Full text of DOI:10.1111/j.1553-2712.2001.tb01111.x

1030
PROCEEDINGS
Richardson, Hwang • ACCESS TO CARE
Access to Care: A Review of the Emergency
Medicine Literature
LYNNE D. RICHARDSON, MD, ULA HWANG, MD, MPH
Abstract. The authors review the evolution of the denying emergency care and more sophisticated anal-
emergency medicine literature regarding emergency yses of actual costs led to reconsideration of initia-
department (ED) use and access to care over the past tives to limit access to ED care and renewed focus on
20 years. They discuss the impact of cost containment the critical role of the ED as a safety net provider. In
and the emergence of managed care on prevailing recent years, ‘‘de facto’’ denials of emergency care due
views of ED utilization. In the 1980s, the character- to long ED waiting times and other adverse conse-
ization of ‘‘nonurgent ED visits’’ as ‘‘inappropriate’’ quences of ED crowding have begun to dominate the
and high ED charges led to the targeting of non- emergency medicine health services literature. Key
emergency ED care as a potential source of savings. words: access to care; safety net; emergency depart-
During the 1990s the literature reveals multiple at- ment utilization; health services access; health ser-
tempts to identify ‘‘inappropriate’’ ED visits and to vices research. ACADEMIC EMERGENCY MEDI-
develop strategies to triage these visits away from the CINE 2001; 8:1030–1036
ED. By the late 1990s, demonstration of the risks of
HE emergency department (ED) is a unique of their income, race, ethnicity, insurance status,
practice setting in two respects. The first is or special needs. We are the ultimate safety net
its capacity to deliver a full range of medical ser- for those whom other providers turn away.
vices to acutely ill or injured patients, regardless
Because of the unique position of the ED within
6
T
of the nature of the presenting complaint. The the health care system, the use of the ED is a func-
other defining characteristic of the ED is its sin- tion of: 1) the nature, acuity, and severity of the
gular accessibility; 24 hours per day, seven days presenting complaint; and 2) the patient’s experi-
per week, the ED provides care to all who seek it, ence with and access to other health care settings
regardless of their ability to pay. Access to emer- prior to the ED visit. Analyzing and interpreting
gency care is ensured not only by the professional patterns of ED utilization gives valuable informa-
and ethical standards of emergency physicians tion about the other health care services available
1
(
EPs), but also by federal law. Either or both of in the community, as well as information regarding
these two defining characteristics may influence a the health status of the population. Many health
patient’s decision to use the ED.
As every EP knows, our health care system has have looked at ED utilization. The ED is used dis-
failed to provide timely and effective access to proportionately by patients without insurance, pa-
services researchers who are interested in access
health care for all individuals. For many people2,
the lack of access is due to their lack of insurance.
For many others, barriers to care exist because of
inadequate coverage, because of educational, cul-
tural, linguistic, logistical, psychosocial, environ-
mental, or institutional factors, or because of the
nature of their health problems. For vulnerable
and disenfranchised populations, the ED may be
the only accessible source of health care. We treat
all persons who come to us seeking care, regardless
tients with Medicaid, patients without primary
care physicians, members of racial and ethnic mi-
norities, and other ‘‘vulnerable populations.’’
7
–14
We review the emergency medicine (EM) liter-
ature on access over the past 20 years, examining
the impact of cost containment and the emergence
of managed care on prevailing views of ED utili-
zation. We discuss the role of the ED as a safety
net provider, and review the literature on ED
crowding in light of growing pressures on all safety
net providers due to the increasing number of un-
2
–5
From the Department of Emergency Medicine, Mount Sinai insured, the changing managed care market, and
School of Medicine, New York, NY (LDR, UH).
decreasing funding and reimbursement for uncom-
Received July 3, 2001; accepted July 12, 2001.
Presented at the Academic Emergency Medicine Consensus
Conference, ‘‘The Unraveling Safety Net,’’ Atlanta, GA, May
2
pensated care.
2
001.
ED VISITS FOR
NONEMERGENCY CARE
Address for correspondence and reprints: Lynne D. Richardson,
MD, 1 Gustave L. Levy Place, Box 1149, New York, NY 10029.
Fax: 212-426-1946; e-mail: lynne.richardson@mssm.edu
During the latter part of the 1970s and the 1980s,

ACADEMIC EMERGENCY MEDICINE • November 2001, Volume 8, Number 11
1031
decades of concern about the rising costs of health some for individual institutions, some for specific
care began to stimulate new strategies to contain populations, each using its own definiion of ‘‘non-
11–13,17,19–26
costs. The previous mechanisms, which relied pri- urgent’’.
marily on controlling charges, were supplanted by Several governmental reports addressing this
a new focus on controlling utilization of health care topic were issued; and analysts examined the lim-
services. Managed care began to gain favor as a ited data on ED visits available in national data-
method of organizing care that prevented the un- bases from surveys such as the National Health
necessary use of costly services. The relatively Interview Survey (NHIS) and the National Medi-
high charges for emergency care (reported to be up cal 9 Care Utilization and Expenditures Sur-
,10,15,27
to five times that of the average charge for a clinic vey.
In 1992, data collection was begun for
10,15
or physician’s office in the same community ), the National Hospital Ambulatory Medical Care
resulted in a great deal of attention to the use of Survey (NHAMCS) providing, for the first time,
the ED for problems that are not medical emer- comprehensive national data on emergency vis-
2
8–32
gencies.
its.
Unfortunately, the definition of a ‘‘nonur-
It is certainly true that some portion of ED vis- gent emergency visit’’ varies from source to source;
its are for problems that are neither emergent nor there are no standard criteria and no uniform
urgent, and that could competently be handled in categorization methodology. Some of the most com-
another setting. However, the characterization and monly quoted references on the proportion of ‘‘non-
quantification of these ‘‘nonurgent emergency vis- urgent ED visits’’ are described in Table 1.
its’’ have been problematic. The literature in this
Although ‘‘nonurgent ED visits’’ have yet to be
area is distinguished by a remarkable lack of pre- uniformly defined and methods to validly and re-
cision and widespread confusion between retro- liably quantify them have yet to be developed, a
spective and prospective judgments on the acuity consensus developed among government officials,
of the presenting problem. As early as 1980, pub- policymakers, insurers, and managed care organi-
lished articles describing ED visits on the basis of zations that the ‘‘the use of the emergency depart-
urgency began to appear in the EM and public ment for non-emergency care is frequent and
1
6–18
33
health literature.
Over the next two decades, costly.’’ In the early 1990s, concern about the
many authors published estimates of the propor- high cost of emergency care crystallized into the
tion of ED visits that are for nonurgent problems; widespread belief that diverting these ‘‘inappropri-
TABLE 1. Estimates of Nonurgent Emergency Department (ED) Visits
1
990: 43% of ED visits are nonurgent. U.S. General Accounting Office. Emergency Departments: Unevenly Affected by Growth
and Change in Patient Use. Publication No. B-251319. Washington, DC, Jan 1993.
A survey of 1,025 nonfederal hospitals conducted by the GAO. Data source was the opinion of the hospital official who responded
to the survey; this was in some cases an administrator and in others the ED medical director. Visits were classified as
emergent, urgent, or nonurgent. Nonurgent was defined as ‘‘not life or limb threatening or did not require immediate care
and probably could have been treated in a doctor’s office or clinic.’’
1
992: 55% of ED visits are nonurgent. McCaig LF. National Hospital Ambulatory Medical Care Survey: 1992 Emergency
Department Summary; advance data from Vital and Health Statistics. Hyattsville, MD: National Center for Health Statistics,
1
994; 245:1–12.
A national probability survey of visits to nonfederal hospital EDs and outpatient departments conducted by the National Center
for Health Statistics. The 1992 data were collected from 437 EDs; data source was a patient record form completed by hospital
staff at the selected hospitals during a four-week reporting period in 1992. Visits were classified as urgent or nonurgent.
Nonurgent visit defined as ‘‘those made by patients who did not require immediate attention or attention within a few hours.’’
1
994: 37% of ED visits are nonurgent. Young, GP, Wagner MB, Kellermann AL, Ellis J, Bouley D. Ambulatory visits to hospital
emergency departments. Patterns and reasons for use. 24 Hours in the ED Study Group. JAMA. 1996; 276:460–5.
An observational study of 6,187 patients arriving in 56 hospital EDs across the United States during a single 24-hour period.
Data source was the triage classification assigned to the patient by a nurse upon arrival in the ED. Each hospital used its
own institutional triage criteria. The percent of nonurgent ED visits reported by individual hospitals ranged from 23% to 72%
of ED visits.
1
997: 9.7% of ED visits are nonurgent. Nourjah P. National Hospital Ambulatory Medical Care Survey: 1997 Emergency
Department Summary (advance data). Hyattsville, MD: National Center for Health Statistics, 1999; 304:1–12.
A national probability survey of visits to nonfederal hospital EDs and outpatient departments conducted by the National Center
for Health Statistics. The 1997 data were collected from 392 EDs; data source was a patient record form completed by hospital
staff at the selected hospitals during a four-week reporting period in 1997. The 1997 NHAMCS contained a new data element
to assess the immediacy with which a patient should be seen; this was assigned upon arrival in the ED by triage staff. The
ED visits were classified as emergent (should be seen within 15 minutes of arrival), urgent (15–60 minutes), semiurgent (1–
2
hours), or nonurgent (2–24 hours). In the 1997 data, 21% were emergent, 32% were urgent, 15.4% were semiurgent, 9.7%
were nonurgent, and 21.9% were ‘‘unknown or no triage.’’

1032
PROCEEDINGS
Richardson, Hwang • ACCESS TO CARE
ate ED visits’’ to physician offices or other primary computerized algorithm-directed triage to direct
care settings would result in substantial sav- patients away from the ED to the acute care clinic
3
4,35
45
ings.
Third-party payers began to deny pay- at Brooke Army Medical Center. Seven hundred
ment for emergency services based on retrospec- thirty-three patients (1.2%) of the 58,282 patients
tive determinations of appropriateness based on triaged away were sent back to the ED for care.
ED diagnosis, and some managed care organiza- Based on these data, it was concluded that com-
tions began to require preauthorization for emer- puterized algorithm-directed triage, using mini-
3
6,37
gency visits.
mally trained personnel, was an effective system
Several authors cautioned that the characteri- for separating ‘‘walk-in’’ patients from emergency
zation of ‘‘nonurgent ED visits’’ as ‘‘inappropriate patients. The author did not suggest that the sys-
11,38,39
ED visits’’ might be inappropriate.
The view tem was an acceptable method to identify patients
of the patient regarding ‘‘appropriateness’’ may di- who could safely be refused care.
verge significantly from that of the provider; the
In 1990, Derlet and Nishio from University of
perspective of the EP diverges from that of the in- California, Davis (UC Davis), reported their expe-
surer, and opinions among physicians vary with rience with a policy of refusing care to patients
4
0–42
their training and experience.
In the presence who presented to the ED after an extensive medi-
of existing financial, temporal, and institutional cal screening examination performed by specially
4
7
barriers to appropriate primary care, the decision trained triage nurses. Using a protocol reviewed
by a patient to forgo an ED visit for a nonurgent and approved by the institution’s legal counsel,
problem may lead to the later occurrence of a life- 19% (4,186) of the 22,390 patients who presented
threatening problem that necessitates an emer- to the ambulatory triage area between July and
4
3,44
gent ED visit.
From the perspective of an in- December 1988 were determined to have nonemer-
dividual with limited resources, for whom an ED gencies; these patients were given a list of clinics
visit may be the only available source of health (on site and off site) or referred to their personal
care, a nonurgent ED visit may be far more appro- physicians. Follow-up consisted of identifying pa-
4
,39
priate than seeking no care at all.
tients who returned to the authors’ ED within 48
Despite these cautionary voices, cost-conscious hours and a survey (via letters or telephone call)
gatekeepers began to erect financial and adminis- of the clinics to which patients had been referred
trative barriers to emergency care. Believing the regarding ‘‘adverse effects.’’
ED to be ‘‘the most expensive place of all’’ (so de-
Derlet and Nishio’s conclusion that patients
scribed by President Bill Clinton in his televised could safely be turned away from the ED sparked
address to the joint session of Congress in Septem- a major controversy among EM academicians.
ber 1993), ‘‘nonurgent ED visits’’ became a target Critics challenged the lack of adequate outcome
for cost reduction. Policymakers sought to develop data: patients were not individually tracked and
strategies to delay or deny emergency care; re- specific patient health outcomes, rehospitalization
searchers attempted to develop criteria to deter- rates, morbidity/mortality rates, or rates of patient
mine who should receive emergency care and who subsequent presentation to another ED after re-
should not.
ferral to clinics, were not measured. Nevertheless,
UC Davis continued its policy of denying care to
‘‘nonemergencies’’ in their ED; in 1992 Derlet et al.
published data for three years, reporting a subse-
DENYING CARE IN THE ED
Almost every ED in the country uses a triage sys- quent hospitalization rate of 0.02% among those
55
tem to prioritize patients’ need for care. Many re- patients who were refused ED care.
searchers began to look at triage as a possible
Both Birnbaum et al. and Lowe et al., in sepa-
mechanism to limit nonemergency ED care. Cri- rate studies, were unable to replicate Derlet et al.’s
46,48
teria similar to those used to determine which pa- findings.
In a historical cohort study, Lowe et
tients need to be seen first in the ED were evalu- al. identified 496 patients who presented to the ED
ated for utility in determining who needs to be at San Francisco General Hospital in July 1990
seen at all. A review of the EM literature during who met the inclusion criteria used by the Derlet
the last decade reveals multiple attempts to de- study; 106 of these patients would have been re-
4
8
velop a reliable method to prospectively determine fused care by the Derlet triage guidelines. Using
the necessity for emergency care. Investigators two separate definitions of ‘‘appropriate,’’ one
used a variety of methods including specific clinical based on explicit clinical criteria, the second on
criteria, computer-driven algorithms, individual expert opinion, 33% (35) were deemed to be ‘‘ap-
expert assessment, and multidisciplinary physi- propriate’’ visits by both definitions; and 3.8% (4)
4
5–54
cian-panel-developed screening exams.
In 1989, Berman et al. used a retrospective au-
were hospitalized.
In a prospective, observational, cohort study,
dit of 98,086 charts to evaluate the effectiveness of Birnbaum et al. used the Derlet published criteria

ACADEMIC EMERGENCY MEDICINE • November 2001, Volume 8, Number 11
1033
for refusal of care to identify a convenience sample panel; they used a cohort of 1,187 consecutive
of 534 adults who presented to Bronx Municipal adult walk-in ED patients at the Los Angeles Vet-
Hospital Center ED in New York from July to mid- erans Affairs Medical Center with selected com-
4
6
October 1992. The study endpoint was ED dis- plaints to prospectively assess the criteria’s utility
position; no patient was lost to follow-up. Of the to avoid hospitalization within seven days, and
5
1
5
34 patients who met the Derlet criteria for refusal death within 30 days. Two hundred six (19%) pa-
of care, 1.1% (n = 6, 95% CI = 0.4% to 2.4%) were tients met screening criteria for deferred care;
hospitalized; this was more than 50 times the there was no hospitalization or death within the
0
.02% hospitalization rate reported by Derlet and study period. The study’s limitations included
colleagues.
small sample size, idiosyncratic criteria, and lim-
Both of these studies, Lowe et al. and Birnbaum ited outcome measures.
et al., challenged the safety of using the Derlet tri-
During the past 20 years, no valid and reliable
age guidelines to refuse ED patient care. In addi- definition of what constitutes an ‘‘appropriate ED
tion to the lack of sensitivity of such guidelines for visit’’ has emerged from the literature. Lowe and
predicting important medical outcomes, these au- Bindman warn that ‘‘limiting patients’ access to
thors point out that refusing care to patients pre- the emergency department without the aid of such
senting to an ED can be challenged on ethical, fi- a definition could result in barriers to needed care
56
nancial, and legal grounds. Both cautioned and harm to patients’ health.’’ The failure to de-
institutions that might be considering the imple- velop accepted criteria to determine the need for
mentation of triage guidelines to refuse care to ED emergency care, fueled by growing concern about
patients.
increasingly stringent Emergency Medical Treat-
Despite these studies repudiating the Derlet ment and Labor Act (EMTALA) regulations led
model, efforts to develop a system to safely assess many to question whether patients could safely be
1
the need for emergency care continued. Waldrop et refused care in the ED. Other authors have chal-
al. attempted to determine the sensitivity and lenged the assertion that decreasing ‘‘nonurgent
specificity for predicting admission of an assigned ED visits,’’ even if it could be done safely, would in
5
7,58
triage acuity of ‘‘nonemergency’’ in an established fact save money.
ED triage system at an urban hospital in Baton
Rouge, Louisiana. Their findings were limited by
the loss to follow-up of 8.25% of their cohort. Brill-
5
0
THE COST OF ED CARE
man et al. used a randomized partial crossover de- The perception that ED overuse is an important
sign to compare triage assessments by computer, cause of high medical care costs arose because of
nurses, and physicians in 5,106 consecutive pa- the markedly higher charges of EDs relative to
35
tients presenting to a university teaching hospital clinic or physician office charges. In 1996, using
5
4
in June and July 1992. None of the three triage 1991–1993 cost data for hospital and physicians’
methods performed well in predicting which pa- services from six community hospitals in Michi-
tients required admission. The authors concluded gan, Williams reported a detailed analysis of ED
that ‘‘Until triage methods are standardized and costs relative to physician office costs, demonstrat-
validated, triage decisions should not be used to de- ing that an accurate comparison can be made only
termine the timeliness of access to emergency care.’’ by examining the actual costs of services in the two
In 1997, O’Brien et al. used chart reviews of 892 different settings, rather than simply comparing
34
ED visits to compare a predetermined list of non- the charges. Because the ED functions 24 hours
urgent complaints, triage assessment by inter- per day, seven days per week, it has high fixed
nists, and triage assessment by EPs to identify pa- costs (those costs that are not dependent on vol-
tients who ‘‘could have been taken care of within ume) for medical staff, ancillary services, supplies,
2
4 hours by a primary care physician without overhead, and administration; and very low mar-
harm to the patient’’; this study found only mod- ginal costs (the additional cost for one additional
4
2
erate rates of agreement (␬ = 0.47). In 1998, visit). Based on these data, Williams concluded
Wuerz et al. had 87 participants, emergency med- that ‘‘the potential savings from a diversion of non-
ical technicians (EMTs) and ED nurses, rate urgent visits to private physicians’ offices may be
scripted patient scenarios for severity, urgency, much less than is widely believed.’’
likely disposition, and medical resource utilization;
Furthermore, it has been suggested that redi-
poor interrater agreement and intrarater incon- rection of ‘‘nonurgent ED visits’’ from EDs to other
gruities led the authors to question the reliability practice settings would require that such settings
5
3
of current ED triage practice.
be similarly equipped and staffed, and immedi-
As recently as 2000, Washington et al. pub- ately available at all hours; total costs to the sys-
lished a set of clinical criteria for deferred care tem would actually increase by creating 24-hour-
4
from a 17-member multidisciplinary physician per-day walk-in capacity in primary care centers.

1034
PROCEEDINGS
Richardson, Hwang • ACCESS TO CARE
For example, the cost of seeing an additional pa-
tient at 4 PM may be slightly lower in a physician’s
office than in the ED. The cost of seeing that pa-
tient at 4 AM when the physician’s office is not
open is much greater; the marginal cost of caring
C
ROWDING IN THE ED
Throughout the 1990s, ED visits steadily in-
creased, while the number of hospital EDs de-
6
0,61
creased.
The increase in ED visits has been at-
34
for this patient in the ED at 4 AM is very low.
tributed to the growing number of uninsured and
underinsured individuals, aggressive gatekeeping
by managed care primary care providers, the aging
of our population, and declining access to primary
care during this period. The number of uninsured
is now estimated to be more than 44 million, 18%
Also in 1996, Tyrance et al. looked at data
released by the National Medical Expenditure Sur-
vey (NMES) of 35,000 persons in 14,000 house-
holds representative of the U.S. civilian, noninsti-
5
7
tutionalized population. Emergency department
expenditures accounted for only 1.9% of the total
national health expenditures. Thus, ED use ac-
counts for a small share of U.S. medical care costs.
The authors concluded that limiting ED use cannot
generate substantial savings and may instead tar-
get the poor, who receive much of their outpatient
care in the ED.
2,62
of the nonelder U.S. population.
Despite this
growing burden of uncompensated care, pressures
to reduce health care costs have resulted in de-
creased direct and indirect funding for uncompen-
6
3–65
sated care.
In 2000, the Institute of Medicine (IOM) issued
a report describing the forces that threaten the fu-
ture survival of safety net providers, providers who
care for the uninsured and other vulnerable pop-
ulations. The ED, the ultimate safety net provider,
is being adversely impacted by many of these
In still another 1996 article, Selby et al. dem-
onstrated that requiring a copayment for using the
ED resulted in a decrease in ED visits after con-
trolling for age, sex, socioeconomic status, and pre-
2
3
3
66
vious level of ED use. Although the authors con-
cluded that instituting a copayment ‘‘can safely
reduce inappropriate use of the emergency depart-
ment,’’ the data revealed that the copayment sig-
nificantly reduced visits for presenting conditions
that were classified as ‘‘often an emergency’’ or
forces. The unprecedented growth of ED visits, in
the face of decreasing primary care capacity and
decreasing hospital bed capacity, has resulted in
dangerously crowded EDs in various parts of the
6
7–70
country.
It is ironic that as attempts to triage
patients away from the ED are being largely aban-
doned, ED crowding is resulting in a form of ‘‘de-
facto’’ triage of patients out of the ED, also limiting
‘
‘sometimes not an emergency’’ as well as those
that were classified as ‘‘often not an emergency.’’
In an earlier paper, Young pointed out that de-
lays in the care of patients with emergency condi-
tions may result in increased morbidity and mor-
37,71,72
patient access to health care.
In an observational cohort study, Bindman et
al. surveyed 700 patients who were not referred by
triage for immediate care in the San Francisco
General Hospital ED waiting area in July 1990,
comparing the responses of patients who left with-
out being seen by a physician with the responses
of those who were seen by a physician. Patients
were contacted again seven and 14 days later;
medical record information from all hospitals and
neighborhood clinics in San Francisco was re-
viewed and deaths during the study period were
3
7
tality. He suggested that comparisons between
ED costs and primary care office or clinic costs
should factor in time lost from work as patients
make daytime appointments. The ED, with its 24-
hour availability and concentration of diagnostic
and therapeutic resources, may be the most cost-
effective way to deliver acute care, regardless of
71
3
7,58
the type of health care delivery system.
With the myth of substantial savings from the
control of ED use dispelled, and with more vigor- tracked via the San Francisco County Vital Statis-
ous enforcement of EMTALA regulations by the tics Registry. The ED waiting times and the
Health Care Financing Administration (HCFA), changes in patients’ self-reported health were mea-
enthusiasm for refusing care to patients who pre- sured. Patients were more likely to leave as wait-
sent to the ED had waned substantially by the end ing times increased, and those who left without be-
of the last decade. The federal EMTALA statute ing seen were twice as likely to report worsening
mandates a medical screening examination for all of their presenting problems; 4% of those who left
who seek emergency care, yet the U.S. government the ED without seeing a physician required sub-
has never assumed financial responsibility for this sequent hospitalization and 27% returned to an
1
service. As Williams pointed out in a March 2000 ED. The authors concluded that the health of some
editorial, under our market-driven multipayer sys- patients may be jeopardized by long waits for
tem, there is a strong role for the government to emergency care.
5
8
guarantee access for uninsured individuals. How-
Baker et al. conducted a prospective study of
ever funding of the ‘‘health care safety net’’ re- 186 patients who left the Harbor–UCLA Medical
mains problematic. Currently only 50% of all emer- Center ED during a two-week block in 1990 with
5
9
gency care is compensated.
follow-up at seven days, comparing them with

ACADEMIC EMERGENCY MEDICINE • November 2001, Volume 8, Number 11
1035
2
000, pp 16–46.
a random sample of 211 patients who waited to be
seen by a physician during the same time period.
3. Rask KJ, Williams MV, Parker RM, McNagny SE. Obstacles
72
predicting lack of a regular provider and delays in seeking care
Forty-six percent of those patients who left were for patients at an urban public hospital. JAMA. 1994; 27:1931–
3
3.
judged to need immediate medical attention by the
triage nurse and 29% were assessed as needing
4
. Richardson LD. Access, the emergency department and
asthma. In: Brenner BE (ed). Emergency Asthma. New York:
care within 24–48 hours; 11% of those who left Marcel Dekker, 1999, pp 557–63.
5
. Woloshin S, Bickell NA, Schwartz LM, Gany F, Welch HG.
were hospitalized within the next week, and three
individuals required emergency surgery. The co-
Language barriers in medicine in the U.S. JAMA. 1995; 273:
7
24–8.
hort of patients who left did not differ significantly 6. Richardson LD, Hwang U. America’s health care safety net:
intact or unraveling? SAEM Newslett. 2001; XIII(2):7.
from those who stayed with respect to measures of
illness acuity. The authors concluded that ED over-
7
. Commonwealth Fund. National Comparative Survey of Mi-
nority Health Care. New York: Commonwealth Fund, 1995.
crowding in fact restricts access to needed ambu- 8. Krieger N, Rowley DL, Herman AA, Avery B, Phillips MJ.
Racism, sexism and social class: implications for the studies
latory care.
of health, disease, and well-being. Am J Prev Med. 1993;
In recent years, articles focusing on the scope
9
(6 suppl):82–122.
and severity of ED crowding have begun to domi- 9. U.S. General Accounting Office. Emergency Departments:
Unevenly Affected by Growth and Change in Patient Use. Pub-
nate the EM health services literature. In 1999,
lication No. B-251319. Washington, DC, Jan 1993.
Graff et al. labeled ED crowding as an ‘‘interna-
1
0. U.S. Department of Health and Human Services, Office of
69
tional symptom of health care system failure.’’ In the Inspector General. Use of Emergency Rooms by Medicaid
Recipients. Washington, DC, 1992.
2
000, Derlet and Richards, having ended the prac-
1
1. Pane GA, Farner MC, Salness KA. Health care access
tice of denying ED care described in earlier
problems of medically indigent emergency department walk-in
publications, authored an article entitled: ‘‘Over- patients. Ann Emerg Med. 1991; 20:730–3.
1
2. Petersen LA, Burstin HR, O’Neil AC, Orav EJ, Brennan
crowding in the nation’s emergency departments:
TA. Nonurgent emergency department visits—the effect of
6
8
complex causes and disturbing effects.’’
having a regular doctor. Med Care. 1998; 26:1249–55.
1
3. Baker DW, Stevens CD, Brook RH. Regular source of am-
bulatory care and medical care utilization by patients present-
CONCLUSIONS
ing to a public hospital emergency department. JAMA. 1994;
2
71:1909–12.
1
4. Purdie FRJ, Honigman B, Rosen P. The chronic emergency
A review of the EM literature regarding ED use
and access to care over the past 20 years reveals
department patient. Ann Emerg. Med. 1981; 10:2998–301.
1
5. U.S. Department of Health and Human Services, Office of
significant evolution. In the 1980s, pressure to re- the Inspector General. Nonurgent use of hospital emergency
department by Medicaid and Medicare beneficiaries: a pro-
duce costs through controlling utilization led to
scrutiny of the use of highly priced EDs for non-
gram inspection for the Office of the Inspector General. Wash-
ington, DC, 1983.
emergency care. ‘‘Nonurgent ED visits’’ were re- 16. Gifford MJ, Franaszek JB, Gibson G. Emergency physi-
cians’ and patients’ assessments: urgency of need for medical
garded as ‘‘inappropriate’’ and targeted as a poten-
tial source of savings. This led to multiple attempts
care. Ann Emerg Med. 1980; 9:502–7.
1
7. Guterman JJ, Franaszek JB, Murdy D, Gifford M. The
in the 1990s to identify those visits that were ‘‘non- 1980 patient urgency study: further analysis of the data. 1985;
1
4:1191–8.
urgent’’ or ‘‘inappropriate’’ and to develop strate-
gies to triage these ‘‘inappropriate visits’’ away
from the ED.
Demonstration of the risks of denying emer-
gency care and more sophisticated analyses of the
1
8. Hurley R, Freund D, Taylor D. Emergency room use of
primary care case management: evidence from four Medicaid
demonstration programs, Am J Public Health. 1989; 79:843–7.
1
9. Shesser R, Kirsch T, Smith J, Hirsch R. An analysis of
emergency department use by patients with minor illness. Ann
Emerg Med. 1991; 20:743–8.
actual cost of providing nonemergency care in the 20. Rask KJ. Ambulatory health care use by patients in a pub-
lic hospital emergency department. J Gen Intern Med. 1998;
ED led to reconsideration of initiatives to deny
emergency care. The unprecedented growth in the
1
2
3:614–20.
1. Buesching DP, Jablonowski A, Vesta E, et al. Inappropri-
number of uninsured, in the face of continued ef- ate emergency department visits. Ann Emerg Med. 1985; 14:
6
72–6.
forts to reduce health care costs, has heightened
the critical role of EDs as safety net providers. In
2
2. Fong C. The influence of insurance status on non-urgent
pediatric visits to the ED. Acad Emerg Med. 1999; 6:744–8.
recent years, ED crowding has been shown to re- 23. Strange GR, Chen EH, Sanders AB. Use of emergency de-
partments by elderly patients: projections from a multicenter
strict access to needed care. Current literature is
data base. Ann Emerg Med. 1992; 21:819–24.
focused on the dangers of patient overcrowding in
our nation’s EDs.
2
4. Weinerman RE. Yale studies in ambulatory medical care
V: determinants of use of hospital emergency services. Am J
Public Health. 1966; 56:1037–56.
2
5. Ziv A, Boulet JR, Slap GB. Emergency department utili-
zation by adolescents in the United States. Pediatrics. 1998;
01:987–94.
. Emergency Medical Treatment and Active Labor Act 26. Pilossoph-Gelb S, Mower WR, Ajaelo I, Yang SC. Psycho-
EMTALA), codified as amended at 42 U.S.C. §1395dd, 1990; social difficulties and emergency department use. Acad Emerg
References
1
1
(
Health Care Financing Administration EMTALA Regulations, Med. 1997; 4:589–92.
42 C.F.R. Parts 488, 489, 1003, 1994.
2
27. Chyba M. National Center for Health Statistics: Utiliza-
. Lewin ME, Altman S (eds). Background and Overview. In: tion of Hospital Emergency and Outpatient Departments
America’s Health Care Safety Net: Intact but Endangered. In- (Jan–Jun 1980), National Medical Care Utilization & Expen-
stitute of Medicine. Washington, DC: National Academy Press, diture Survey, 1983, Preliminary report 2. Hyattsville, MD:

1036
PROCEEDINGS
Richardson, Hwang • ACCESS TO CARE
National Center for Health Statistics, (DHHS) 83-20000.
Fink A, Brook RH. Safely directing patients to appropriate lev-
2
8. McCaig LF. National Hospital Ambulatory Medical Care els of care: guideline-driven triage in the emergency service.
Survey: 1992 ED Survey (advance data from Vital and Health Ann Emerg Med. 2000; 36:15–22.
Statistics). Hyattsville, MD: National Center for Health Sta- 52. Wilson LO, Wilson FP Jr, Wheeler M. Computerized triage
tistics, 1994, 245:1–12.
of pediatric patients: automated triage algorithms. Ann Emerg
2
9. Stussman BJ. National Hospital Ambulatory Medical Care Med. 1981; 10:636–40.
Survey: 1995 ED Survey (advance data). Hyattsville, MD: Na- 53. Wuerz R, Fernandes CM, Alarcon J. Inconsistency of emer-
tional Center for Health Statistics, 1997, 285:1–19. gency department triage, Ann Emerg Med. 1998; 32:431–5.
0. McCaig L. National Hospital Ambulatory Medical Care 54. Brillman JC, Doezema D, Tandberg D, et al. Triage: limi-
Survey: 1996 ED Survey (advance data). Hyattsville, MD: Na- tations in predicting need for emergency care and hospital ad-
tional Center for Health Statistics, 1997, 293:1–20. mission. Ann Emerg Med. 1996; 2:493–500.
1. Nourjah P. National Hospital Ambulatory Medical Care 55. Derlet RW, Nishio DA, Cole LM, Silva J. Triage of patients
Survey: 1997 ED Survey (advance data). Hyattsville, MD: Na- out of the emergency department: three year experience. Am J
tional Center for Health Statistics, 1999, 304:1–12. Emerg Med. 1992; 10:195–9.
2. McCaig L. National Hospital Ambulatory Medical Care 56. Lowe RA, Bindman AB. Judging who needs emergency de-
Survey: 1998 ED Survey (advance data). Hyattsville, MD: Na- partment care: a prerequisite for policy-making. Am J Emerg
tional Center for Health Statistics, 2000; 313:1–24. Med. 1997; 15:133–6.
3. Selby J, Fireman BH, Swain BE. Effect of a co-pay on the 57. Tyrance PH, Himmelstein DU, Woolhandler SW. U.S.
use of emergency department in a health maintenance orga- emergency department costs: no emergency. Am J Public
3
3
3
3
nization. N Engl J Med. 1996; 334:635–41.
Health. 1996; 86:1527–31.
3
4. Williams R. The costs of visits to emergency departments. 58. Williams R. Single-payer and all-payer systems: implica-
N Engl J Med. 1996; 334:642–6.
3
tions for emergency medicine in the United States. Ann Emerg
5. Baker LC, Baker LS. Excess cost of emergency depart- Med. 2001; 37:337–9.
ments for non-urgent care. Health Aff (Milwood). 1994; 13(5): 59. Knopp RK. Health care reform and the safety net. Ann
162–71.
3
Emerg Med. 1996; 27:234–46.
6. Shaw KN, Selbst SM, Gill FM. Indigent children who are 60. American College of Emergency Physicians. EM Statistical
denied care in the emergency department. Ann Emerg Med. Profile; March 23, 2000, Fact Sheet. Website: http://www.acep.
1
990; 19:59–62.
7. Young G. Health care reform and emergency medicine. 61. Fields W (ed). Defending America’s Safety Net. Report of
the American College of Emergency Physicians 1998–1999
8. Lowe RA, Young GP, Reinke B, White JD, Auerbach PS. Safety Net Task Force. Dallas, TX: American College of Emer-
Indigent health care in emergency medicine: an academic per- gency Physicians, 1999.
spective. Ann Emerg Med. 1991; 20:790–4. 62. Fronstin P. Sources of health insurance and characteristics
9. Kellermann AL. Visits to a hospital ED for non-emergency of the uninsured: analysis of the March 1999 Current Popu-
care: inappropriate use or unmet need? Nashville, TN: Ten- lation Survey (EBRI Issue Brief No 217). Washington, DC: Em-
nessee State Legislature, March 29, 1993. ployee Benefit Research Institute, 2000.
0. Gill J. Non-urgent use of hospital emergency departments: 63. Rosenbaum S, Darnell J. An analysis of the Medicaid and
org/2,381,0.html.
3
Ann Emerg Med. 1995; 25:666–74.
3
3
4
urgency from the patient’s perspective. J Fam Pract. 1996; 42: health-related provisions of the Personal Responsibility and
4
91–6.
Work Opportunity Reconciliation Act of 1997 (PL 104-193).
4
1. Lowe RA, Bindman AB. Judging who needs emergency de- Washington, DC: Center for Health Policy Research, George
partment care: a prerequisite for policy-making. Am J Emerg Washington University, 1997.
Med. 1997; 15:133–6. 64. Schneider A. Overview of Medicaid managed care provi-
4
2. O’’Brien GM, Shapiro MJ, Fagan MJ, Wollard RW, sions in the Balanced Budget Act of 1997. Washington, DC: The
O’Sullivan PS, Stein MD. Do internists and emergency physi- Henry J. Kaiser Family Foundation, 1997.
cians agree on the appropriateness of emergency department 65. National Association of Community Health Centers. Com-
visits? J Gen Intern Med. 1997; 12:188–91.
4
promise delays phase-out of health center payment systems-
3. Medicaid Access Study Group. Access of Medicaid recipi- orders congressional report on impact and alternative payment
ents to outpatient care. N Engl J Med. 1994; 330:1426–30.
mechanisms. Website: http://www.nachc.com/FSA/Federal/
4
4. Rask KJ. Ambulatory health care use by patients in a pub- Agenda/PPS/Compromise%20Announcement.htm.
lic hospital emergency department. J Gen Intern Med. 1998; 66. Richardson LD, Hwang U. America’s health care safety
13:614–20.
4
net: intact or unraveling? Acad Emerg Med. 2001; 8:1056–63.
5. Berman DA, Coleridge ST, McMurry TA. Computerized al- 67. American College of Physicians Overcrowding Task Force.
gorithm directed triage in the emergency department. Ann Measures to deal with emergency department overcrowding.
Emerg Med. 1989; 18:141–4. Ann Emerg Med. 1990; 19:944–5.
6. Birnbaum AB, Gallagher J, Utkewicz M, Gennis P, Carter 68. Derlet RW, Richards JR. Overcrowding in the nation’s
4
W. Failure to validate a predictive model for refusal of care to emergency departments: complex causes and disturbing ef-
emergency department patients. Acad Emerg Med. 1994;1: fects. Ann Emerg Med. 2000; 35:63–8.
2
13–7
69. Graff L. Overcrowding in the ED: an international symp-
4
7. Derlet RW, Nishio DA. Refusing care to patients who pre- tom of health care system failure. Am J Emerg Med. 1999; 17:
sent to an emergency department. Ann Emerg Med. 1990; 19: 208–9.
2
62–7.
70. Taylor T. Position Statement on the Critical State of Emer-
8. Lowe RA, Bindman AB, Ulrich SK, et al. Refusing care to gency Care in Arizona. Arizona Emergency Care Crisis Meet-
4
emergency department patients: evaluation of published triage ing Summary, 12/18/2000. Phoenix, AZ: Governor’s Adminis-
guidelines. Ann Emerg Med. 1994; 23:286–93. trative Offices Conference, 2000.
9. Verdile VP, Paris PM, Stewart RD, Verdile LA. Emergency 71. Bindman AB, Grumbach K, Keane D, Rauch L, Luce JM.
4
department telephone advice. Ann Emerg Med. 1989; 18:278– Consequences of queuing for care at a public hospital emer-
282.
5
gency department. JAMA. 1991; 266:1091–6.
0. Waldrop RD, Harper DE, Mandry C. Prospective assess- 72. Baker DW, Stevens CD, Brook RH. Patients who leave a
ment of triage in an urban emergency department. South Med public hospital emergency department without being seen by
J. 1997; 90:1208–12. a physician, causes and consequences. JAMA. 1991; 266:1085–
1. Washington DL, Stevens CD, Skekelle PG, Baker DW, 90.
5