- Email: [email protected]
Food, Shelter and Safety Needs Motivating Homeless Persons' Visits to an Urban Emergency Department
HEALTH POLICY AND CLINICAL PRACTICE/BRIEF RESEARCH REPORT
Food, Shelter and Safety Needs Motivating Homeless Persons’ Visits to an Urban Emergency Department Robert M. Rodriguez, MD Jonathan Fortman, BS Chris Chee, BS Valerie Ng, BS Daniel Poon, BS
From the Departments of Medicine and Emergency Medicine, San Francisco General Hospital, University of California San Francisco School of Medicine, San Francisco, CA.
Study objectives: We determine whether homeless persons present to the emergency department (ED) for food, shelter, and safety and whether the availability of alternative sites for provision of these needs might decrease their ED presentations. Methods: In July to August 2006 and February to March 2007, adult homeless and control (not homeless) patients, who self-presented (nonambulance) to an urban county ED, were interviewed with a structured instrument. Results: One hundred ninety-one homeless and 63 control subjects were enrolled. Homeless persons spent a mean (standard deviation [SD]) of 3.5 (3.0) nights/week sleeping without shelter and ate a mean (SD) of 2.1 (1.1) meals per day; 51% stated they had been assaulted on the street. On an analog scale, in which 0⫽no problem and 10⫽worst possible problem in their daily lives, the mean (SD) homeless subject responses for hunger, lack of shelter, and safety were 4.8 (3.7), 6.1 (4.2), and 5.1 (4.0), respectively. More homeless (29% [55/189]) than not homeless (10% [6/63]) persons replied that hunger, safety concerns, and lack of shelter were reasons they came to the ED (⌬⫽20%; 95% confidence interval 10% to 29%). If offered a place that would provide food, shelter, and safety at all times, 24% of homeless subjects stated they would not have come to the ED. Conclusion: Homeless persons commonly come to the ED for food, shelter, and safety. Provision of these subsistence needs at all times at another site may decrease their ED presentations. [Ann Emerg Med. 2009; 53:598-602.] 0196-0644/$-see front matter Copyright © 2008 by the American College of Emergency Physicians. doi:10.1016/j.annemergmed.2008.07.046
SEE EDITORIAL, P. 603. INTRODUCTION In addition to higher chronic disease burden, homeless peoples’ higher rates of emergency department (ED) utilization has been attributed to their lack of access to primary care clinics and other medical care entrance points.1-4 EDs function as the health care safety net for millions of homeless persons and other disenfranchised individuals.4,5 As the only site where federal law requires that anyone seeking help be treated at any hour, EDs may also be filling another key safety net role—that being a place of last resort for the homeless and others to receive basic subsistence needs (food, shelter, and safety).6,7 The ED is thus not only an essential medical provider but also an underappreciated, vital social welfare institution, without which many indigent people might 598 Annals of Emergency Medicine
not survive.7 Although most physicians and others certainly recognize the value of provision of subsistence needs to the poor, the dilemma is this: Visits for nonmedical needs may clog an already overburdened ED system, delaying care for acutely ill patients and siphoning off scarce resources.2 Examining general populations of ED patients, investigators have characterized their social deprivation, such as lack of food and home heating, and how hunger affects food versus medicine decisions.8-10 The degree to which subsistence needs contribute to ED visits, however, has not been reported. Our objective in this prospective study was to characterize the effect that subsistence needs have on homeless persons’ decisions to present to the ED. Specifically, we sought to determine whether homeless persons present to the ED for food, shelter, and safety and whether the availability of alternative sites for provision of these needs might decrease their ED presentations. Volume , . : May
Rodriguez et al
Nonmedical Needs Motivating Visits of the Homeless
Editor’s Capsule Summary
What is already known on this topic Homeless persons who present to emergency departments (EDs) frequently have both medical and social problems. What question this study addressed To what extent homeless persons use the ED for social issues such as the unavailability of adequate food, shelter, and safety. What this study adds to our knowledge This 254-patient case-control study in a single urban ED showed that the homeless reported greater problems obtaining food, shelter, and safety than the nonhomeless. Twenty-four percent of homeless persons indicated that they would not have come to the ED had they had other options for food, shelter, and safety. How this might change clinical practice These data should lead to confirming studies including examination of whether alternative resources actually result in reduced ED visits by the homeless.
MATERIALS AND METHODS We conducted this prospective case-control study in the ED of an urban county hospital, which had 50,172 patient visits in 2006, of which 9,806 (19.5%) were by homeless persons. During an 8-week block (July to August 2006) and a 6-week block (February to March 2007), all adult homeless patients in the treatment areas of the ED (not the waiting room) between 8 AM and 2 PM and 8 PM and midnight on 4 weekdays were considered for enrollment. We chose these daily periods because a 1-week trial period and review of triage logs revealed that these were the times with the highest number of homeless person presentations. Selection of Participants The following patients were excluded: (1) trauma patients; (2) patients transferred from another hospital, chronic care facility, nursing home, or hospice; (3) patients unable to participate in an interview because of intoxication, altered mental status, or critical illness; (4) incarcerated patients; (5) patients undergoing psychiatric holds; and (6) patients brought in by ambulance or by police—these were excluded because we sought to examine patients who made their own decision to present to the ED. We defined “homelessness” according to the standard method used in other studies as lack of stable housing for the Volume , . : May
preceding 2 months.11 We initially identified homeless patients according to a data field on their registration sheet. At our institution, registration personnel ask all patients a series of questions about their current housing; “homeless” designation on registration is inclusive— even someone who has lost their housing for a few days is designated as such. To determine whether patients designated homeless by registration met our criteria for homelessness, they were asked about their housing for the previous 6 months. Using the same exclusion criteria during the same blocks of time, control subjects were concurrently enrolled on a 1:3 (control:homeless) basis. After enrollment of 3 homeless subjects, the next (according to triage time) nonhomeless patient who met inclusion and exclusion criteria was approached for consent and interviewed in the same manner as the homeless persons. Control (not homeless) persons were also initially identified according to their registration data field and were asked the same questions to confirm their previous stable housing. Two faculty experts in health care literacy, who were not otherwise involved in the study, reviewed the interview instrument for content validity, and we pilot-tested it on 6 homeless persons to confirm test-retest validity. In addition to demographic questions, the instrument consisted of 14 questions assessing subsistence needs and the effect of these needs on their ED presentation: 8 yes/no questions (Is hunger a main reason you came to the ED today?); 3 analog scale questions (On a scale in which 0⫽no problem and 10⫽worst possible problem, how much of a problem is hunger in your daily life?); and 3 other numeric-answer questions (On an average day, how many meals do you eat?). See Appendix E1 (available online at http://www.annemergmed.com) for the interview instrument. Four second-year medical students and 2 postbaccalaureate students, all of whom received a 4-hour orientation session to ensure standard interview technique, conducted the interviews. Subjects were consented and interviewed in private rooms and semiprivate (screen-partitioned) gurneys during their ED stay. The number of ED visits in the past year, insurance status, and total minutes (triage to discharge time) for the index ED visit of subjects were derived from review of computerized patient records. Our hospital’s institutional review board approved the study. All data were entered into Microsoft Excel (Microsoft, Redmond, WA) using double data entry checking. Statistical tests were performed using Stata v 9.0 (StataCorp, College Station, TX). Demographic data were summarized and reported in aggregate form. Responses to questions about hunger, shelter, safety, and usual medical care were tabulated and reported as frequency percentages. The mean plus SD of analog scale questions and the mean difference in proportions with 95% confidence intervals (CIs) of the homeless and control groups were calculated. Annals of Emergency Medicine 599
Nonmedical Needs Motivating Visits of the Homeless Table 1. Subject characteristics. Homeless, nⴝ191
Not Homeless, nⴝ61
148 (77) 43 (23) 45.5 (11)
37 (61) 24 (39) 45.3 (14.4)
67 (35) 93 (49) 21 (11) 5 (3) 3 (2) 17 (9) 39/121 (32) 62/132 (47) 27/123 (22) 426 (289) 5.8 (2.2) 45.7 (93.4)
25 (41) 22 (36) 8 (13) 5 (8) 1 (2) 8 (13) 11 (18) 8 (13) 5 (8) 391 (321) 3.0 (4.9) N/A
Characteristic Sex (%) Male Female Mean age, y (SD) Race/ethnicity (%) White Black White-Hispanic Asian-American Other Admitted to hospital (%) Alcohol use (%) Illicit drug use (%) Injection of drugs (%) Mean No. of minutes in ED (SD) Mean No. ED visits past year (SD) Mean No. months homeless (SD) Site of usual health care (%) EDs Free clinics Private clinics Other or none Insurance status Medicare Medi-Cal Private None
74 (39) 70 (37) 9 (5) 38 (20)
17 (28) 15 (25) 18 (30) 11 (18)
21 (11) 47 (25) 0 122 (64)
14 (23) 11 (18) 6 (10) 30 (49)
SD, standard deviation; ED, emergency department; N/A, not asked.
Rodriguez et al 3.7) versus 1.6 (SD 2.9) on the 0 to 10 analog scale. In the previous 3 months, 63% of homeless and 8% of control subjects reported eating less than they wanted. Lack of shelter was a greater problem for the homeless group, 6.1 (SD 4.2) versus 3.5 (SD 4.0); 17% of homeless subjects stated that lack of shelter was a primary reason for their presentation to the ED. Despite considerably colder ambient temperatures during the winter (mean 10.4°C [50.7°F]; range 3.8°C to 18.8°C [39°F to 66°F]) than the summer (mean 18.5°C [65.3°F]; range 11°C to 31°C [52°F to 88°F]) enrollment blocks of our study,12 there was no important difference between the blocks in terms of the percentage of homeless individuals who cited shelter as a reason for presentation. Spending an average (SD) of 3.5 (3.0) nights per week sleeping on the street, only 31% of homeless persons stated they had adequate clothing. On the 0 to 10 scale, personal safety was a greater problem for homeless persons, mean (SD) 5.1 (4.0) versus 1.8 (3.2). More than half of homeless subjects reported having been assaulted on the streets, and 15% stated that safety was a primary reason for their ED presentation. Overall, more homeless persons (29% [55/189]) than not homeless persons (10% [6/63]) replied that hunger, safety concerns, and lack of shelter were primary reasons they came to the ED (⌬⫽20%; 95% CI 10% to 29%). If they could go to a place that would provide food, shelter, and safety at all hours, 24% (46/189) of homeless subjects stated they would not have come to the ED.
LIMITATIONS RESULTS During the total 14-week study period, there were 2,698 ED visits by homeless persons. Of the 503 homeless persons who arrived during study enrollment times, 66 came by ambulance and 203 had other exclusion criteria, leaving 234 eligible subjects approached. Seven of these patients refused to be interviewed and 36 others fell asleep or could not otherwise participate meaningfully in the interviews, resulting in a total of 191 homeless subjects interviewed. Of 110 control (not homeless) patients initially screened, 18 arrived by ambulance and 20 had other exclusion criteria, leaving 72 not homeless patients initially approached. Three of these refused to be interviewed and 6 others fell asleep or could not otherwise participate meaningfully in the interviews, resulting in 63 control subjects concurrently enrolled. See Table 1 for a summary of subject characteristics. Similar percentages (79% and 79%) of homeless and control subjects believed that their presenting problems were emergencies, but fewer (40% versus 57%) homeless persons believed that they could get their problem cared for in a clinic (⌬⫽17%; 95% CI 3% to 32%). See Table 2 for a summary of subject responses to interview questions. Homeless persons reported eating fewer mean meals per day (standard deviation [SD] 2.1 [1.1] versus 3.6 [SD 5.9]) and cited hunger as a greater problem in their daily lives, 4.8 (SD 600 Annals of Emergency Medicine
Our results are subject to the limitations of other interviewbased studies, most notably social desirability bias and a failure of subjects to respond to all of the questions. Given that patients may be reluctant to admit that they present for food, shelter, or safety instead of for true medical care, our findings may underestimate the effect of subsistence needs on ED presentation. We did not elucidate the reasons for incomplete survey responses. Although we established test-retest validity on a pilot group of patients, other measures of validity were not confirmed. Our exclusion of patients with true medical need (trauma, intoxicated, critically ill, and transported by ambulance) artificially inflates the percentage of homeless persons who we determined presented for subsistence needs. Assuming that standard triage would leave less severely ill homeless persons in the waiting room, however, this inflation may be countered to a certain degree by our restriction of interviews to homeless patients within the ED care area. Given that this was a singlecenter study in a city with a large homeless population, our findings may not apply to EDs in other cities. The most important limitations of our study are the inherent difficulties in defining the true reasons patients came to the ED and in making the distinction between true medical need and true social/welfare need. Many patients likely presented for both needs, and in this population hunger, shelter, and safety may be Volume , . : May
Rodriguez et al
Nonmedical Needs Motivating Visits of the Homeless
Table 2. Summary of subject responses to interview questions. Yes Responses Question Hunger Mean No. meals/day (SD) Currently receive food stamps (%) How much of a problem is hunger in your daily life?* (SD) Last 3 mo, eaten less than you wanted because of no money (%) Is hunger a main reason you came to ED today? (%) Shelter/clothing Mean No. nights/week on street (SD) How much of a problem is lack of shelter in your life?* (SD) Do you have adequate clothing? (%) Lack of shelter a main reason came to ED today? (%) Safety Have you ever been assaulted or attacked on street? (%) Mean No. of times, if responded yes (SD) How much of a problem for you is safety on streets?* (SD) Fear for safety a main reason came to ED today? Do you believe your problem is an emergency today? (%) Could you go to a clinic instead for this problem? (%)
Homeless, nⴝ191
Not Homeless, nⴝ63
2.1 (1.1) 39/189 (21) 4.8 (3.7) 119/189 (63) 25/189 (13)
3.6 (5.9) 10 (16) 1.6 (2.9) 5 (8) 3 (5)
3.5 (3.0) 6.1 (4.2) 57/185 (31) 32/185 (17)
N/A 3.5 (4.0) N/A N/A
92/182 (51) 10.4 (53) 5.1 (4.0) 27/182 (15%) 142/179 (79) 68/169 (40)
0 0 1.8 (3.2) 4 (6%) 50 (79) 35/61 (57)
*Analog scale 0-10: 0⫽no problem, 10⫽worst possible problem.
intrinsically tied to and perceived as medical conditions. Any measure targeting subsistence presentations to the ED must consider these links between social and medical need.
DISCUSSION Although the ED’s position as the health care safety net is well established and understood by medical and lay communities, the fact that EDs also commonly serve as an urban welfare safety net may be less obvious to those outside of ED staff.6,7 In this study, we confirmed what many ED health care providers have assumed—that nonmedical subsistence needs are a common reason for homeless patients to present to the ED. Almost a third of homeless persons walking into our ED cited hunger, safety, and lack of shelter as primary reasons for their presentation. In this era of severe ED crowding and closures, any extra burden imposed by nonemergency care needs of the homeless and others may further jeopardize not only the ED’s vital health care safety net function but also care for critically ill patients.13,14 To redirect homeless persons seeking health care from the ED to clinics and other sites of care, O’Toole et al15 proposed a renewed focus on outreach, education, and increased medical insurance coverage. Other investigators have reported decreases in ED utilization by homeless persons after implementation of intensive aid and primary care programs.2,16 In a study examining the effect of the federally funded Health Care for the Homeless Program, Han and Wells2 reported that having 2 or more Health Care for the Homeless Program visits was associated with fewer inappropriate ED visits. Similarly, in a Canadian ED trial, Redelmeier et al16 reported a one-third reduction of return ED visits in homeless persons randomized to a program of satisfaction directed “compassionate care.” Volume , . : May
Kushel et al,11 however, have proposed that to decrease homeless ED utilization, it may not be enough to merely provide health insurance and alternative ambulatory care settings. Our findings support this more comprehensive view of the homeless individual–ED problem, specifically, that any program directed at decreasing homeless person ED visits must also address subsistence needs presentations. An obvious primary measure to reduce subsistence needs presentations is the establishment of alternative sites where homeless persons can receive basic necessities at all times. The first step in referral of subsistence needs patients to these alternative sites away from the ED would be identifying and distinguishing them from other true medical needs patients. Incorporation of standard questions about hunger, shelter, and safety at ED triage could allow patients to express these needs in a nonjudgmental setting, without the requirement of presenting a medical chief complaint. Patients with primary subsistence problems could thereby opt out of ED care and receive these necessities outside of ED treatment areas or by transport to an always open shelter. In our study, 24% of subjects reported that they would have utilized such a center instead of the ED if it were always available. Provision of food, shelter, and safety without utilization of ED beds for even a fraction of homeless person presentations would undoubtedly save time and resources. Gordon7 outlined several models of such “social triage centers,” in which EDs could provide a variety of screening, referral, and care management options. The author implemented a pilot program of one of these models, which identified many patients with substantial social deprivations and was well received by patients and health care workers.17 Although some may posit that these alternative basic needs sites already exist in the form of homeless shelters, shelter Annals of Emergency Medicine 601
Nonmedical Needs Motivating Visits of the Homeless
Rodriguez et al
systems are overburdened, with most closing to new occupants late at night. San Francisco currently has 9 homeless shelters with approximately 1,200 beds, and an annually conducted homeless person count revealed a conservative estimate of 6,377 homeless persons in San Francisco in 2007.18 In an assessment of shelter referrals conducted from mid-April until mid-May 2008 by the Social Work Department at San Francisco General Hospital, only 15 of 53 (28%) of the homeless persons in the ED seeking shelter referral were successfully placed, primarily because of lack of beds. The ED serves as a welfare safety net for homeless persons who commonly present because of hunger, lack of shelter, and safety concerns. Provision of these subsistence needs at all times at another site may decrease their ED presentations. Supervising editor: David L. Schriger, MD, MPH Author contributions: RMR conceived and oversaw the study. RMR and JF analyzed the data. JF, CC, VN, and DP interviewed patients. JF finalized the data. CC, VN, and DP assisted with the background research for the project. RMR takes responsibility for the paper as a whole. Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article that might create any potential conflict of interest. The authors have stated that no such relationships exist. See the Manuscript Submission Agreement in this issue for examples of specific conflicts covered by this statement. Publication dates: Received for publication March 7, 2008. Revisions received June 6, 2008, and July 16, 2008. Accepted for publication July 23, 2008. Available online October 5, 2008. Reprints not available from the authors. Address for correspondence: Robert M. Rodriguez, MD, Department of Emergency Services, San Francisco General Hospital, 1001 Potrero Avenue, San Francisco 94110; E-mail [email protected].
REFERENCES 1. Kushel MB, Vittinghoff E, Haas JS. Factors associated with the health care utilization of homeless persons. JAMA. 2001;285: 200-206.
2. Han B, Wells BL. Inappropriate emergency department visits and use of the Health Care for the Homeless Program services by homeless adults in the northeastern United States. J Public Health Manag Pract. 2003;9:530-537. 3. Little GF, Watson DP. The homeless in the emergency department: a patient profile. J Accid Emerg Med. 1996;13:415417. 4. Richardson LD, Hwang U. America’s health care safety net: intact or unraveling? Acad Emerg Med. 2001;11:1056-1063. 5. Ong Eng Hock M, Ornato JP, Cosby C, et al. Should the emergency department be society’s health safety net? J Public Health Policy. 2005;3:269-281. 6. Adams JG. Confronting hunger in the emergency department. Acad Emerg Med. 1999;6:1082-1084. 7. Gordon JA. The hospital emergency department as a social welfare institution. Ann Emerg Med. 1999;33:321-325. 8. Biros MH, Hoffman PL, Resch K. The prevalence and perceived health consequences of hunger in emergency department patient populations. Acad Emerg Med. 2005;12:310-317. 9. Kersey MA, Beran MS, McGovern PG, et al. The prevalence and effects of hunger in an emergency department patient population. Acad Emerg Med. 1999;6:1109-1114. 10. D’Amore J, Hung O, Chiang W, et al. The epidemiology of the homeless population and its impact on an urban emergency department. Acad Emerg Med. 2001;8:1051-1055. 11. Kushel MB, Perry S, Bangsberg D, et al. Emergency department use among the homeless and marginally housed: results from a community-based study. Am J Public Health. 2002;92:778-784. 12. National Weather Service Forecast Office. San Francisco Bay area/Monterey. Available at: http://www.weather.gov/climate/ index.php?wfo⫽mtr. Accessed May 13, 2008. 13. Cowan RM, Trzeciak S. Clinical review: emergency department overcrowding and the potential impact on the critically ill. Crit Care. 2005;3:291-295. 14. Trzeciak S, Rivers EP. Emergency department overcrowding in the United States: an emerging threat to patient safety and public health. J Emerg Med. 2003;5:402-405. 15. O’Toole TP, Gibbon JL, Hanusa BH, et al. Preferences for sites of care among urban homeless and housed poor adults. J Gen Intern Med. 1999;14:599-605. 16. Redelmeier DA, Molin JP, Tibshirani RJ. A randomized trial of compassionate care for the homeless in an emergency department. Lancet. 1995;345:1131-1134. 17. Gordon JA, Chudnofsky CR, Hayward RA. Where health and welfare meet: social deprivation among patients in the emergency department. J Urban Health. 2001;78:104-111. 18. Swarns RL. Chronic homeless count down 30%, government says reduction from ’05 to ’07 may reflect changes in reporting. San Francisco Chronicle. July 30, 2008;A-2. Available at: http://www.sfgate.com/cgibin/article.cgi?f⫽/c/a/2008/07/30/ MNAF121EL3.DTL&hw⫽homeless&sn⫽001&sc⫽1000. Accessed July 31, 2008.
Did you know? You can personalize the new Annals of Emergency Medicine Web site to meet your individual needs.
Visit www.annemergmed.com today to see what else is new online!
602 Annals of Emergency Medicine
Volume , . : May
APPENDIX E1. ED data extraction form-study patients. Questions: How long have you been homeless/without permanent shelter?______ I. Hunger: A. On a typical day, how many meals do you eat? B. Where do you get your food? C. Do you receive food stamps? □Yes □No D. On a scale of 0 to 10 (where 0 ⫽ no problem at all and 10 ⫽ worst possible problem), how much of a problem is hunger for you in your daily life? ____/10 E. In the last 3 months, have you ever eaten less than you felt you should have because there wasn’t enough money for food? □Yes □No F. Was lack of food/hunger one of the main reasons you came to the ED today? □Yes □No ● Have you ever come to the ED for food/hunger reasons in the past? □Yes___# □No G. If you could go to a place that would provide you with free food at all hours, would you have come to the ED today? □Yes □No II. Shelter: A. In a typical week, how many nights do you sleep/spend on the street without shelter? _______ B. If you have shelter some nights, what type of place is it (a house, apartment, group shelter)? ___________________ C. On a scale of 0 to 10 (where 0 ⫽ no problem at all and 10 ⫽ worst possible problem), how much of a problem is lack of shelter for you in your daily life? ___/10 D. Do you feel you have adequate clothing when you are without shelter? □Yes □No
Volume , . : May
E. Was lack of a place to sleep/get out of the cold one of the main reasons you came to the ED today? □Yes □No ● Have you ever come to the ED for shelter in the past? □Yes___# □No F. If you could go to a place that would provide you with free shelter at all hours, would you have come to the ED today? □Yes □No III. Primary care: A. Where do you usually get your healthcare? Emergency departments Private clinics Free clinics None B. Do you believe your problem today is an emergency? □Yes □No C. If you could go to a clinic at any hour for your current problem, would you have come to the ED today? □Yes □No IV. Safety: A. In your time as a homeless person, have you ever been assaulted or attacked? □Yes □No B. On a scale of 0 to 10 (where 0 ⫽ no problem at all and 10 ⫽ worst possible problem), how much of a problem is personal safety on the streets for you in your daily life? ___/10 C. Was fear for your safety one of the main reasons you came to the ED today? □Yes □No ● Have you ever come to the ED for safety reasons in the past? □Yes___# □No D. If you could go to a place that would provide you with safety at all hours, would you have come to the ED today? □Yes □No V. If you could go to a place that would provide you food, shelter and safety at all hours would you have come to the ED today? □Yes □No Is there anything else you would like to tell me?
Annals of Emergency Medicine 602.e1
Food, Shelter and Safety Needs Motivating Homeless Persons’ Visits to an Urban Emergency Department Robert M. Rodriguez, MD Jonathan Fortman, BS Chris Chee, BS Valerie Ng, BS Daniel Poon, BS
From the Departments of Medicine and Emergency Medicine, San Francisco General Hospital, University of California San Francisco School of Medicine, San Francisco, CA.
Study objectives: We determine whether homeless persons present to the emergency department (ED) for food, shelter, and safety and whether the availability of alternative sites for provision of these needs might decrease their ED presentations. Methods: In July to August 2006 and February to March 2007, adult homeless and control (not homeless) patients, who self-presented (nonambulance) to an urban county ED, were interviewed with a structured instrument. Results: One hundred ninety-one homeless and 63 control subjects were enrolled. Homeless persons spent a mean (standard deviation [SD]) of 3.5 (3.0) nights/week sleeping without shelter and ate a mean (SD) of 2.1 (1.1) meals per day; 51% stated they had been assaulted on the street. On an analog scale, in which 0⫽no problem and 10⫽worst possible problem in their daily lives, the mean (SD) homeless subject responses for hunger, lack of shelter, and safety were 4.8 (3.7), 6.1 (4.2), and 5.1 (4.0), respectively. More homeless (29% [55/189]) than not homeless (10% [6/63]) persons replied that hunger, safety concerns, and lack of shelter were reasons they came to the ED (⌬⫽20%; 95% confidence interval 10% to 29%). If offered a place that would provide food, shelter, and safety at all times, 24% of homeless subjects stated they would not have come to the ED. Conclusion: Homeless persons commonly come to the ED for food, shelter, and safety. Provision of these subsistence needs at all times at another site may decrease their ED presentations. [Ann Emerg Med. 2009; 53:598-602.] 0196-0644/$-see front matter Copyright © 2008 by the American College of Emergency Physicians. doi:10.1016/j.annemergmed.2008.07.046
SEE EDITORIAL, P. 603. INTRODUCTION In addition to higher chronic disease burden, homeless peoples’ higher rates of emergency department (ED) utilization has been attributed to their lack of access to primary care clinics and other medical care entrance points.1-4 EDs function as the health care safety net for millions of homeless persons and other disenfranchised individuals.4,5 As the only site where federal law requires that anyone seeking help be treated at any hour, EDs may also be filling another key safety net role—that being a place of last resort for the homeless and others to receive basic subsistence needs (food, shelter, and safety).6,7 The ED is thus not only an essential medical provider but also an underappreciated, vital social welfare institution, without which many indigent people might 598 Annals of Emergency Medicine
not survive.7 Although most physicians and others certainly recognize the value of provision of subsistence needs to the poor, the dilemma is this: Visits for nonmedical needs may clog an already overburdened ED system, delaying care for acutely ill patients and siphoning off scarce resources.2 Examining general populations of ED patients, investigators have characterized their social deprivation, such as lack of food and home heating, and how hunger affects food versus medicine decisions.8-10 The degree to which subsistence needs contribute to ED visits, however, has not been reported. Our objective in this prospective study was to characterize the effect that subsistence needs have on homeless persons’ decisions to present to the ED. Specifically, we sought to determine whether homeless persons present to the ED for food, shelter, and safety and whether the availability of alternative sites for provision of these needs might decrease their ED presentations. Volume , . : May
Rodriguez et al
Nonmedical Needs Motivating Visits of the Homeless
Editor’s Capsule Summary
What is already known on this topic Homeless persons who present to emergency departments (EDs) frequently have both medical and social problems. What question this study addressed To what extent homeless persons use the ED for social issues such as the unavailability of adequate food, shelter, and safety. What this study adds to our knowledge This 254-patient case-control study in a single urban ED showed that the homeless reported greater problems obtaining food, shelter, and safety than the nonhomeless. Twenty-four percent of homeless persons indicated that they would not have come to the ED had they had other options for food, shelter, and safety. How this might change clinical practice These data should lead to confirming studies including examination of whether alternative resources actually result in reduced ED visits by the homeless.
MATERIALS AND METHODS We conducted this prospective case-control study in the ED of an urban county hospital, which had 50,172 patient visits in 2006, of which 9,806 (19.5%) were by homeless persons. During an 8-week block (July to August 2006) and a 6-week block (February to March 2007), all adult homeless patients in the treatment areas of the ED (not the waiting room) between 8 AM and 2 PM and 8 PM and midnight on 4 weekdays were considered for enrollment. We chose these daily periods because a 1-week trial period and review of triage logs revealed that these were the times with the highest number of homeless person presentations. Selection of Participants The following patients were excluded: (1) trauma patients; (2) patients transferred from another hospital, chronic care facility, nursing home, or hospice; (3) patients unable to participate in an interview because of intoxication, altered mental status, or critical illness; (4) incarcerated patients; (5) patients undergoing psychiatric holds; and (6) patients brought in by ambulance or by police—these were excluded because we sought to examine patients who made their own decision to present to the ED. We defined “homelessness” according to the standard method used in other studies as lack of stable housing for the Volume , . : May
preceding 2 months.11 We initially identified homeless patients according to a data field on their registration sheet. At our institution, registration personnel ask all patients a series of questions about their current housing; “homeless” designation on registration is inclusive— even someone who has lost their housing for a few days is designated as such. To determine whether patients designated homeless by registration met our criteria for homelessness, they were asked about their housing for the previous 6 months. Using the same exclusion criteria during the same blocks of time, control subjects were concurrently enrolled on a 1:3 (control:homeless) basis. After enrollment of 3 homeless subjects, the next (according to triage time) nonhomeless patient who met inclusion and exclusion criteria was approached for consent and interviewed in the same manner as the homeless persons. Control (not homeless) persons were also initially identified according to their registration data field and were asked the same questions to confirm their previous stable housing. Two faculty experts in health care literacy, who were not otherwise involved in the study, reviewed the interview instrument for content validity, and we pilot-tested it on 6 homeless persons to confirm test-retest validity. In addition to demographic questions, the instrument consisted of 14 questions assessing subsistence needs and the effect of these needs on their ED presentation: 8 yes/no questions (Is hunger a main reason you came to the ED today?); 3 analog scale questions (On a scale in which 0⫽no problem and 10⫽worst possible problem, how much of a problem is hunger in your daily life?); and 3 other numeric-answer questions (On an average day, how many meals do you eat?). See Appendix E1 (available online at http://www.annemergmed.com) for the interview instrument. Four second-year medical students and 2 postbaccalaureate students, all of whom received a 4-hour orientation session to ensure standard interview technique, conducted the interviews. Subjects were consented and interviewed in private rooms and semiprivate (screen-partitioned) gurneys during their ED stay. The number of ED visits in the past year, insurance status, and total minutes (triage to discharge time) for the index ED visit of subjects were derived from review of computerized patient records. Our hospital’s institutional review board approved the study. All data were entered into Microsoft Excel (Microsoft, Redmond, WA) using double data entry checking. Statistical tests were performed using Stata v 9.0 (StataCorp, College Station, TX). Demographic data were summarized and reported in aggregate form. Responses to questions about hunger, shelter, safety, and usual medical care were tabulated and reported as frequency percentages. The mean plus SD of analog scale questions and the mean difference in proportions with 95% confidence intervals (CIs) of the homeless and control groups were calculated. Annals of Emergency Medicine 599
Nonmedical Needs Motivating Visits of the Homeless Table 1. Subject characteristics. Homeless, nⴝ191
Not Homeless, nⴝ61
148 (77) 43 (23) 45.5 (11)
37 (61) 24 (39) 45.3 (14.4)
67 (35) 93 (49) 21 (11) 5 (3) 3 (2) 17 (9) 39/121 (32) 62/132 (47) 27/123 (22) 426 (289) 5.8 (2.2) 45.7 (93.4)
25 (41) 22 (36) 8 (13) 5 (8) 1 (2) 8 (13) 11 (18) 8 (13) 5 (8) 391 (321) 3.0 (4.9) N/A
Characteristic Sex (%) Male Female Mean age, y (SD) Race/ethnicity (%) White Black White-Hispanic Asian-American Other Admitted to hospital (%) Alcohol use (%) Illicit drug use (%) Injection of drugs (%) Mean No. of minutes in ED (SD) Mean No. ED visits past year (SD) Mean No. months homeless (SD) Site of usual health care (%) EDs Free clinics Private clinics Other or none Insurance status Medicare Medi-Cal Private None
74 (39) 70 (37) 9 (5) 38 (20)
17 (28) 15 (25) 18 (30) 11 (18)
21 (11) 47 (25) 0 122 (64)
14 (23) 11 (18) 6 (10) 30 (49)
SD, standard deviation; ED, emergency department; N/A, not asked.
Rodriguez et al 3.7) versus 1.6 (SD 2.9) on the 0 to 10 analog scale. In the previous 3 months, 63% of homeless and 8% of control subjects reported eating less than they wanted. Lack of shelter was a greater problem for the homeless group, 6.1 (SD 4.2) versus 3.5 (SD 4.0); 17% of homeless subjects stated that lack of shelter was a primary reason for their presentation to the ED. Despite considerably colder ambient temperatures during the winter (mean 10.4°C [50.7°F]; range 3.8°C to 18.8°C [39°F to 66°F]) than the summer (mean 18.5°C [65.3°F]; range 11°C to 31°C [52°F to 88°F]) enrollment blocks of our study,12 there was no important difference between the blocks in terms of the percentage of homeless individuals who cited shelter as a reason for presentation. Spending an average (SD) of 3.5 (3.0) nights per week sleeping on the street, only 31% of homeless persons stated they had adequate clothing. On the 0 to 10 scale, personal safety was a greater problem for homeless persons, mean (SD) 5.1 (4.0) versus 1.8 (3.2). More than half of homeless subjects reported having been assaulted on the streets, and 15% stated that safety was a primary reason for their ED presentation. Overall, more homeless persons (29% [55/189]) than not homeless persons (10% [6/63]) replied that hunger, safety concerns, and lack of shelter were primary reasons they came to the ED (⌬⫽20%; 95% CI 10% to 29%). If they could go to a place that would provide food, shelter, and safety at all hours, 24% (46/189) of homeless subjects stated they would not have come to the ED.
LIMITATIONS RESULTS During the total 14-week study period, there were 2,698 ED visits by homeless persons. Of the 503 homeless persons who arrived during study enrollment times, 66 came by ambulance and 203 had other exclusion criteria, leaving 234 eligible subjects approached. Seven of these patients refused to be interviewed and 36 others fell asleep or could not otherwise participate meaningfully in the interviews, resulting in a total of 191 homeless subjects interviewed. Of 110 control (not homeless) patients initially screened, 18 arrived by ambulance and 20 had other exclusion criteria, leaving 72 not homeless patients initially approached. Three of these refused to be interviewed and 6 others fell asleep or could not otherwise participate meaningfully in the interviews, resulting in 63 control subjects concurrently enrolled. See Table 1 for a summary of subject characteristics. Similar percentages (79% and 79%) of homeless and control subjects believed that their presenting problems were emergencies, but fewer (40% versus 57%) homeless persons believed that they could get their problem cared for in a clinic (⌬⫽17%; 95% CI 3% to 32%). See Table 2 for a summary of subject responses to interview questions. Homeless persons reported eating fewer mean meals per day (standard deviation [SD] 2.1 [1.1] versus 3.6 [SD 5.9]) and cited hunger as a greater problem in their daily lives, 4.8 (SD 600 Annals of Emergency Medicine
Our results are subject to the limitations of other interviewbased studies, most notably social desirability bias and a failure of subjects to respond to all of the questions. Given that patients may be reluctant to admit that they present for food, shelter, or safety instead of for true medical care, our findings may underestimate the effect of subsistence needs on ED presentation. We did not elucidate the reasons for incomplete survey responses. Although we established test-retest validity on a pilot group of patients, other measures of validity were not confirmed. Our exclusion of patients with true medical need (trauma, intoxicated, critically ill, and transported by ambulance) artificially inflates the percentage of homeless persons who we determined presented for subsistence needs. Assuming that standard triage would leave less severely ill homeless persons in the waiting room, however, this inflation may be countered to a certain degree by our restriction of interviews to homeless patients within the ED care area. Given that this was a singlecenter study in a city with a large homeless population, our findings may not apply to EDs in other cities. The most important limitations of our study are the inherent difficulties in defining the true reasons patients came to the ED and in making the distinction between true medical need and true social/welfare need. Many patients likely presented for both needs, and in this population hunger, shelter, and safety may be Volume , . : May
Rodriguez et al
Nonmedical Needs Motivating Visits of the Homeless
Table 2. Summary of subject responses to interview questions. Yes Responses Question Hunger Mean No. meals/day (SD) Currently receive food stamps (%) How much of a problem is hunger in your daily life?* (SD) Last 3 mo, eaten less than you wanted because of no money (%) Is hunger a main reason you came to ED today? (%) Shelter/clothing Mean No. nights/week on street (SD) How much of a problem is lack of shelter in your life?* (SD) Do you have adequate clothing? (%) Lack of shelter a main reason came to ED today? (%) Safety Have you ever been assaulted or attacked on street? (%) Mean No. of times, if responded yes (SD) How much of a problem for you is safety on streets?* (SD) Fear for safety a main reason came to ED today? Do you believe your problem is an emergency today? (%) Could you go to a clinic instead for this problem? (%)
Homeless, nⴝ191
Not Homeless, nⴝ63
2.1 (1.1) 39/189 (21) 4.8 (3.7) 119/189 (63) 25/189 (13)
3.6 (5.9) 10 (16) 1.6 (2.9) 5 (8) 3 (5)
3.5 (3.0) 6.1 (4.2) 57/185 (31) 32/185 (17)
N/A 3.5 (4.0) N/A N/A
92/182 (51) 10.4 (53) 5.1 (4.0) 27/182 (15%) 142/179 (79) 68/169 (40)
0 0 1.8 (3.2) 4 (6%) 50 (79) 35/61 (57)
*Analog scale 0-10: 0⫽no problem, 10⫽worst possible problem.
intrinsically tied to and perceived as medical conditions. Any measure targeting subsistence presentations to the ED must consider these links between social and medical need.
DISCUSSION Although the ED’s position as the health care safety net is well established and understood by medical and lay communities, the fact that EDs also commonly serve as an urban welfare safety net may be less obvious to those outside of ED staff.6,7 In this study, we confirmed what many ED health care providers have assumed—that nonmedical subsistence needs are a common reason for homeless patients to present to the ED. Almost a third of homeless persons walking into our ED cited hunger, safety, and lack of shelter as primary reasons for their presentation. In this era of severe ED crowding and closures, any extra burden imposed by nonemergency care needs of the homeless and others may further jeopardize not only the ED’s vital health care safety net function but also care for critically ill patients.13,14 To redirect homeless persons seeking health care from the ED to clinics and other sites of care, O’Toole et al15 proposed a renewed focus on outreach, education, and increased medical insurance coverage. Other investigators have reported decreases in ED utilization by homeless persons after implementation of intensive aid and primary care programs.2,16 In a study examining the effect of the federally funded Health Care for the Homeless Program, Han and Wells2 reported that having 2 or more Health Care for the Homeless Program visits was associated with fewer inappropriate ED visits. Similarly, in a Canadian ED trial, Redelmeier et al16 reported a one-third reduction of return ED visits in homeless persons randomized to a program of satisfaction directed “compassionate care.” Volume , . : May
Kushel et al,11 however, have proposed that to decrease homeless ED utilization, it may not be enough to merely provide health insurance and alternative ambulatory care settings. Our findings support this more comprehensive view of the homeless individual–ED problem, specifically, that any program directed at decreasing homeless person ED visits must also address subsistence needs presentations. An obvious primary measure to reduce subsistence needs presentations is the establishment of alternative sites where homeless persons can receive basic necessities at all times. The first step in referral of subsistence needs patients to these alternative sites away from the ED would be identifying and distinguishing them from other true medical needs patients. Incorporation of standard questions about hunger, shelter, and safety at ED triage could allow patients to express these needs in a nonjudgmental setting, without the requirement of presenting a medical chief complaint. Patients with primary subsistence problems could thereby opt out of ED care and receive these necessities outside of ED treatment areas or by transport to an always open shelter. In our study, 24% of subjects reported that they would have utilized such a center instead of the ED if it were always available. Provision of food, shelter, and safety without utilization of ED beds for even a fraction of homeless person presentations would undoubtedly save time and resources. Gordon7 outlined several models of such “social triage centers,” in which EDs could provide a variety of screening, referral, and care management options. The author implemented a pilot program of one of these models, which identified many patients with substantial social deprivations and was well received by patients and health care workers.17 Although some may posit that these alternative basic needs sites already exist in the form of homeless shelters, shelter Annals of Emergency Medicine 601
Nonmedical Needs Motivating Visits of the Homeless
Rodriguez et al
systems are overburdened, with most closing to new occupants late at night. San Francisco currently has 9 homeless shelters with approximately 1,200 beds, and an annually conducted homeless person count revealed a conservative estimate of 6,377 homeless persons in San Francisco in 2007.18 In an assessment of shelter referrals conducted from mid-April until mid-May 2008 by the Social Work Department at San Francisco General Hospital, only 15 of 53 (28%) of the homeless persons in the ED seeking shelter referral were successfully placed, primarily because of lack of beds. The ED serves as a welfare safety net for homeless persons who commonly present because of hunger, lack of shelter, and safety concerns. Provision of these subsistence needs at all times at another site may decrease their ED presentations. Supervising editor: David L. Schriger, MD, MPH Author contributions: RMR conceived and oversaw the study. RMR and JF analyzed the data. JF, CC, VN, and DP interviewed patients. JF finalized the data. CC, VN, and DP assisted with the background research for the project. RMR takes responsibility for the paper as a whole. Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article that might create any potential conflict of interest. The authors have stated that no such relationships exist. See the Manuscript Submission Agreement in this issue for examples of specific conflicts covered by this statement. Publication dates: Received for publication March 7, 2008. Revisions received June 6, 2008, and July 16, 2008. Accepted for publication July 23, 2008. Available online October 5, 2008. Reprints not available from the authors. Address for correspondence: Robert M. Rodriguez, MD, Department of Emergency Services, San Francisco General Hospital, 1001 Potrero Avenue, San Francisco 94110; E-mail [email protected].
REFERENCES 1. Kushel MB, Vittinghoff E, Haas JS. Factors associated with the health care utilization of homeless persons. JAMA. 2001;285: 200-206.
2. Han B, Wells BL. Inappropriate emergency department visits and use of the Health Care for the Homeless Program services by homeless adults in the northeastern United States. J Public Health Manag Pract. 2003;9:530-537. 3. Little GF, Watson DP. The homeless in the emergency department: a patient profile. J Accid Emerg Med. 1996;13:415417. 4. Richardson LD, Hwang U. America’s health care safety net: intact or unraveling? Acad Emerg Med. 2001;11:1056-1063. 5. Ong Eng Hock M, Ornato JP, Cosby C, et al. Should the emergency department be society’s health safety net? J Public Health Policy. 2005;3:269-281. 6. Adams JG. Confronting hunger in the emergency department. Acad Emerg Med. 1999;6:1082-1084. 7. Gordon JA. The hospital emergency department as a social welfare institution. Ann Emerg Med. 1999;33:321-325. 8. Biros MH, Hoffman PL, Resch K. The prevalence and perceived health consequences of hunger in emergency department patient populations. Acad Emerg Med. 2005;12:310-317. 9. Kersey MA, Beran MS, McGovern PG, et al. The prevalence and effects of hunger in an emergency department patient population. Acad Emerg Med. 1999;6:1109-1114. 10. D’Amore J, Hung O, Chiang W, et al. The epidemiology of the homeless population and its impact on an urban emergency department. Acad Emerg Med. 2001;8:1051-1055. 11. Kushel MB, Perry S, Bangsberg D, et al. Emergency department use among the homeless and marginally housed: results from a community-based study. Am J Public Health. 2002;92:778-784. 12. National Weather Service Forecast Office. San Francisco Bay area/Monterey. Available at: http://www.weather.gov/climate/ index.php?wfo⫽mtr. Accessed May 13, 2008. 13. Cowan RM, Trzeciak S. Clinical review: emergency department overcrowding and the potential impact on the critically ill. Crit Care. 2005;3:291-295. 14. Trzeciak S, Rivers EP. Emergency department overcrowding in the United States: an emerging threat to patient safety and public health. J Emerg Med. 2003;5:402-405. 15. O’Toole TP, Gibbon JL, Hanusa BH, et al. Preferences for sites of care among urban homeless and housed poor adults. J Gen Intern Med. 1999;14:599-605. 16. Redelmeier DA, Molin JP, Tibshirani RJ. A randomized trial of compassionate care for the homeless in an emergency department. Lancet. 1995;345:1131-1134. 17. Gordon JA, Chudnofsky CR, Hayward RA. Where health and welfare meet: social deprivation among patients in the emergency department. J Urban Health. 2001;78:104-111. 18. Swarns RL. Chronic homeless count down 30%, government says reduction from ’05 to ’07 may reflect changes in reporting. San Francisco Chronicle. July 30, 2008;A-2. Available at: http://www.sfgate.com/cgibin/article.cgi?f⫽/c/a/2008/07/30/ MNAF121EL3.DTL&hw⫽homeless&sn⫽001&sc⫽1000. Accessed July 31, 2008.
Did you know? You can personalize the new Annals of Emergency Medicine Web site to meet your individual needs.
Visit www.annemergmed.com today to see what else is new online!
602 Annals of Emergency Medicine
Volume , . : May
APPENDIX E1. ED data extraction form-study patients. Questions: How long have you been homeless/without permanent shelter?______ I. Hunger: A. On a typical day, how many meals do you eat? B. Where do you get your food? C. Do you receive food stamps? □Yes □No D. On a scale of 0 to 10 (where 0 ⫽ no problem at all and 10 ⫽ worst possible problem), how much of a problem is hunger for you in your daily life? ____/10 E. In the last 3 months, have you ever eaten less than you felt you should have because there wasn’t enough money for food? □Yes □No F. Was lack of food/hunger one of the main reasons you came to the ED today? □Yes □No ● Have you ever come to the ED for food/hunger reasons in the past? □Yes___# □No G. If you could go to a place that would provide you with free food at all hours, would you have come to the ED today? □Yes □No II. Shelter: A. In a typical week, how many nights do you sleep/spend on the street without shelter? _______ B. If you have shelter some nights, what type of place is it (a house, apartment, group shelter)? ___________________ C. On a scale of 0 to 10 (where 0 ⫽ no problem at all and 10 ⫽ worst possible problem), how much of a problem is lack of shelter for you in your daily life? ___/10 D. Do you feel you have adequate clothing when you are without shelter? □Yes □No
Volume , . : May
E. Was lack of a place to sleep/get out of the cold one of the main reasons you came to the ED today? □Yes □No ● Have you ever come to the ED for shelter in the past? □Yes___# □No F. If you could go to a place that would provide you with free shelter at all hours, would you have come to the ED today? □Yes □No III. Primary care: A. Where do you usually get your healthcare? Emergency departments Private clinics Free clinics None B. Do you believe your problem today is an emergency? □Yes □No C. If you could go to a clinic at any hour for your current problem, would you have come to the ED today? □Yes □No IV. Safety: A. In your time as a homeless person, have you ever been assaulted or attacked? □Yes □No B. On a scale of 0 to 10 (where 0 ⫽ no problem at all and 10 ⫽ worst possible problem), how much of a problem is personal safety on the streets for you in your daily life? ___/10 C. Was fear for your safety one of the main reasons you came to the ED today? □Yes □No ● Have you ever come to the ED for safety reasons in the past? □Yes___# □No D. If you could go to a place that would provide you with safety at all hours, would you have come to the ED today? □Yes □No V. If you could go to a place that would provide you food, shelter and safety at all hours would you have come to the ED today? □Yes □No Is there anything else you would like to tell me?
Annals of Emergency Medicine 602.e1