | Brown et al. 2010 [18] | 2010 | UK | To determine whether the rate of attendance was related to the outside temperature | A retrospective study of routine ED computer records from 2003 to 2008 | Patients presenting to the Northern General Hospital ED and data from the Weston Park Weather Station | 2930 (0.55%) out of 528,573 visits to the ED were made by persons identified as homeless | |||
| 5 | Cheung et al. 2015 [19] | 2015 | Canada | To examine the relationship between ED use and substance dependence among homeless individuals with concurrent mental illness who participated in a ‘Housing First’ intervention trial | Analysis of administrative data and findings from Vancouver At Home survey data | Homeless or precariously housed individuals who met criteria for a mental disorder with or without concurrent substance use dependence; administrative data collected from six urban hospitals in the Vancouver Coastal Health Authority | Average of 2.1 ED visits per person, per year | |||
| 6 | Brown et al. 2013 [20] | 2013 | USA | To compare the ED visit characteristics of younger homeless adults with those of older homeless adults | Analysis of a systematic random sample of ED visits using NHAMCS dataset from 2005 to 2009 | Patients presenting to non-federal hospital ED and outpatient departments | 2,808,000 (0.6%) out of 468,000,000 ED visits were made by homeless adults | Psychiatric issues were more frequent in younger than older homeless adults (23% vs 15%; P = 0.01). Older homeless adults were more likely to suffer injuries (28% vs 21%; P = .04) and cardiovascular complaints (11% vs 5%; P = 0.02) than younger homeless adults | ||
| 7 | Raven et al. 2017 [21] | 2017 | USA | To identify ED use patterns and factors associated with ED use in adults 50 and older | Initial screen for study eligibility followed by analysis of baseline interview and medical records | Random sample of adults aged 50 years and older from homeless encampments, recycling centres, overnight homeless shelters and meal programmes | Out of 348 recorded visits to the ED, 23.9% presented for chronic illness, 21.6% for new illness, 19.2% for pain, 16.4% for injury, 8.3% for analgesic requirement and 5.8% for mental health issues | |||
| 8 | Ku et al. 2010 [22] | 2010 | USA | To assess whether homelessness or associated characteristics independently predicted ED use | Descriptive, cross-sectional secondary analysis of ED visits using NHAMCS dataset for the years 2005 and 2006 | Patients presenting to non-federal hospital ED and outpatient departments | 1.1 million (0.5%) out of 234 million weighted ED attendances during the 2-year time frame were by homeless patients | Out of 550,000 recorded visits to the ED, 304,000 (55.3%) patients presented due to injuries, 100,000 (18.3%) due to alcohol or other drug use, 57,000 (10.4%) due to psychiatric diagnoses and 38,500 (7%) due to respiratory diagnoses | Average of 0.72 ED visits per person, per year | |
| 9 | Feldman et al. 2017 [23] | 2017 | USA | To explore whether prevalence of homelessness in the ED varied between weekdays and weekends and between seasons | Prospective, 5-question homelessness screening survey of eligible participants attending the ED between May 2015 and February 2016 | Patients, who are not critically ill, registered with 3 EDs in north-eastern Pennsylvania | 309 (7.03%) out of 4395 participants were cited as experiencing homelessness | |||
| 10 | Jackson et al. 2019 [24] | 2019 | USA | To describe demographics and proportion of ED patients who have experienced homelessness | Cross-sectional survey of a convenience sample of patients presenting to the ED from September to December 2016 | Patients presenting to Urban Atlanta ED | 475 (51.5%) out of 923 ED patients who completed the survey stated some degree of homelessness in the previous year | |||
| 11 | Lee et al. 2019 [25] | 2019 | Australia | To compare the prevalence of homelessness in consecutive patients presenting to the ED | Prospective screening of housing status and retrospective audit of administrative data for patients presenting to the ED during a 7-day period in 2017 |
Patients presenting to an inner metropolitan hospital ED in Melbourne Sample size: 1275 ED presentations involving 1208 individual patients (7-day period) |
40 (7.9%) of the 504 prospectively screened patients were identified as homeless and 16 (2.3%) of the 704 non-screened patients were identified as homeless | |||
| 12 | Tsai et al. 2013 (a) [26] | 2013 | USA | To examine the proportion of homeless veterans among users of Veteran Affairs EDs and compare homeless and housed VA ED users’ clinical characteristics | Cross-sectional study analysing national VA ED user’s administrative data from the fiscal year 2010 | Homeless veterans presenting to VA EDs | 64,091 (6.89%) VA ED users identified as homeless out of 930,712 veterans that utilised VA EDs | Out of 64,091 recorded visits to the ED, 13.55% presented for alcohol disorder, 11.92% for drug disorder, 37.72% for psychiatric diagnoses (35.68% for non-substance misuse related), 12.84% for any pain diagnosis, 26.30% for congestive heart failure and 7.49% for chronic pulmonary disease | Average of 3.38 (SD=4.01) ED visits per person per year compared with 2.07 (SD=1.09) for non-homeless users | |
| 13 | Rodriguez et al. 2009 [27] | 2009 | USA | To determine the extent that people experiencing homeless present to the ED for social issues | Prospective case-control study conducting interviews between July 2006 and March 2007 | Patients in the treatment areas of one urban hospital ED | 9806 (19.5%) out of 50,172 visits to the ED in 2006 | Out of 191 homeless patients, 29% stated that hunger, safety and lack of shelter were the primary reasons for presenting to the ED | Average of 5.8 ED visits per person, per year (SD= 2.2) | |
| 14 | Lin et al. 2015 [28] | 2015 | USA | To determine which factors are associated with frequent ED visits and hospitalisations among the insured homeless population | Retrospective, cross-sectional study using BHCHP electronic database from January to December 2010 | Homeless Medicaid recipients who received service from BHCHP | Out of 25,771 recorded visits to the ED, 15.2% of patients presented for alcohol-related disorders, 7.6% for psychiatric disorders (not including substance misuse-related conditions), 5.3% for drug-related disorders, 14% for injury and poisoning, 7% for respiratory disorders and 5% for circulatory disorders | Average of 3.97 ED visits per person, per year | ||
| 15 | Mackelprang et al. 2014 [29] | 2014 | USA | To describe injury characteristics and circumstances among individuals identified as homeless in the ED | Cross-sectional, case-control study using the NEISS database between January 2007 and December 2011 | Patients with product-related injuries who presented to NEISS EDs | 268 (0.0142%) out of 1,885,274 unique cases that presented to NEISS ED’s with product-related injuries involved a homeless person | Out of 268 recorded visits to the ED, 13.8% had alcohol involvement and 3.4% had drug/substance use involvement | ||
| 16 | Doran et al. 2016 [30] | 2016 | USA | To quantify the presence of housing instability, homelessness, and other selected social determinants of health in ED patients | Cross-sectional survey of a random sample of ED patients from June to August 2014 | Patients presenting to an urban public hospital ED | Out of 625 visits to the ED, 19.6% reported homelessness or lack of stable housing in the past 2 months | |||
| 17 | Moore et al. 2011 [31] | 2011 | Australia | To describe patterns of service use and predict risk factors for re-presentation to an ED among homeless persons | Retrospective analysis using computerised patient administration system from January 2003 to December 2004 | Patients presenting to a principal referral hospital ED | 1595 (3.9%) out of 40,942 individual patients were homeless | 6689 (10.4%) out of 64,177 visits to the ED were made by the homeless population | Average of 2.1 ED visits per person, per year | |
| 18 | Hammig et al. 2014 [32] | 2014 | USA | To determine the clinical characteristics of homeless patients presenting to the ED, focusing on unintentional and intentional injury events and related comorbid conditions | Retrospective cohort study analysing ED visits from the NHAMCS database from 2007 to 2010 | Patients presenting to non-federal hospital ED and outpatient departments | 603,000 (0.5%) out of 119,993,000 visits to the ED annually were made by homeless patients | Out of 603,000 reorded visits to the ED, 55% were injury related and 45% were non-injury related | ||
| 19 | Mackelprang et al. 2015 [33] | 2015 | USA | To analyse the prevalence and characteristics of ED and inpatient admissions among homeless and unstably housed youth | Retrospective cohort study using electronic medical records from July 2009 to June 2012 | Patients presenting to the ED or inpatient departments of two urban teaching hospitals | Out of 1151 recorded visits to the ED, 30.06% were injury related, 23.28% were due to psychiatric illness, 7.99% were alcohol related, 21.29% were drug related and 57.34% were due to a chronic medical condition | Average of 0.97 ED visits per person, per year | ||
| 20 | Feldman et al. 2018 [34] | 2018 | USA | To assess the prevalence of homelessness by gender | Retrospective survey from May 2015 to February 2016 | Patients presenting to 3 EDs (a level trauma centre, a suburban hospital and an inner-city hospital) | 309 (7%) out of 4395 unique participants were homeless | |||
| 21 | Tsai et al. 2013 (b) [35] | 2013 | USA | To determine the ED use among homeless and domiciled VA service users | Retrospective cohort study using VA administrative workload databases from fiscal year 2010 | Homeless and domiciled veterans presenting to the ED | 64,099 (6.89%) out of 930,598 visits to the ED were made by homeless people | |||
| 22 | Moulin et al. 2018 [36] | 2018 | USA | To determine the ED utilisation for patients with a primary mental health diagnosis | Retrospective analysis of OSHPD data from 2009 to 2014 | Patients with a primary mental illness visiting acute care hospitals’ EDs | 6153 (0.73%) out of 846,867 visits made to the ED by adult patients with mental illness were by homeless ED users | |||
| 23 | Cheallaigh et al. 2017 [37] | 2017 | Ireland | To compare the use of unscheduled ED and inpatient care between housed and homeless patients | Observational cross-sectional study using electronic patient data in 2015 | All ED visits and unscheduled admissions to one teaching hospital | 2966 (6.3%) out of 47,174 ED attendances were made by homeless patients | Out of 2966 recorded visits to the ED, 7.6% presented for overdose and poisoning, 6.6% for alcohol-related issues, 5.6% for head injury, 4.8% for mental illness, 3.8% for abdominal pain and 2.9% for chest pain | Average of 3 ED visits per person per year and housed individuals had an average of 0.16 ED visits per, person per year | |
| 24 | Yeniocak et al. 2017 [38] | 2017 | Turkey | To determine the sociodemographic and clinical characteristics of Turkish homeless patients who were brought to the ED by ambulance | Retrospective cross-sectional study from January to December 2014 | Homeless adult patients brought to a Tertiary Training and Research Hospital by ambulance | 167 (0.0835%) homeless patients attended the ED which serves an average of 200,000 patients each year | Out of 167 visits to the ED, 14.7% presented due to respiratory difficulty, 12.57% due to abdominal pain, 23.35% for clouded consciousness, 15.57% for generally impaired condition, 7.78% for traffic incidents and 5.39% for sharp object injury | ||
| 25 | Lloyd et al. 2017 [39] | 2017 | Australia | To understand the profile and expressed needs of people seen by HEDLO in the ED in comparison to the general hospital population | Retrospective chart audit of data recorded in ED referral database and HEDLO files from October 2013 to January 2015 | Homeless persons referred to HEDLOs in Queensland Health ED |
117,996 presentations to the ED over 16-month period. Of these, 221 homeless people were referred to HEDLO |
Out of 221 recorded visits to the ED, 25% presented due to mental health, 19% due to alcohol- and other drug-related issues, 39% for chronic medical conditions and 15% for social reasons | ||
| 26 | Lombardi et al. 2019 [40] | 2019 | USA | To analyse national survey data to elucidate the differences between homeless and non-homeless patients’ ED visits | Retrospective study using NHAMCS dataset from 2005 to 2015 | Patients presenting to non-federal hospital ED and outpatient departments | 2750 (0.91%) out of 303, 326 visits to the ED were made by homeless persons | Out of 2750 recorded visits to the ED, 28.4% presented due to psychiatric diagnoses, (16.29% were not substance misuse related) 17.7% were drug use related, 1.2% were alcohol related, 1.78% were respiratory related and 1.09% were cardiovascular related | ||
| 27 | Hastings et al. 2013 [41] | 2013 | USA | To determine predictors of repeat health service use in older veterans treated and released from the ED | Retrospective cohort study analysing VHA administrative datasets and the Vitals Mini File from 1 October 2007 to 30 June 2008 | Patients aged 65 or over who were treated and released from a Veterans Affairs Medical Centre ED or urgent care clinic | 374 (1.2%) out of 31,206 visits to the ED were made by homeless veterans | |||
| 28 | Lam et al. 2016 [42] | 2016 | USA | To assesses the impact of homelessness on 30-day ED revisits and hospital readmissions among patients presenting with mental disorders | Secondary analysis of administrative data in the ED looking at visits made in 2012 | Homeless patients presenting to the ED in an urban, safety-net hospital | 4210 (4.6%) 0ut of 92,307 unique patients were homeless at any time during the study period | 15,159 (10.9%) out of 139,414 visits to the ED were made by persons who were homeless at any time during the study period | Out of 15,159 recorded visits to the ED, 39.25% presentations were mental disorders and 60.75% were non-mental disorders | |
| 29 | Stenius-Ayoade. 2017 [43] | 2017 | Finland | To examine the role of mental disorders in relation to the use of primary healthcare services among homeless shelters in Helsinki | Retrospective analysis of electronic health records made by physicians and nurses working in primary health care from 2005 to 2008 | Homeless persons in 4 shelters operating in the Helsinki metropolitan area | Out of 587 recorded visits to the Primary Health Care Emergency Rooms, 11% were for mental health and substance abuse, 38% were for trauma, 11% were for infections and 19% were for intoxications and convulsions | |||
| 30 | Post et al. 2013 [44] | 2013 | USA | To determine the prevalence and types of ‘new media’ use among homeless patients who present to the ED | Observational cross-sectional survey from July to August 2012 | Patients presenting to 3 urban, high-volume EDs in Connecticut | 249 (4.3%) out of 5788 subjects enrolled in the study, reported episodes of homelessness in the past year. | |||
| 31 | Moore et al. 2012 [45] | 2012 | Australia | To evaluate the accuracy of a predictive model to identify homeless people at risk of re-presentation to the ED | Prospective cohort study conducted from 1 April 2009 to 30 April 2009 | Patients presenting to an adult, tertiary referral hospital ED, excluding those who died during study period | 211 (7.31%) out of 2888 unique individuals who visited the ED were homeless | 327 (9.92%) out of 3298 visits to the ED were made by homeless persons | ||
| 32 | Doran et al. 2018 [46] | 2018 | USA | To characterise alcohol and drug use in a sample of homeless vs. non-homeless ED patients | Baseline survey interviews with patients at public hospital ED from November 2016 to September 2017 | Random sample of patients who presented to an urban public hospital ED | 316 (13.69%) out of 2309 patients were currently experiencing homelessness | Out of 316 recorded visits to the ED, 25% were substance use related | ||
| 33 | Doran et al. 2013 [47] | 2013 | USA | To determine what multi-dimensional patient-level factors are most strongly associated with a 6-level gradient of VHA ED use | Cross-sectional analysis of data obtained from national VHA databases for fiscal year 2010 | Veterans presenting to VHA ED services | 64,091 (6.9%) out of 930,712 patients who visited the ED were homeless | |||
| 34 | Ku et al. 2014 [48] | 2014 | USA | To examine the study characteristics and costs associated with homeless ED frequent users | Retrospective cross-sectional review of hospital and financial records for ED visits in 2006 | Frequent users of the ED in an urban academic medical centre with a level 1 trauma and annual census of greater than 60,000 visits | 74 (13.7%) out 542 frequent users were homeless | 845 (15.5%) out of 5440 visits made by frequent users were made by homeless persons | Out of the 845 presentations to the ED, 12.9% were due to substance abuse, 10.9% were nervous system related, 8.9% were respiratory problems, 7.1% were cardiovascular problems and 8.3% were due to traumatic disorders | |
| 35 | Coe et al. 2015 [49] | 2015 | USA | To compare homeless patients’ utilisation of the urban ED in the USA with non-homeless patients | Cross-sectional study of the NHAMCS-ED electronic database for 2009 to 2010 | Patients presenting to non-federal hospital ED and outpatient departments | 1,302,256 (0.65%) out of 200,645,347 visits to the ED were made by homeless patients | |||
| 36 | Amato et al. 2018 [50] | 2018 | USA | To compare emergency care utilisation between individuals with documented homelessness to those enrolled in Medicaid without documented homelessness | Retrospective cohort study using medical chart review between for the years 2013 and 2014 | Patients presenting to a single, urban, academic, tertiary care centre | 7532 (5.17%) out of 145,662 visits to the ED were made by persons with documented homelessness | Out of 7532 recorded visits to the ED, 20.1% of patients presented for mental health disorders, 13.4% were alcohol related, 1.12% were for drug overdose, 9.3% were for abdominal pain, 8.7% were for chest pain and 7.8% were for trauma |
BHCHP, Boston Health Care for the Homeless Program; DSM, Diagnostic and Statistical Manual; ED, emergency department; HEDLO, Homeless Emergency Department Liaison Officers; DSM-IV: NEISS, National Electronic Injury Surveillance System; NHAMCS, National Hospital Ambulatory Care Survey; OSHPD, California’s Office of Statewide Health Planning and Development; VA, Veteran Affairs; VHA, Veteran Health Affairs