Managing Indigent Care: A Case Study of a Safety-Net Emergency Department

Abstract

Objective

To examine how one safety-net emergency department (ED) managed problems associated with the provision of indigent care in everyday life.

Data Sources/Study Setting

Interview and observational data collected in County Hospital ED, a public teaching hospital in a California city, during 6 months of 1999.

Study Design

The study used ethnographic methods to document and understand day-to-day routines and practices for providing indigent care in a safety-net facility.

Data Collection/Extraction Methods

One- to 2-hour semistructured interviews with a snowball sample of eight ED physicians were tape recorded, and fieldnotes were recorded in situ during 10–30 hours of participant observation per week in all areas of the ED. Data were coded to highlight themes of interest and to identify recurrent patterns of behavior.

Principal Findings

In everyday life, providers at County ED relied on graduate medical education (GME) to manage two everyday problems, social use and tenuous financing, associated with the provision of indigent care. GME helped manage problematic social visits to the ED by defining them as interesting cases. GME helped with tenuous finances by creating a work culture that encouraged the provision of uncompensated work.

Conclusions

Safety-net facilities often face problems similar to those in County ED. Future research should assess the extent to which the everyday management of these problems in County ED resembles that in other safety-net facilities.

Emergency departments (EDs) of safety-net hospitals that treat a low-income population face two problems on a daily basis: social use and tenuous financing. Problematic social use arises because EDs are required to see “all people at all times regardless of the problem” (Gordon 1999, p.321), even if this means assessing and treating patients for social problems that typically fall outside of the purview of the acute-care hospital (Malone 1995; Pope et al. 2000). Tenuous financing is a problem because hospital services are often inadequately reimbursed by patients who have Medicaid or are uninsured (Rask et al. 1994). Although EDs have faced these problems for decades (Roth and Douglas 1983), recent market-oriented changes in social welfare provision and in health care financing have exacerbated the problems of social use (Gordon 1999; Malone 1998) and tenuous financing (Cunningham et al. 1999; Grogan and Gusmano 1999; Mann et al. 1995).

A considerable amount of research exists, including the studies cited previously here, on the structural, demographic, and political causes of inappropriate social use and tenuous financing. However, previous research has not examined how these two problems are managed in everyday life inside of the safety-net ED. To provide some initial insights into how inappropriate use and tenuous financing manifest themselves and are managed, a case study of one safety-net ED was conducted using ethnographic research methods.

Like many safety-net facilities (Fishman 1997), the studied ED was affiliated with a teaching hospital, and indigent care was usually provided in conjunction with graduate medical education (GME). In this article, the relationship between GME and indigent care in the ED is examined. While acknowledging the substantial role of GME in financing indigent care, the focus here is on how GME shaped day-to-day work inside the safety-net ED. Itisargued that GME shaped the social and cultural mechanisms by which EDstaffers managed the problems of social use and tenuous financing in everyday life.

Methods

Setting

Research was conducted at County Hospital (a pseudonym, as are all proper names in this article), a publicly funded teaching institution that has historically served the low-income population in the California city where it is located. These characteristics of County Hospital are typical for health care facilities used by poor people in the urban United States (Baxter and Mechanic 1997). Within County Hospital, the ED is located in approximately one dozen rooms, each with one to four treatment areas, opening onto a hallway. Outside the entrance to the ED is a small lobby that can hold additional patients and beyond this lobby are cubicles for primary triage, small offices for secondary triage and intake, an urgent care clinic, and the ED waiting area. The waiting area can accommodate approximately 100 people, and it serves the ED as well as other hospital wards.

According to hospital administrators, the most recent yearly ED census was approximately 70,000, and at least three quarters of the patients who came to County ED for care were indigent. County ED staff did not carefully distinguish medical indigence from socioeconomic indigence. Their practical understanding of indigent care was that all people who came to County ED were seen and treated regardless of ability to pay. Some of the indigent were working but uninsured. Others were destitute or homeless. Some were underinsured via Medicaid. Some were frail seniors who lacked the resources to meet their health care needs outside of the hospital, and some were prisoners and criminal suspects. County Hospital is a trauma center, and thus, it also treated some adequately insured trauma victims.

Data Collection Procedures

Fieldwork at County Hospital was conducted during 6 consecutive months in 1999, beginning with formal interviews of senior physicians. A snowball sampling technique—soliciting referrals during informal conversations and at the close of formal interviews—was used to identify and arrange interviews with physicians with administrative responsibility, physicians who worked many shifts, and physicians who had a long tenure at County ED. In all, interviews were conducted with eight physicians, nearly half of the ED’s active attending staff, including all physicians with significant administrative responsibilities and many of the physicians with the heaviest workloads during the months that the fieldwork was conducted. Interviews lasted 60 to 120 minutes and were tape recorded. The interview was semistructured and open ended. It addressed respondents’ views of how and why people came to the ED at County Hospital, how physicians interacted with patients inside the ED, respondents’ personal and professional backgrounds, their professional history at County ED, and their views of the rewards and drawbacks of being an emergency doctor at County Hospital. All of the interviews took place at County Hospital.

During 5 months at County ED, 2 to 4 days (10–30 hours) per week were spent conducting participant observation and recording fieldnotes using standard procedures (Emerson, Fretz, and Shaw 1995). Time was spent in nearly all parts of the ED, including waiting areas, hallways, treatment areas, and staff-only residents’ and nurses’ rooms. Activities were observed during all three shifts and on every day of the week. As in EDs in general (Nourjah 1999, p. 11), patient volume at County ED was higher in the evenings and on weekends, and thus, more time was spent at the hospital at those times. Interest in GME led to spending more time in places where interactions between attending physicians and student doctors commonly took place, such as the hallways, trauma rooms, and residents’ room of the ED. In order to compare observational with interview data, the author spent a shift observing the routine activities of the physician in the days preceding a scheduled interview.

Participant observation involved several kinds of interaction at County ED. Informal conversations were held with clients (including patients, family, and friends), physicians, residents, nurses, and other staff. At times, our conversations had nothing whatsoever to do with areas of research interest, but at other times, the conversation included discussion of issues of interest, such as their experiences as a client or as a worker in County ED or in health care settings. As a participant, menial assistance was provided with the provision of patient care by, for example, handing a package of gauze to a staff member or relaying a patient’s message to a family member. As both observer and participant, the researcher sought to disturb the social setting as little as possible. After inviting staffers’ and clients’ participation in the research project (on approximately six occasions, staffers or clients refused invitation and asked to not have observation of their activities in the ED), activities were observed from an out-of-the-way location.

There was always more going on in County ED than could be recorded in fieldnotes. To focus participant observation activities, each fieldwork day began by defining the goals of the day’s observations with respect to the project’s central questions. When leaving the field, a summary of the day’s events, which focused on the issues in the fieldnote write-up, was included. Without excessively narrowing purview, these daily plans and summaries focused research efforts on the issues that motivated the project.

Data Analysis

Data analysis began during fieldwork in a manner consistent with that used in grounded theory (Strauss 1987). Transcribed interviews and fieldnotes were organized in a Folio Views database for computer-assisted qualitative data analysis according to format (interview transcript, participant observation fieldnote, daily fieldwork plan or summary, or analytical memo) and the date collected. Database passages related to GME and the provision of indigent care were coded during data entry and by reading through and searching the completed database. Sample codes include attitudes toward the indigent and student doctors, incidents of explicit didactic instruction, and expressions of opinion about the medical training program at County Hospital. After completing fieldwork, supplemental notes were added to the database about the background of research subjects, connections between items in the database that were distinct in format or chronology, and preliminary analyses and interpretations. The final analysis step was to write analytical memos that integrated raw fieldnotes, fieldnote codes, and supplemental notes. These memos condense large amounts of participant observation and interview data into coherent interpretations of repeatedly observed patterns of interaction.

Results

The results presented here are repeatedly observed interactions, as condensed and interpreted in analytical memos, and are presented in three parts. First, how the problems of social use and tenuous financing manifested themselves in everyday life at County ED is described. Second, how GME helped manage the problem of social use is discussed. The third item discussed is how the organizational culture surrounding GME helped address the problem of tenuous financing.

Social Use and Tenuous Financing at County ED

Social use of County ED was self-evident in everyday life. Patients regularly came to the ED seeking relief from social problems, and police and emergency medical technicians (EMTs) brought clients with social problems to County ED for evaluation and treatment. Tenuous financing was manifest in the patient population at County. The ED was usually overcrowded with patients who were obviously poor. Wealthier patients rarely came to County ED, and those who did usually left quickly.

Clients seeking relief from social problems were highly visible in the ED waiting area. All people who presented at County Hospital were triaged by EDstaff, and patients with serious medical conditions quickly entered the ED proper. However, patients whose primary complaint stemmed from social dislocations usually faced a long wait. The situation one Friday night shortly after midnight was typical. The waiting area contained approximately 15 people who had minor medical complaints. Most appeared to be homeless, which is a serious social problem, and were sleeping. By 9 a.m. the next morning, only one person, a middle-aged man who had complained several times about the long wait to have a doctor look at a rash that kept him awake, had been seen in the ED. For the others in the waiting area at midnight, ED triage had justified their spending the night in the waiting area and thus addressed their social needs. After several hours of indoor sleep, they left in the morning without entering the ED proper.

Social use was not confined to the waiting area of County ED. In treatment areas, staffers regularly confronted socially problematic patients who were involuntarily brought to County ED by police and EMTs. Police brought prisoners and criminal suspects, as well as all people deemed a danger to themselves or others, to County because it was a public hospital. EMTs often chose to transport people with cognitive or emotional problems and people who had overdosed on alcohol or drugs to County rather than to private hospitals nearby. As one doctor explained, “A patient, dead drunk, completely passed out will come in here and on the run sheet [a form filled out by the EMT], it will say ‘requested to go to County.’ I mean that’s ridiculous, the person is passed out dead drunk. They didn’t say anything, but EMT does that because they know if they go to Burgess or Park [private hospitals nearby], they’ll just get sent over here anyway.” Patients who were involuntarily transported to County ED were often physically restrained on gurneys, and their gurneys were often left in the hallway in order to reserve treatment areas for patients with more serious medical problems. Thus, the hallway of County ED was often filled with patients with substantial social problems who confronted staff and other patients as they passed by in the hallway.

Gurneys in the hallway were also one of the many visible manifestations of County ED’s tenuous financial situation. Low-income people referred themselves specifically to County ED because they could not afford medical care provided in nearby private facilities (such a case is discussed later here). Transfers of low-income patients from nearby private health care facilities were also common. One Wednesday afternoon, for example, a note arrived saying that Burgess hospital, a nearby private nonprofit, wanted to transfer a patient. The note provided no further details, and the attending physician at County was surprised because in his experience, the Burgess doctor who requested the transfer usually did not “dump” poor patients on County. He interpreted the transfer request cynically. As he waited to speak to the Burgess ED physician, he remarked, “They’ll take a look at him, and then do a wallet biopsy. If he doesn’t have the money, they’ll send him over here because we take anyone. But there’s no medical reason. We’re not a higher level [care] facility unless it’s a trauma.” After speaking with the Burgess ED physician, however, the County attending characterized the transfer differently. “This man is HIV positive and he’s a working man, so they tell him [to] come over here and get his care. If he stays at Burgess Hospital, he’ll run up $20,000 of care and he’ll never get out from under the debt. So they tell him he can get a transfer to County, and he does it. Here, we’ll take him and treat him, and if he can’t pay, [we will] set up some kind of financing.”

Tenuous financing was a part of everyday life at County not only because poor people occupied the ED but also because wealthier people seldom referred themselves to County and rarely stayed at County when transported by others. The only regular source of wealthier patients at the ED was the trauma service. Higher income people brought to County after a trauma such as a motor vehicle accident were, as one physician said, “People who wouldn’t come here in a million years, but are here by what are called the X-factor. Something happened and here they are. They usually leave as quickly as they can.” One higher income person was brought to County ED following a car accident, and he remained in the ED for less than 1 hour. Physicians determined that he had no life-threatening injuries, fit him with a neck brace, and called his relatives; shortly after arriving at County ED, his family secured the man’s transfer by ambulance to nearby Burgess Hospital where he was to receive further treatment.

GME, Teaching Opportunities, and Social Use

GME helped manage the problem of social use at County ED by allowing staff to recast social visits to the ED as teaching opportunities. GME thus allowed ED staff to respond legitimately to the social needs of patients at the same time thatit provided an excuse to treat social visits, which were often interesting medical cases.

Staff at County ED often encountered patients who came to County Hospital in search of social support or in response to social problems. The social needs of these patients could not be addressed within the context of the medical mission of the ED, but the presence of the GME program at County provided some flexibility. For example, procedures that might not be justified on strictly biomedical grounds could be justified as an opportunity for graduate medical instruction. The case of Mr. Jones illustrated this dynamic. Mr. Jones was a frequent user of County ED who arrived one cold and rainy night seeking admission to the hospital. He had a variety of social problems, but no medical condition warranted his admission. Staffers were concerned that Mr. Jones would react poorly to being refused admission. GME helped defuse this problematic and potentially explosive instance of social use.

A senior resident presented the particulars of Mr. Jones’s case to an attending physician. “Mr. Jones has a long history at County,” the resident said. “He is an African American man, in his 40s, who is a drug and alcohol user and often comes in for treatment. His visits have become more frequent since the New Year, when he became homeless. In addition, Mr. Jones can be difficult to manage. He has gotten aggressive and hostile in the past. When he arrived at the ED last night or the day before, I saw him, and he was totally compliant until we told him we were sending him home and not going to be admitting him. At that point, he became very aggressive, and I had to get security to help me move him out of the ED. Things might be a bit better now because the last time he was here, Mr. Jones saw social services, and they managed to find him a place in an SRO hotel. Mr. Jones has chronic circulation problems, especially in his legs. He returned tonight because of swelling. I don’t know what to do with him. I don’t think there is anything too serious going on, but he keeps saying he doesn’t want to leave.”

GME provided an artful way to address Mr. Jones’s social problem—lack of shelter—without admitting him to the hospital. The attending physician who heard the case presentation asked a medical student to assess Mr. Jones’s condition using an ultrasound diagnostic procedure. The procedure provided a reason for Mr. Jones to remain in the ED for many additional hours. It provided an opportunity for the attending physician to teach the medical student the use of the ultrasound equipment, and it provided an opportunity for the student to practice his technique. None of the medical staff were surprised to find that the results of the test were negative. However, after the procedure, Mr. Jones left the ED voluntarily, which was seen as a positive outcome to the case.

Attending physicians at County ED did not experience social visits by poor people, such as the visit of Mr. Jones, solely as a problem to be overcome. Many physicians said that they valued treating low-income patients whose social problems and limited access to health care made them interesting cases. Among the poor, as one physician at County ED said in an interview, “You get people who not only have had pneumonia, but have had pneumonia for 6 days before they seek care, so you are basically seeing the natural course of disease untreated as it progresses.” The “natural course of disease untreated” was intellectually interesting and exciting, as the physician explained, because “the medical problems associated with the indigent are much more complex and advanced” than among wealthier people. When practicing in the ED, physicians were often excited to see indigent patients with social problems during their shift. Staffers took pictures of the gangrenous hand of a mentally ill intravenous drug abuser, covered a whiteboard with differential diagnoses of a woman whose autoimmune disease had progressed untreated for months, and gathered residents to see the open chest of a man with multiple gunshot wounds. Cases like these were exciting, even though physicians recognized that these patient visits were caused by social circumstances—physicians saw poverty, lack of education, lack of access to health care, social isolation, drug abuse, and violence as social causes of ED visits—as much as by biomedical problems.

Medical residents also valued the social visits of a low-income patient population. EDs with a low-income client base, such as County, had a greater number of more interesting patients than slower paced EDs that provided less indigent care. The comments of one medical student, who was visiting several EDs before applying to residency, captured how student doctors viewed County ED. The student compared his experiences at County ED with the time he spent at a private hospital in Utah: “I did a rotation up in Salt Lake City, and we didn’t see anything up there. Here, I figure having a rotation here will help me get into a good residency. I’d love to come here.” Patients with “disease untreated” could be found throughout County Hospital, but low-income patients were particularly valuable for ED training. This was reflected in the relative popularity of the ED residency program at County compared with other programs. One administrator familiar with GME at County Hospital said that the ED typically attracted more applicants from good medical schools than it could accommodate. Internal medicine, on the other hand, often had to rely on what the administrator termed “lower quality” international medical graduates.

Social use taught student doctors to recognize and disentangle the social and medical reasons for ED visits—a crucial part of patient assessment, diagnosis, and treatment. For example, an inexperienced resident named Paul and an experienced resident named Ron both saw patients with medical and social problems within a few hours of each other one night at County ED. An attending physician, Jerry, consulted with the residents on both cases. The similarities and differences between the cases illustrate the importance of recognizing the social as well as the medical natures of ED visits.

At approximately 8 p.m., Paul, who spoke Spanish, was asked to interview a 35-year-old Spanish-speaking man with a history of cocaine use who was suffering from chest pains. In his initial interview, Paul asked the patient how he was feeling and performed an examination. While observing the interview, the smell of alcohol was noticeable; however, this was not unusual at County ED, and Paul appeared not to notice it. Later, Paul saw Jerry in the hallway and recommended that an EKG be performed on the patient. The two doctors returned to the treatment area to examine the patient together. Jerry immediately commented on the smell of alcohol. He explained, for the researcher’s benefit, that cocaine generally is flushed through the system quickly, but when it is combined with alcohol, it can metabolize into a form that remains for a long period of time and can cause problems. Jerry instructed Paul to ask the patient about his alcohol use. When asked, the patient replied that he had not been drinking, and Paul appeared inclined to believe him. However, Jerry remained skeptical, and he instructed Paul to ask the patient specifically whether “he maybe had a few beers earlier in the day.” Prodded, the patient said that he had had some beers. Paul’s recommended EKG was deferred pending the results of a blood alcohol test.

At approximately midnight that same evening, Ron examined a 60-year-old White male who also had chest pains. Two EKGs had been performed, one earlier that day at a local clinic and a second that evening at County ED. The following fieldnotes relate a conversation about the case between Ron and Jerry.It shows that Ron, in contrast to Paul, had been aware of, and had inquired into, the social circumstances surrounding the patient’s medical condition.

Jerry (looking over the EKGs with Ron): “They look like they come from two different patients.” Ron (pointing to one part of the EKG readout): “There are some similarities here on this part. I don’t really know what is going on. I’m thinking about getting another EKG and comparing it to these two to see if there is anything going on over time.”

Jerry: “Where’s his pain?”

Ron: “He says it is right here (indicates the upper part of his chest on both sides of his body). It hurts him when I push on it, but I don’t know. He takes a good amount of pressure.”

Jerry:: “Does he have a home? Is he looking for a place to stay?”

Ron: “No, he has a home; he lives in a trailer.”

Paul and Ron’s handling of patients with similar medical complaints shows how experience with the social dimensions of disease led to improved medical assessment and treatment. Both patients had similar medical symptoms (chest pain), but proper interpretation of those symptoms required information about the patient’s social condition. Ron, with greater experience in County ED, paid attention to an obscure and seemingly irrelevant feature of his patient’s social condition: that he lived in a trailer and was not just “looking for a place to stay.” Thus, he correctly focused on the man’s puzzling EKG. Paul, with less experience at County, focused prematurely on the patient’s EKG. He did not recognize the relevance of the smell of alcohol in the treatment area, nor did he press the patient when the man denied having had anything to drink. As Jerry’s interactions with the residents highlight, proper diagnosis and treatment in both cases required attention to the patient’s social situation as well as his medical condition. These examples illustrate the final reason that inappropriate social visits were valued in the context of GME: They provided opportunities for student doctors to practice recognizing and disentangling social from medical problems.

GME, Organizational Culture, and Tenuous Financing

GME addressed the problem of tenuous financing by sustaining a work culture that encouraged the provision of uncompensated work. The organizational culture of County ED revolved around medical training. Within this cultural context, working overtime without pay, undertaking specific tasks that fell outside of one’s formal job definition, and other forms of uncompensated work were understood to be both appropriate and necessary. Uncompensated work occurred in three ways: First, attending staff required interns and residents to provide uncompensated work. Second, both student doctors and attendings often provided uncompensated work absent of coercion. Third, ED administrators actively sought to maintain and reproduce this feature of the organizational culture of the ED. All three of these forms of uncompensated work were part of the everyday work culture surrounding GME at County ED, and all three provided a means for addressing the problem of tenuous financing.

As part of GME, attending physicians regularly required student doctors to provide uncompensated work. Most commonly, attending physicians told residents and interns to work beyond the end of a scheduled shift. Residents were not paid hourly wages, and thus, staying beyond the end of a shift constituted unpaid work. Residents often found themselves completing several required tasks after their shift ended. For example, one spring evening, a resident presented his final case to an attending physician at 7 p.m., after already having worked 2 hours of uncompensated overtime. The resident’s final patient was a young girl with a mouth injury, and the attending agreed that antibiotics were indicated. As the attending signed off on the young girl’s chart, the resident began gathering his personal possessions to leave the ED. The attending asked, “Were they [the antibiotics] administered?” The resident replied that he had ordered the medications. “Not good enough,” said the attending. “You have to make sure they actually got to her.” The resident, clearly eager to leave the ED, pointed out the medications he had ordered. “No. That’s not enough,” replied the attending, her voice rising. “You have to make sure she got them.” Exasperated, the resident relented and remained in the ED an additional 30 minutes to administer the antibiotics.

Often, physicians at County ED provided uncompensated work voluntarily. Although they often found it frustrating, physicians saw uncompensated work as an unavoidable part of everyday life for two reasons: First, they believed that the patient population at County ED required a great deal of personal attention, and second, experience taught physicians that they could not rely on support from other workers in the ED. The comments of one attending physician, made toward the end of one busy overnight shift, illustrate both of these beliefs.

I stepped outside with Dr. Roberts, and he lit a cigarette and began to say how he had been paid more money for less work at Park Hospital and at suburban hospitals nearby. “The reality is,” he said, “that I work harder at County during slow shifts than I did during busy shifts other places. You never get a night like tonight at other places. And you make more money at other places.” I asked him what accounted for that. Some of the extra work at County, he said, was because the patients are sicker and poorer. Dealing with poor, sick, and uneducated patients always requires more work. But a lot of the excess work is because the staff at County hospital is no good. Docs in the ED have to do everything themselves. You have to double check to be sure all the tests get done, to make sure all the meds get administered, do all the paperwork to get someone admitted, even to call someone’s family to come pick them up. “That’s why so many burn out,” he concluded. I suggested that the way to avoid burnout was to avoid doing all the excess work—just work your shift and go home. Dr. Roberts disagreed. “No, that doesn’t work. Someone has to do it. The way to avoid burnout is to keep perspective. You have to have a sense of humor. Ride with it. On busy nights like tonight, I say, you just have to surf the chaos. There’s no way to take it easy. But if you take it too seriously all the time, you go crazy and burn out.”

Senior staff at County ED nurtured a work culture that encouraged uncompensated work in their selection of new residents and their recruitment of new attending faculty. In terms of residents, County ED always had more applicants than its program could accommodate. Administrators wanted residents to have impressive credentials, but they also wanted residents who would complement the work culture at County ED. In an interview, an administrator said that he and his colleagues were particularly interested in “the volunteer experiences that people put on their applications because that tells us a little bit about what their real interests are.” They were particularly attracted to applicants who had “worked in third-world countries to help out in some point of their career” because those applicants were more likely to understand and support the work that goes into providing indigent care at County ED. Someone “who grew up in the projects in Los Angeles and has a great success story and made it through college and medical school and now wants to do something where he can go back and work in indigent care environment” would also be a good candidate. Residents who lacked this preferred orientation toward indigent care could have a difficult time at County ED. During one shift, for example, an attending complained to a colleague about one intern who she thought had become overly frustrated with alcoholic patients. The attending said that if the intern was not able to be more patient and understanding, it would perhaps be better for the intern to go to a different hospital for the remainder of training. As the comments of the administrator and attending show, in both selection and evaluation of residents, County ED staff desired residents who understood and supported a work culture in which the provision of uncompensated work in the name of indigent care was valued as an end unto itself.

Attending staff at County ED also looked for a demonstrated willingness to provide uncompensated work when searching for new faculty. They expected that prospective attendings at County ED understood that uncompensated work was common. Comparing County to a nearby teaching hospital that was less well known for providing indigent care, the administrator quoted previously said, “Rarely do we get people looking at County Hospital to work who are also considering University Hospital. The patient populations are so different. The jobs are entirely different.” An attending faculty member recently recruited to County ED had previous work experience in public hospital EDs in other states. This new physician appreciated, as one veteran said during a break from a busy shift, that it was possible to make more money working elsewhere but there were nonmonetary rewards of working at County ED. The work was never dull because the ED was busy and the patient population was interesting, and working at County gave physicians a sense of accomplishing something worthwhile.

Conclusion

In this case study ED, everyday manifestations of inappropriate social use and tenuous financing were largely managed through GME. As part of GME, physicians regularly recast inappropriate social visits as valuable teaching opportunities—thus managing the problem of inappropriate use by taking advantage of the fact that patient visits are also the source of “teaching material” for GME (Stevens 1989, p. 10). The presence of the GME program helped with the problem of tenuous financing by encouraging physicians to provide uncompensated work—thus shaping work routines and expectations to provide a cultural solution to the problem of inadequate resources. This cultural mechanism complemented GME’s structural role financing academic health care centers that provide indigent care (Fishman and Bentley 1997).

The case study of indigent care at County ED illustrates three points. First, physicians at County ED faced day-to-day difficulties in providing health care to the poor because of social use and tenuous financing. Second, the GME training program at County Hospital helped ED physicians manage social use by redefining problematic social visits as valuable opportunities for graduate medical training. Third, GME helped address the problem of tenuous financing by sustaining a work culture that encouraged the provision of uncompensated work. The implications of these findings extend in three directions.

First, safety-net EDs are often called on to provide services to the medically needy, but researchers have noted that EDs are increasingly being looked to for a broader range of services by poor people in general (Gordon 1999). Increased service demand at EDs may reflect the fact that safety-net EDs are still better able to provide nonmedical social services to the poor than dedicated social service agencies (Malone 1998). This case study suggests that EDs manage to respond to patients’ social needs thanks to an elaborate financial and cultural system in which GME plays a substantial part. Any expansion of social service provision through the EDs needs to consider the relationship between social use, tenuous financing, and GME. For example, one indirect way to buttress the urban social safety net may be to train more emergency physicians.

Second, redefining social visits to the ED as opportunities for medical training addresses the practical problem of social use but raises ethical issues. Indigent patients in a safety-net ED are doubly constrained in their ability to opt out of being used as “teaching material” for student doctors. ED patients in general are constrained by the fact that they see their visit to the hospital as a response to an urgent need. Poor people, who may face social and financial barriers that prevent their going to EDs that are not part of teaching hospitals, may inevitably end up in the EDs of teaching hospitals. Addressing social needs through GME, therefore, raises considerable ethical questions.

Third, work culture can be a relatively autonomous shaper of everyday health care provision, and this reminder is apropos given recent changes in health care financing. Contemporary reform efforts seek to introduce market principles in health care provision, and researchers have noted organizational proliferation in some arenas of the health care industry (i.e., Bazzoli, Shortell, Dubbs, et al. 1999). However, even in the face of rapid change in financing and organization, core cultural concepts informing the practice of medicine remain (Good 1998). The case study of County ED shows how the practice of medicine may proceed according to a cultural logic, such as that surrounding GME, that may be somewhat insulated from pressures motivating overarching changes in the organizational culture of the health care industry. Although we might expect a safety-net facility to not be at the forefront of market-oriented change in health care, the case study does serve as a reminder that health care institutions continue to be shaped by nonmarket institutions such as medical education.

It is impossible to assess the generalizability of this case study’s findings and implications without further research. The value of a case study is to generate new ideas about what kinds of factors future researchers may want to consider. For example, future researchers might want to measure the frequency of social visits to the ED or the amount of uncompensated overtime worked by clinicians. These data could be collected through nonethnographic means but have not been to date. Another area for future research is to understand how safety-net facilities manage social use and how all health care providers manage tenuous financing absent of GME. Research might address how safety-net care is provided without GME, how GME occurs without a supply of indigent patients, and how safety-net care is affected by differences in reimbursement systems, patient population, amount of public financing, and facility type. In the face of continued market-oriented change in health care financing, it is increasingly important that health care policymakers and administrators understand how providers manage the seemingly inevitable organizational and financial problems of safety-net care in everyday life.