This is the accessible text file for GAO report number GAO-03-460 
entitled 'Hospital Emergency Departments: Crowded Conditions Vary among 
Hospitals and Communities' which was released on March 28, 2003.



This text file was formatted by the U.S. General Accounting Office 

(GAO) to be accessible to users with visual impairments, as part of a 

longer term project to improve GAO products’ accessibility. Every 

attempt has been made to maintain the structural and data integrity of 

the original printed product. Accessibility features, such as text 

descriptions of tables, consecutively numbered footnotes placed at the 

end of the file, and the text of agency comment letters, are provided 

but may not exactly duplicate the presentation or format of the printed 

version. The portable document format (PDF) file is an exact electronic 

replica of the printed version. We welcome your feedback. Please E-mail 

your comments regarding the contents or accessibility features of this 

document to Webmaster@gao.gov.



Report to the Ranking Minority Member, Committee on Finance, U.S. 

Senate:



United States General Accounting Office:



GAO:



March 2003:



HOSPITAL EMERGENCY DEPARTMENTS:



Crowded Conditions Vary among Hospitals and Communities:



GAO Highlights:



Highlights of GAO-03-460, a report to the Ranking Minority Member, 

Committee on Finance, U.S. Senate 



Why GAO Did This Study:



Hospital emergency departments are a major part of the nation’s health 

care safety net. Emergency departments report being under increasing 

pressure, with the number of visits nationwide increasing from an 

estimated 95 million in 1997 to an estimated 108 million in 2000. GAO 

was asked to provide information on emergency department crowding, 

including the extent hospitals located in metropolitan areas are 

experiencing crowding, the factors contributing to crowding, and the 

actions hospitals and communities have taken to address crowding.



To conduct this work, GAO surveyed over 2,000 hospitals and about 74 

percent responded. The survey collected information on crowding, such 

as data on diversion—that is, the extent to which hospitals asked 

ambulances that would normally bring patients to their hospitals to go 

instead to other hospitals that were presumably less crowded.



What GAO Found:



While many emergency departments across the country reported some 

degree of crowding, the problem is more pronounced in certain hospitals

and communities. For example, while 2 of every 3 hospitals reported 

asking ambulances to be diverted to other hospitals at some point in 

fiscal year 2001, a smaller portion—about 1 of every 10—reported being 

on diversion status for more than 20 percent of the year. Hospitals in 

areas with larger populations, areas with high population growth in 

recent years, and areas with higher-than-average percentages of people 

without health insurance reported higher levels of crowding.



While no single factor stands out as the reason why crowding occurs, 

GAO found the factor most commonly associated with crowding was the 

inability to transfer emergency patients to inpatient beds once a 

decision had been made to admit them as hospital patients rather than 

to treat and release them. When patients “board” in the emergency 

department due to the inability to transfer them elsewhere, the space, 

staff, and other resources available to treat new emergency patients 

are diminished.



Hospitals and communities reported a variety of actions to address 

crowding, including expanding their emergency departments and 

developing ways to transfer emergency patients to inpatient beds more 

efficiently. For the most part, these actions have not been 

extensively evaluated, so their effect is unknown. However, the widely 

varying characteristics between hospitals mean that no one approach is 

likely to emerge as a way to address this ongoing concern.



Representatives from the American College of Emergency Physicians and 

the American Hospital Association and an independent reviewer provided 

comments on a draft of this report, which we incorporated as 

appropriate. 



www.gao.gov/cgi-bin/getrpt?GAO-03-460.

To view the full report, including the scope and methodology, click on

the link above. For more information, contact Janet Heinrich on (202)

512-7119. March 2003



[End of section]



GAO-03-460:



Contents:



Letter:



Results in Brief:



Background:



Emergency Department Crowding Is More Pronounced in Some Hospitals and 

Certain Types of Communities:



Availability of Inpatient Beds for Emergency Patients Cited as a Key 

Factor Contributing to Crowding, but Other Factors Also Contribute:



Wide Range of Activities Under Way to Manage Crowding at Hospitals and 

in Communities, but Problems Persist:



Concluding Observations:



Comments from External Reviewers:



Appendix I: Scope and Methodology:



Survey of Hospitals:



Site Visits:



Appendix II: Diversion Policies at the Six Locations GAO 

Visited:



Appendix III: Select Results of GAO Survey of Hospitals 

Regarding Emergency Department Crowding:



Appendix IV: GAO Contact and Staff Acknowledgments:



GAO Contact:



Acknowledgments:



Tables:



Table 1: Indicators of Emergency Department Crowding:



Table 2: Indicators of Crowding, by Population of MSA:



Table 3: Indicators of Crowding, by Population Growth of MSAs:



Table 4: Indicators of Crowding, by Percentage of MSA Population 

without Health Insurance:



Table 5: Indicators of Crowding, by Admissions per Bed in the MSA:



Table 6: Indicators of Crowding, by Percentage of Emergency Visits 

Resulting in Hospital Inpatient Admissions, Fiscal Year 2001:



Table 7: Indicators of Crowding, by Average Occupancy as a Percentage 

of Staffed Inpatient Beds, Fiscal Year 2001:



Table 8: Examples of Expansions of Emergency Departments or Inpatient 

Capacity at Hospitals GAO Visited:



Table 9: Examples of Hospitals’ Increasing Efficiency:



Table 10: Diversion Task Force Activity:



Table 11: Characteristics of Locations Selected for Site Visits:



Table 12: Comparison of EMS Areas and Diversion Policies for Site Visit 

Locations:



Table 13: Characteristics of Hospitals in Survey Universe:



Table 14: Hospitals by Volume of Emergency Department Patient Visits, 

Fiscal Years 1997 and 2001:



Table 15: Mean Number of Emergency Department Standard and Other 

Treatment Spaces and Increase in Treatment Spaces, Last Day of Fiscal 

Years 1997 and 2001:



Table 16: Hospitals Reporting Problems with On-Call Physician Specialty 

Coverage during Fiscal Year 2001:



Table 17: Specialty Areas for Which Hospitals Reported Having Problems 

with On-Call Physician Specialty Coverage in the Emergency Department 

during Fiscal Year 2001:



Table 18: Hospitals on Diversion, Fiscal Year 2001:



Table 19: Hospitals by Percentage of Time on Diversion, Fiscal Year 

2001:



Table 20: Reasons Contributing to the Hospital Not Going on Diversion 

in Fiscal Year 2001:



Table 21: Trauma Center Status and Diversion, Fiscal Year 2001:



Table 22: Conditions Contributing to Hospitals Going on Diversion, 

Fiscal Year 2001:



Table 23: Methods Hospitals Used to Minimize Diversion, Hospitals That 

Diverted in Fiscal Year 2001:



Table 24: Hospitals Reporting State or Local Laws or Rules That 

Restrict When Hospitals Can Go on Diversion:



Table 25: Hospitals’ Knowledge of When Other Hospitals Are on 

Diversion:



Table 26: Methods for Learning about Other Hospitals’ Diversion:



Table 27: Type of Care the Hospital Was Unable to Receive or Accept for 

the Most Recent Episode of Diversion:



Table 28: Hospitals Boarding Patients 2 Hours or More, Past 

12 Months:



Table 29: Hospitals by Percentage of Patients Boarded 2 Hours or More, 

Past 12 Months:



Table 30: Hospitals by Average Hours of Patients Boarding, Past 

12 Months:



Table 31: Conditions Contributing to Boarding Patients 2 Hours or More 

in Past 12 Months:



Table 32: Hospitals by Percentage of Patients Who Left after Triage but 

before a Medical Evaluation, Fiscal Year 2001:



Table 33: Indicators of Crowding, by Number of Staffed Inpatient Beds, 

Last Day of Fiscal Year 2001:



Table 34: Indicators of Crowding, by Number of Emergency Department 

Visits per Standard Treatment Space, Fiscal Year 2001:



Table 35: Indicators of Crowding, by Number of Emergency Department 

Standard Treatment Spaces per Staffed Inpatient Hospital Bed, Last Day 

of Fiscal Year 2001:



Table 36: Indicators of Crowding, by Emergency Department Admissions 

per Staffed Inpatient Bed, Fiscal Year 2001:



Table 37: Indicators of Crowding, by Hospital Ownership:



Table 38: Indicators of Crowding, by Trauma Center Status:



Table 39: Indicators of Crowding, by Teaching Status:



Table 40: Indicators of Crowding, by Select Payer Sources for Emergency 

Department Visits, Fiscal Year 2001:



Table 41: Hospitals Applying for Regulatory Approval to Increase 

Licensed Beds, since Start of Fiscal Year 2001:



Table 42: Types of Beds Requested since Start of Fiscal Year 2001:



Table 43: Average Proportion of Emergency Visits Covered by Medicare, 

Medicaid and SCHIP, and Self-Pay, Fiscal 

Year 2001:



Figures:



Figure 1: Hospitals by Percentage of Time on Diversion, Fiscal 

Year 2001:



Figure 2: Percentage of Hospitals on Diversion More than 

10 Percent of the Time, by MSA, Fiscal Year 2001:



Figure 3: Hospitals by Percentage of Patients Boarding 2 Hours or More 

and Average Number of Hours Boarding, Past 

12 Months:



Figure 4: Percentage of Hospitals Boarding More than Half of Patients 

for an Average of 8 Hours or More, by MSA:



Figure 5: Hospitals by Percentage of Patients Who Left Before a Medical 

Evaluation, Fiscal Year 2001:



Figure 6: Percentage of Hospitals with at Least 5 Percent of Patients 

Leaving before a Medical Evaluation, by MSA, Fiscal Year 2001:



Figure 7: Conditions Hospitals Reported as Contributing to Diversion, 

Fiscal Year 2001:



Figure 8: Conditions Hospitals Reported as Contributing to Boarding 

Patients in the Past 12 Months:



Figure 9: Primary Payer Source of Routine and Emergency Department 

Admissions, 2000:



Abbreviations:



AHRQ: Agency for Healthcare Research and Quality

CCU: critical care unit

CAT: computed axial tomography

CT: computed tomography

DRG: diagnosis related group

EMS: emergency medical services

HHS: Department of Health and Human Services

ICU: intensive care unit

MRI: magnetic resonance imaging

MSA: metropolitan statistical area

SCHIP: State Children’s Health Insurance Program:



This is a work of the U.S. Government and is not subject to copyright 

protection in the United States. It may be reproduced and distributed 

in its entirety without further permission from GAO. It may contain 

copyrighted graphics, images or other materials. Permission from the 

copyright holder may be necessary should you wish to reproduce 

copyrighted materials separately from GAO’s product.



United States General Accounting Office:



Washington, DC 20548:



March 14, 2003:



The Honorable Max Baucus

Ranking Minority Member

Committee on Finance

United States Senate:



Dear Senator Baucus:



Open 24 hours a day, 7 days a week, hospital emergency departments are 

a major part of the nation’s health care safety net. Emergency 

departments report being under increasing pressure, with the number of 

visits increasing about 14 percent from an estimated 95 million in 1997 

to an estimated 108 million in 2000, and the number of hospitals with 

emergency departments decreasing by about 2 percent. This pressure has 

led to reports of crowding. For example, considerable attention has 

been given to reports that emergency departments request that 

ambulances that would normally bring patients to their hospitals go 

instead to other hospitals that are presumably less crowded--a practice 

known as diversion. Crowded conditions in emergency departments can 

also lead to long waits for care, which can prolong pain and suffering.



There are no standard measures of the extent to which emergency 

departments are experiencing crowded conditions--hospital officials 

say “they know it when they see it.” However, there are several 

indicators that, according to experts, point to situations in which 

crowding is likely occurring. One indicator is the number of hours a 

hospital is on diversion status. A second indicator is the proportion 

of patients and the length of time patients “board” or remain in the 

emergency department--and therefore tie up space and staff resources--

after a decision has been made to admit them as inpatients or transfer 

them to other facilities rather than releasing them. Finally, a third 

indicator is the proportion of patients who leave the emergency 

department before receiving a medical evaluation, generally because 

they tire of waiting.



While considerable attention has been focused on this topic, much of it 

has centered on anecdotal reports or on data from a limited number of 

communities or emergency departments. You asked us to determine if data 

could be assembled from a broader, more national scope in order to 

provide more perspective on the issue. We conducted a review that 

encompassed hospitals located in the nation’s metropolitan statistical 

areas (MSA).[Footnote 1] We excluded nonmetropolitan areas because 

available information and contacts with rural health organizations 

indicated that emergency department crowding is not a major problem in 

these areas. Our work addressed the following questions:



* To what extent are hospitals in MSAs experiencing crowded conditions 

in their emergency departments, and is crowding more severe in some 

types of MSAs than in others?



* What factors contribute to emergency department crowding?



* What actions have hospitals and communities taken to address 

crowding?



To conduct this work, we sent a mail questionnaire to all community 

hospitals located in MSAs that reported having emergency departments in 

2000--more than 2,000 hospitals in all,[Footnote 2] of which about 74 

percent responded. The survey collected information related to three 

indicators of crowding: (1) the number of hours on diversion, (2) the 

percentage of patients who were boarding in the emergency department 

for 2 hours or more and the average number of hours boarded, and (3) 

the proportion of patients who left before a medical 

evaluation.[Footnote 3] In analyzing these responses, we weighted 

responses to adjust for a lower response rate from investor-owned (for-

profit) hospitals to provide estimates for the universe of hospitals. 

To examine which factors contributed to crowding, we analyzed 

information provided by the surveyed hospitals and other data on 

hospital and MSA characteristics. To provide information on actions 

taken by hospitals and communities to address crowding, as well as 

emergency medical services (EMS) systems and diversion at the community 

level, we conducted site visits in six locations where problems had 

been reported regarding crowded emergency departments--Atlanta, 

Boston, Cleveland, Los Angeles, Miami, and Phoenix. We selected these 

sites because they varied in geographic location, proportion of people 

without health insurance, population, and recent population growth. In 

these locations, we interviewed EMS officials, professional 

associations, and hospital officials, and we observed emergency 

departments in 24 hospitals. We supplemented this work with analysis of 

existing national data and reviews of relevant studies. We also 

interviewed persons knowledgeable about the issues, including health 

services researchers; representatives from hospital associations, 

provider associations, and emergency medical associations; and federal, 

state, and local health officials. Appendix I explains our methodology 

in more detail. We conducted our work from July 2001 through February 

2003 in accordance with generally accepted government auditing 

standards.



Results in Brief:



Although most emergency departments across the country experienced some 

degree of crowding, the problem is much more pronounced in some 

hospitals and areas than in others. For example, while our nationwide 

survey of hospitals found that about two of every three emergency 

departments reported going on diversion at some point in fiscal year 

2001, a much smaller portion--nearly 1 of every 10 hospitals--was on 

diversion more than 20 percent of the time. In general, hospitals that 

reported the most problems with crowding were in the largest MSAs, MSAs 

with high population growth, and MSAs with higher percentages of people 

without health insurance. For example, hospitals in MSAs with 

populations of 2.5 million or more had a median of about 162 hours of 

diversion in 2001, compared with about 9 hours for hospitals in MSAs 

with populations of less than 1 million. Similarly, hospitals in MSAs 

with higher percentages of uninsured people had almost twice as high of 

a median percentage of patients leaving the emergency department prior 

to a medical evaluation as those in MSAs with fewer uninsured.



Crowding is a complex issue and no single factor tends to explain why 

crowding occurs. However, one key factor contributing to crowding at 

many hospitals involves the inability to move patients out of emergency 

departments and into inpatient beds when these patients must be 

admitted to the hospital rather than released after treatment. With no 

inpatient beds available for them, these patients then have to board in 

the emergency department, reducing the emergency department’s ability 

to see additional patients. In particular, hospitals that we surveyed 

and that we visited cited the inability to move emergency patients into 

critical care or telemetry (instrument-monitored) beds as contributing 

to crowding. Our analysis of survey data found that indicators of 

emergency department crowding were higher at hospitals in MSAs with 

more demand for inpatient hospital beds and at hospitals with higher 

occupancy. Reasons given by hospital officials and researchers we 

interviewed for not always having enough inpatient beds to meet the 

demand from emergency patients included (1) economic incentives to 

staff only the number of inpatient beds that will nearly always be 

full--a practice that limits a hospital’s ability to meet periodic 

spikes in demand, and (2) competition for available beds with scheduled 

admissions such as surgery patients. Other factors cited by researchers 

and hospital officials as contributing to crowding included closures of 

nearby hospitals and inadequate availability of physicians and other 

providers in the community.



At the six sites we visited, hospitals and communities reported a 

variety of actions to address crowding. At hospitals, these actions 

generally fell into two categories--increasing capacity and increasing 

efficiency. For example, two-thirds of the hospitals we visited had 

expanded or planned to expand their emergency departments. Officials at 

some of the hospitals we visited also reported holding meetings of key 

hospital staff members to quickly identify and make available inpatient 

beds to minimize boarding in the emergency department. At the community 

level, actions included developing standard definitions and policies 

for when hospitals can go on diversion and improving communication 

among hospitals and EMS providers. However, the extent to which these 

actions address crowding is unknown. Hospital officials and others 

involved in these efforts said that their actions have helped better 

manage the problem of crowded emergency departments, but have not 

solved it. Some efforts are under way to better measure and track 

crowding at individual hospitals, which may facilitate future 

evaluation of efforts to address crowding.



Representatives from the American College of Emergency Physicians and 

the American Hospital Association and an independent reviewer provided 

comments on a draft of this report, which we incorporated as 

appropriate.



Background:



In 2000, about 3,900 nonfederal, general medical hospitals nationwide 

reported providing emergency care in emergency departments. Of these, 

just over half were located in MSAs. From 1997 through 2000, while the 

number of emergency department visits increased about 14 percent, the 

number of hospitals with emergency departments decreased by about 2 

percent. The result was that the average number of visits per emergency 

department increased by 16 percent.[Footnote 4] Many hospitals expanded 

the physical space and number of treatment spaces in their emergency 

departments during that time.



Recent reports have raised concern that many of the nation’s emergency 

departments are experiencing high demand and crowded conditions. An 

April 2002 report for the American Hospital Association, while limited 

in scope and the proportion of hospitals responding, found that 

officials at many hospitals in urban areas described their emergency 

departments as operating at or above capacity.[Footnote 5] While there 

are no comprehensive studies on the consequences of crowded conditions, 

health care researchers and clinicians report that crowding has 

multiple effects, including prolonged pain and suffering for some 

patients, long patient waits, increased transport times for ambulance 

patients, inconvenience and dissatisfaction for the patients and their 

families, and increased frustration among medical staff.[Footnote 6] In 

addition to delays in treatment, some emergency department directors 

have reported that patient care was compromised and patients 

experienced poor outcomes as a result of crowded conditions in 

emergency departments.[Footnote 7]



Because the medical conditions of patients who come to the emergency 

department can range from mild injuries such as ankle sprains to 

serious traumas such as from automobile accidents--and can also include 

patients with chronic conditions such as asthma or diabetes--the space, 

equipment, and medical personnel resources required to treat patients 

vary. As a result, there are no specific criteria, such as a ratio of 

patients to staff, to define when an emergency department is too 

crowded and its providers are overloaded. Rather, emergency department 

administrators and physicians say “they know it when they see it.” In 

the absence of specific criteria to define when an emergency department 

is crowded, health care researchers suggest using several available 

indicators to point to crowded conditions. Based on our review of 

studies and discussions with experts, we chose three indicators of 

emergency department crowding. As shown in table 1, all three are 

useful indicators but all three also have limitations.



Table 1: Indicators of Emergency Department Crowding:



Indicator: Diversion; Definition: Hospitals request that ambulances 

bypass their emergency departments and transport patients that would 

have been otherwise taken to those emergency departments to other 

medical facilities.; Usefulness: For hospitals that can go on 

diversion, it is an indicator of how often these emergency departments 

believe that they cannot safely handle additional ambulance patients.; 

Limitations: The number of hours on diversion is a potentially 

imprecise measure of crowding because whether a hospital can go on 

diversion and the circumstances under which it can do so vary from 

location to location, according to both individual hospital policy and 

communitywide guidelines or rules.; Our measure of this indicator: Our 

survey asked if hospitals ever went on diversion in fiscal year 2001[A] 

and the total number of hours they were on diversion for any reason in 

fiscal year 2001. In the six sites we visited, we collected available 

data on diversion for 2000, 2001, and 2002.



Indicator: Boarding; Definition: The decision to admit or transfer an 

emergency patient has been made, and the patient waits to leave the 

emergency department for a minimum period.; Usefulness: Patients 

boarding in the emergency department take space and resources that 

could be used to treat other emergency department patients. Boarding is 

an indicator that an emergency department’s capacity to treat 

additional patients is diminished.; Limitations: Boarding can be used 

to indicate a hospital’s ability to move a patient out of the emergency 

department and into an inpatient bed; however, it is possible for an 

emergency department to be boarding several patients while also having 

available treatment spaces to see additional patients that come to the 

emergency department.; Our measure of this indicator: Our survey asked 

if hospitals ever boarded patients for 2 hours or more. For those that 

did, we asked for the percentage of patients boarded for 2 hours or 

more and the average number of hours patients boarded in the past 12 

months.



Indicator: Left before a medical evaluation; Definition: The number of 

patients who left after triage[B] but before a medical evaluation as a 

percentage of emergency department visits.; Usefulness: The most common 

reason for patients leaving the emergency department before being 

treated is excessive waiting time, which can occur when an emergency 

department is crowded and unable to treat the patients waiting to be 

seen in a reasonable amount of time.; Limitations: Since emergency 

department staff triage patients, those with nonemergent conditions 

generally wait the longest and may be most likely to tire of waiting 

and leave before receiving a medical evaluation.; Our measure of this 

indicator: Based on survey data, we calculated the percentage of 

patients who left after triage but before a medical evaluation for 

fiscal year 2001.



Source: GAO.



[A] We asked hospitals to provide data for their fiscal year 2001.



[B] The process of sorting patients based on their need for immediate 

medical treatment.



[End of table]



Crowding Indicator 1: Diversion:



One indicator of a crowded emergency department is the number of hours 

a hospital is on diversionary status. Under federal law, all hospitals 

that participate in Medicare are required to screen--and if an 

emergency medical condition is present, stabilize--any patient who 

comes to the emergency department, regardless of the individual’s 

ability to pay.[Footnote 8] Under certain circumstances where a 

hospital lacks staffing or facilities to accept additional emergency 

patients, the hospital may place itself on “diversionary status” and 

direct en route ambulances to divert to another hospital.[Footnote 9] 

In general, hospitals ask EMS providers to divert ambulances to other 

medical facilities because their emergency department staff are 

occupied and unable to promptly care for new arrivals or specific 

services within the hospitals, such as the intensive care units, are 

filled and unable to accommodate the specialized needs of new ambulance 

arrivals.



While on diversion, hospitals must still treat any patients who arrive 

by ambulance, and in some cases, local community protocols allow 

ambulances to go to a hospital that is on diversion when the patient 

asks to go to that hospital or if the patient needs immediate medical 

treatment. In addition, even while on diversion, the emergency 

department is still required to screen and treat nonambulance patients-

-those patients who walk in or otherwise arrive at the hospital--and 

these patients make up the vast majority of visits to the emergency 

department. The Department of Health and Human Service’s (HHS) National 

Center for Health Statistics estimates that in 2000 about 14 percent of 

emergency department visits were made by patients who arrived by 

ambulance, while 78 percent of visits were made by patients who arrived 

at the emergency department by “walking in.” For the remaining visits, 

the patients were brought in by the police or social services (1.5 

percent), or the mode of arrival was unknown (6.3 percent).[Footnote 

10]



As a measure of crowding, diversion has limitations in that some 

hospitals, even when crowded, do not have the option to divert 

ambulances due to state or local regulations, because there are no 

other medical facilities nearby, or because of individual hospital 

policies. Hospital practices may vary regarding the threshold at which 

a hospital goes on diversion. Local community or hospital policies may 

also differ regarding the length of time a hospital may remain on 

diversion. (See app. II for the local community policies for the six 

sites we visited). However, for those hospitals that can go on 

diversion, it is an indicator of how often these emergency departments 

believe they can no longer handle additional ambulance patients.



Crowding Indicator 2: Boarding:



A second indicator suggested by health care researchers is the number 

of patients who are “boarding” in the emergency department. These 

patients remain in the emergency department after the decision has been 

made to admit them to the hospital or transfer them to another 

facility. Many factors can contribute to the length of time a patient 

is boarded in the emergency department, such as inpatient bed 

availability, staffing levels, and the complexity of a patient’s 

condition. Regardless of the reason, while waiting for an inpatient bed 

or transfer, these patients still require care and take up treatment 

space, equipment, and staff time in the emergency department, shrinking 

the department’s resources available to treat other emergency patients. 

A limitation of using boarding as an indicator is that many hospitals 

do not collect this information regularly and can only estimate how 

often and how long patients board in their emergency departments. In 

addition, it is possible that emergency departments board patients 

while also having available treatment spaces to see additional 

patients.



Crowding Indicator 3: Leaving before a Medical Evaluation:



Finally, the proportion of patients who leave after triage but before 

receiving a medical evaluation is another indicator suggested by health 

care researchers that could indicate a crowded emergency department. 

Long waits in the emergency department can delay needed care and 

contribute to an increase in the number of people who choose to leave 

the emergency department before receiving a medical evaluation. A 

limitation to this indicator is that, because emergency department 

staff triage patients, those with nonemergent conditions generally wait 

the longest and may be most likely to tire of waiting and leave before 

a medical evaluation. However, relatively mild conditions could 

potentially become more serious if patients do not receive needed 

medical care because they leave the emergency department before being 

evaluated and treated. A study of the consequences of leaving the 

emergency department prior to a medical evaluation at one public 

hospital found that 46 percent of those who left were judged to need 

immediate medical attention, and 11 percent who left were hospitalized 

within the next week.[Footnote 11]



Emergency Department Crowding Is More Pronounced in Some Hospitals and 

Certain Types of Communities:



Although most emergency departments across the country reported some 

degree of crowding on one or more of the three indicators, the problem 

is much more pronounced in some hospitals than in others. In addition, 

hospitals in the largest metropolitan areas (those with populations of 

2.5 million or more), communities with high population growth, and 

communities with above average percentages of people without health 

insurance had higher levels of crowding.



Indicators Show Varying Degrees of Crowding Nationwide:



Analysis of responses to our nationwide survey showed substantial 

variation in the degree of crowding reported across all three 

indicators--diversion, boarding, and patients leaving before a medical 

evaluation. Hospitals ranged from little or no crowding to crowding 

that persisted for a substantial part of the time.



Diversion. In total, we estimate that about 2 of every 3 of the 

hospitals in our survey universe went on diversion at least once during 

fiscal year 2001. We estimate that about 2 in every 10 of these 

hospitals were on diversion for more than 10 percent of the time, and 

about 1 in every 10 was on diversion for more than 20 percent of the 

time--or about 5 hours per day.[Footnote 12] Figure 1 shows the 

variation in the amount of diversion reported by hospitals in MSAs.



Figure 1: Hospitals by Percentage of Time on Diversion, Fiscal Year 

2001:



[See PDF for image]



Note: Responses were weighted to provide estimates for the entire 

universe of 2,021 hospitals. Data were missing for about 4 percent of 

hospitals.



[End of figure]



Diversion varies greatly by MSA. Figure 2 shows each MSA and the share 

of hospitals within the MSA that reported being on diversion more than 

10 percent of the time--or about 2.4 hours or more per day--in fiscal 

year 2001. Of the 248 MSAs for which data were available,[Footnote 13] 

171 (69 percent) had no hospitals reporting being on diversion more 

than 10 percent of the time. By contrast, 53 MSAs (21 percent) had at 

least one-quarter of responding hospitals on diversion for more than 10 

percent of the time.



Figure 2: Percentage of Hospitals on Diversion More than 10 Percent of 

the Time, by MSA, Fiscal Year 2001:



[See PDF for image]



Note: Percentage of hospitals reflects those hospitals that responded 

to the survey; responses were not weighted to represent all hospitals 

in the MSA.



[A] MSAs with a response rate of 50 percent or less or MSAs with 50 

percent or more of data missing for responding hospitals. In 12 MSAs, 

no hospitals responded; these MSAs were excluded from the map.



[End of figure]



Boarding. Boarding patients for 2 hours or more in the emergency 

department while waiting for an inpatient bed or transfer occurred to 

some extent at an estimated 9 of every 10 hospitals. As part of our 

survey, we examined what percentage of emergency patients who boarded 

spent 2 hours or more in boarding status and the average number of 

hours patients boarded.[Footnote 14] As figure 3 shows, while many 

hospitals reported boarding less than 25 percent of boarded patients 

for 2 hours or more in the past 12 months, about one-third of them 

reported boarding 75 percent or more of their boarded patients for that 

long. About 1 in every 5 hospitals reported an average boarding time in 

their emergency departments of 8 hours or more.



Figure 3: Hospitals by Percentage of Patients Boarding 2 Hours or More 

and Average Number of Hours Boarding, Past 12 Months:



[See PDF for image]



Note: Responses were weighted to provide estimates for the entire 

universe of 2,021 hospitals. Data were missing for about 5 percent of 

hospitals on the percentage of patients boarded 2 hours or more and for 

about 11 percent of hospitals on the average number of hours boarded.



[End of figure]



Boarding varies greatly by MSA. Figure 4 shows each MSA and the extent 

to which responding hospitals within the MSA reported that of those 

patients who boarded in the past 12 months, at least half spent 2 hours 

or more in boarding status, and the average boarding time was 8 hours 

or more. Of the 206 MSAs for which data were available on the 

percentage of patients boarded and the average number of hours 

boarded,[Footnote 15] 112 MSAs (54 percent) had no hospitals reporting 

that they met these criteria. In contrast, 52 of the 206 MSAs (25 

percent) had at least one-fourth of responding hospitals reporting that 

they boarded at least half of their patients for 2 hours or more and 

had an average boarding time of at least 8 hours.



Figure 4: Percentage of Hospitals Boarding More than Half of Patients 

for an Average of 8 Hours or More, by MSA:



[See PDF for image]



Note: Percentage of hospitals reflects those hospitals that responded 

to the survey; responses were not weighted to represent all hospitals 

in the MSA. Boarding data were for the past 12 months.



[A] MSAs with a response rate of 50 percent or less or MSAs with 50 

percent or more of data missing for responding hospitals. In 12 MSAs, 

no hospitals responded; these MSAs were excluded from the map.



[End of figure]



Patients Leaving before a Medical Evaluation. From our nationwide 

survey of hospitals, we estimate that the median percentage of patients 

who left after triage but before a medical evaluation in fiscal year 

2001 was 1.4 percent. We estimate that about 39 percent of hospitals 

had from 1 to 3 percent of patients who left before medical a 

evaluation in fiscal year 2001 while about 7 percent of hospitals 

reported that 5 percent or more of emergency department patients left 

before a medical evaluation (see fig. 5).[Footnote 16]



Figure 5: Hospitals by Percentage of Patients Who Left Before a Medical 

Evaluation, Fiscal Year 2001:



[See PDF for image]



Note: Responses were weighted to provide estimates for the entire 

universe of 2,021 hospitals. Data were missing for about 7 percent of 

hospitals.



[End of figure]



Figure 6 shows each MSA and the extent to which hospitals within the 

MSA reported at least 5 percent of patients leaving before a medical 

evaluation. Of the 243 MSAs for which data were available on the 

percentage of patients who left before a medical evaluation,[Footnote 

17] 183 MSAs (75 percent) had no hospitals reporting that they met 

these criteria. In contrast, 31 of the 243 MSAs (13 percent) had at 

least one-fourth of responding hospitals reporting that at least 5 

percent of patients left before a medical evaluation in fiscal year 

2001.



Figure 6: Percentage of Hospitals with at Least 5 Percent of Patients 

Leaving before a Medical Evaluation, by MSA, Fiscal Year 2001:



[See PDF for image]



Note: Percentage of hospitals reflects those hospitals that responded 

to the survey; responses were not weighted to represent all hospitals 

in the MSA.



[A] MSAs with a response rate of 50 percent or less or MSAs with 50 

percent or more of data missing for responding hospitals. In 12 MSAs, 

no hospitals responded; these MSAs were excluded from the map.



[End of figure]



Crowding More Pronounced in Certain Types of Communities:



We analyzed our three crowding indicators across different MSA 

characteristics, including population, population growth, and level of 

uninsurance. We found all three characteristics were associated with 

reported levels of crowding.[Footnote 18]



MSAs with Larger Populations:



Hospitals in MSAs of 2.5 million or more people reported higher levels 

of all three indicators--diversion, boarding, and patients leaving 

before a medical evaluation--than hospitals in MSAs of less than 1 

million people (see table 2). In these larger areas, hospitals had a 

median of about 162 hours of diversion in 2001 compared with 9 hours 

for hospitals in areas with a population of less than 1 million. 

Similarly, the median percentage of patients boarding 2 hours or more 

was more than twice as high in large MSAs--48 percent versus 23 

percent.[Footnote 19] The median percentage of patients who left before 

a medical evaluation was also higher, though not as dramatically as for 

the two other indicators.



Table 2: Indicators of Crowding, by Population of MSA:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; MSA population: 2.5 million 

or more: 162 hours[A]; MSA population: 1 million or more but less than 

2.5 million: 84 hours[A]; MSA population: Less than 

1 million: 9 hours.



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 12 months; MSA population: 2.5 million 

or more: 48%[A]; MSA population: 1 million or more but less than 

2.5 million: 39%[A]; MSA population: Less than 

1 million: 23%.



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; MSA population: 2.5 million 

or more: 1.6%[A]; MSA population: 1 million or more but less than 

2.5 million: 1.4%[A,B]; MSA population: Less than 

1 million: 1.3%[B].



Source: GAO survey of hospitals, 2002, and U.S. Census Bureau.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] No statistically significant difference between the medians for 

hospitals in MSAs of 1 million or more but less than 2.5 million and 

hospitals in MSAs with populations of 2.5 million or more.



[B] No statistically significant difference between the medians for 

hospitals in MSAs of less than 1 million and hospitals in MSAs of 1 

million or more but less than 2.5 million.



[End of table]



Our site visits show that crowding indicators vary not only across MSAs 

but also between hospitals within MSAs. Four of the six locations we 

visited (Atlanta, Los Angeles, Boston, and Phoenix) were in MSAs with 

populations of over 2.5 million and we found variation among hospitals 

within these communities. For example, the 10 major Boston hospitals 

were on diversion for an average of 322 hours in 2001. However, 2 of 

the 10 hospitals accounted for nearly half of the diversion hours for 

the 10 hospitals, averaging nearly 800 hours of diversion each.



MSAs with High Population Growth:



Hospitals in communities with high population growth from 1996 through 

2000 reported higher levels of diversion and patients leaving before a 

medical evaluation compared to hospitals in communities with lower 

population growth (see table 3). The median number of hours of 

diversion in fiscal year 2001 for hospitals in MSAs with a high 

percentage population growth was about five times that for hospitals in 

MSAs with lower percentage population growth. Similarly, the median 

percentage of patients who left before a medical evaluation was 

significantly higher for hospitals in MSAs with high population growth-

-1.7 percent--than for those in MSAs with low population growth--1.0 

percent. In addition, of hospitals that reported at least 5 percent of 

patients leaving before a medical evaluation in 2001, 31 percent were 

in communities with high population growth compared to 15 percent in 

communities with low population growth.



Table 3: Indicators of Crowding, by Population Growth of MSAs:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; MSA population growth, 1996-2000: Top 25 percent[A]: 50 hours; 

MSA population growth, 1996-2000: Bottom 25 percent[B]: 10 hours.



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 12 months; MSA population growth, 1996-2000: Top 25 

percent[A]: 33%[C]; MSA population growth, 1996-2000: Bottom 25 

percent[B]: 22%[C].



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; MSA population growth, 1996-2000: Top 25 

percent[A]: 1.7%; MSA population growth, 1996-2000: Bottom 25 

percent[B]: 1.0%.



Source: GAO survey of hospitals, 2002, and U.S. Census Bureau.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] Hospitals in the top 25 percent in terms of MSA population growth 

were located in MSAs with a population increase of about 8.4 percent or 

more.



[B] Hospitals in the bottom 25 percent in terms of MSA population 

growth were located in MSAs with a population increase of less than 

about 2.9 percent.



[C] No statistically significant difference between the medians for 

hospitals in the top and bottom quartiles.



[End of table]



Two of the six locations we visited, Atlanta and Phoenix, were in MSAs 

with high population growth from 1996 to 2000--16 percent and 18 

percent growth, respectively. Diversion hours varied among hospitals in 

these communities. For example, in Phoenix, 5 of the 28 hospitals in 

the region made up about 42 percent of the region’s diversion hours in 

2001. Two of these 5 hospitals with high rates of diversion were in the 

city’s central sector. Hospitals in this sector were on diversion an 

average of 10 percent of the time in 2001. By contrast, hospitals in 

the region’s northeast sector, a more suburban area, had the lowest 

average rate of diversion--an average of 3 percent of the time.



MSAs with Higher Levels of Uninsurance:



Hospitals in communities with a higher percentage of people without 

health insurance reported higher levels of diversion and patients 

leaving before a medical evaluation (see table 4). For example, 

hospitals in MSAs where the percentage of uninsured people was above 

average reported having almost twice as many patients leave the 

emergency department prior to a medical evaluation than those in MSAs 

where the percentage of uninsured was below average. Our analysis of 

other national data indicate that waiting times, which are reported to 

be the primary reason patients leave the emergency department before a 

medical evaluation, were longer in communities with more uninsured 

people. For example, in 2000, waiting times for nonemergent visits 

averaged about 25 minutes longer in communities with high levels of 

uninsured people than in communities with low levels of uninsured 

people (90 minutes versus 65 minutes).[Footnote 20]



Table 4: Indicators of Crowding, by Percentage of MSA Population 

without Health Insurance:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Level of uninsurance in the MSA: Significantly above the average 

level of uninsurance: 228 hours; Level of uninsurance in the MSA: 

Significantly below the average level of uninsurance: 72 hours.



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 12 months; Level of uninsurance in the MSA: Significantly 

above the average level of uninsurance: 42%[A]; Level of uninsurance in 

the MSA: Significantly below the average level of uninsurance: 49%[A].



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Level of uninsurance in the MSA: Significantly 

above the average level of uninsurance: 2.2%; Level of uninsurance in 

the MSA: Significantly below the average level of uninsurance: 1.2%.



Source: GAO survey of hospitals, 2002, and UCLA Center for Health 

Policy Research.



Notes: GAO analysis of survey data and UCLA Center for Health Policy 

Research analysis of uninsurance rates for 96 large MSAs (compared to 

the average for those MSAs) based on the 2000 and 2001 Current 

Population Survey. Analysis was limited to hospitals in these 96 large 

MSAs. Responses were weighted to provide estimates for the universe of 

hospitals.



[A] No statistically significant difference between the median 

percentages for areas with above and below average levels of 

uninsurance.



[End of table]



Of the six sites we visited, three (Los Angeles, Phoenix, and Miami) 

were MSAs with significantly higher percentages of people without 

health insurance. The crowding indicators varied among hospitals in 

these MSAs with high levels of uninsurance. For example, the number of 

hours on diversion in 2001 for hospitals in the Los Angeles MSA ranged 

from no diversion at four hospitals to 6,186 hours--about 71 percent of 

the time--at another hospital.[Footnote 21]



Crowding More Pronounced in Certain Types of Hospitals:



We also analyzed differences across a wide range of hospital 

characteristics, including the number of staffed beds; hospital 

ownership; teaching status; trauma center status; and the proportions 

of emergency department visits covered by Medicare, Medicaid or the 

State Children’s Health Insurance Program (SCHIP), or self-pay as the 

payer source. All three indicators of crowding were significantly 

higher in hospitals with more staffed beds and at teaching hospitals, 

while the median numbers of hours on diversion were higher at hospitals 

designated as certified trauma centers and at hospitals with fewer 

patients covered by Medicare. In addition, we found that the median 

proportion of patients who left before a medical evaluation was 

significantly higher in public hospitals than private, not-for-profit 

hospitals, and in hospitals with more emergency department visits 

covered by Medicaid and SCHIP or more patients who were self-pay 

patients.[Footnote 22] (See app. III for additional information on the 

indicators of crowding by select hospital characteristics).



Availability of Inpatient Beds for Emergency Patients Cited as a Key 

Factor Contributing to Crowding, but Other Factors Also Contribute:



No single factor stands out as the reason why crowding occurs. Rather, 

a number of factors, including many outside the emergency department, 

are associated with crowding. In both the opinion of hospitals we 

surveyed and of hospital officials we interviewed, the factor most 

commonly associated with crowding was the inability to transfer 

emergency patients to inpatient beds once decisions had been made to 

admit them as hospital patients rather than to release them after 

treatment. In looking at why hospitals did not have the capacity to 

always meet the demand for inpatient beds from emergency patients, 

hospital officials, researchers, and others pointed to (1) financial 

pressures leading to limited hospital capacity to meet periodic spikes 

in demand for inpatient beds and 

(2) competition between admissions from the emergency department and 

scheduled admissions such as surgery patients, who are generally 

considered to be more profitable. Other factors cited as contributing 

to crowding include closures of nearby hospitals or availability of 

physicians and other providers in the community.



Lack of Available Inpatient Beds for Emergency Patients the Most 

Commonly Cited Factor:



The inability to transfer emergency patients to inpatient beds was the 

condition that surveyed hospitals reported most often as contributing 

to going on diversion and boarding patients. Even when treatment spaces 

are available in the emergency department, hospitals may go on 

diversion for patients who will likely need instrument-monitored beds 

or critical care beds because these types of beds are full. As figure 7 

shows, the most common types of beds that were unavailable were 

intensive care unit (ICU) or critical care unit (CCU) beds, followed by 

instrument-monitored (telemetry) beds. More than three-fourths of 

hospitals that went on diversion reported that the lack of ICU/CCU beds 

contributed to diversion to a moderate, great, or very great extent.



Figure 7: Conditions Hospitals Reported as Contributing to Diversion, 

Fiscal Year 2001:



[See PDF for image]



[A] Responses were weighted to provide estimates for the entire 
universe 

of hospitals. Percentages are based on an estimated 1,389 hospitals 

going on diversion in fiscal year 2001.



[End of figure]



Similarly, lack of inpatient beds was the dominant reason given for the 

need to board patients in the emergency room (see fig. 8). Of hospitals 

that boarded patients for 2 hours or more in the past 12 months, about 

80 percent cited the lack of telemetry or critical care beds as 

contributing to boarding to a moderate, great, or very great extent.



Figure 8: Conditions Hospitals Reported as Contributing to Boarding 

Patients in the Past 12 Months:



[See PDF for image]



[A] Responses were weighted to provide estimates for the entire 
universe 

of hospitals. Percentages are based on an estimated 1,822 hospitals 

boarding patients for 2 or more hours in the past 12 months.



[End of figure]



Our analysis of data collected in our survey generally corroborates 

that a lack of inpatient beds plays a major role in contributing to 

emergency department crowding. We found that those hospitals in 

communities with higher demand for inpatient beds--as measured by 

admissions per inpatient bed--had higher indicators of crowding. As 

table 5 shows, hospitals that rank in the top 25 percent in terms of 

admissions per bed in the MSA had both significantly higher numbers of 

diversion hours and proportions of patients boarding 2 hours or more 

than hospitals in the bottom 25 percent of admissions per bed. For 

example, hospitals in the top 25 percent reported a median of 170 hours 

on diversion in fiscal year 2001, compared with a median of 12 hours 

for hospitals in the lowest 25 percent.



Table 5: Indicators of Crowding, by Admissions per Bed in the MSA:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Admissions per bed[A]: Top 25 percent[B]: 170 hours; Admissions 

per bed[A]: Bottom 25 percent[C]: 12 hours.



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 12 months; Admissions per bed[A]: Top 25 percent[B]: 60%; 

Admissions per bed[A]: Bottom 25 percent[C]: 19%.



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Admissions per bed[A]: Top 25 percent[B]: 1.5%[D]; 

Admissions per bed[A]: Bottom 25 percent[C]: 1.2%[D].



Source: GAO survey of hospitals, 2002, and American Hospital 

Association Annual Survey Database, Fiscal Year 2000.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] Admissions per bed in short-term general medical and surgical 

community hospitals with emergency departments in the MSA based on data 

from the American Hospital Association Annual Survey Database, Fiscal 

Year 2000.



[B] Hospitals in the top 25 percent in terms of admissions per bed in 

the MSA were located in MSAs with more than 48.9 admissions per bed.



[C] Hospitals in the bottom 25 percent in terms of admissions per bed 

in the MSA were located in MSAs with fewer than 40.3 admissions per 

bed.



[D] No statistically significant difference in the medians for 

hospitals in the top and bottom quartiles.



[End of table]



Similarly, hospitals with more demand for inpatient beds from the 

emergency department--that is, a higher proportion of emergency visits 

resulting in hospital admission--also had higher indicators of 

crowding. As table 6 shows, the quarter of hospitals with the highest 

percentages--more than 19.7 percent--of emergency visits resulting in 

inpatient hospital admission reported more diversion and boarding than 

the quarter of hospitals with the smallest percentages--less than 11.8 

percent--of emergency visits resulting in admission.



Table 6: Indicators of Crowding, by Percentage of Emergency Visits 

Resulting in Hospital Inpatient Admissions, Fiscal Year 2001:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Percentage of emergency department patients admitted to the 

hospital: Top 25 percent[A]: 144 hours; Percentage of emergency 

department patients admitted to the hospital: Bottom 25 percent[B]: 4 

hours.



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 12 months; Percentage of emergency department patients 

admitted to the hospital: Top 25 percent[A]: 52%; Percentage of 

emergency department patients admitted to the hospital: Bottom 25 

percent[B]: 9%.



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Percentage of emergency department patients 

admitted to the hospital: Top 25 percent[A]: 1.6%[C]; Percentage of 

emergency department patients admitted to the hospital: Bottom 25 

percent[B]: 1.3%[C].



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] Hospitals in the top 25 percent admitted more than 19.7 percent of 

emergency visits.



[B] Hospitals in the bottom 25 percent admitted fewer than 11.8 percent 

of emergency visits.



[C] No statistically significant difference in the medians for 

hospitals in the top and bottom quartiles.



[End of table]



Finally, our analysis found that hospitals with more patients per bed-

-measured by the average occupancy in fiscal year 2001 as a percentage 

of the total number of staffed inpatient beds on the last day of the 

fiscal year--also had higher indicators of crowding in the emergency 

department (see table 7).[Footnote 23]



Table 7: Indicators of Crowding, by Average Occupancy as a Percentage 

of Staffed Inpatient Beds, Fiscal Year 2001:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Average occupancy as a percentage of staffed inpatient beds[A]: 

Top 25 percent[B]: 101 hours; Average occupancy as a percentage of 

staffed inpatient beds[A]: Bottom 25 percent[C]: 6 hours.



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 

12 months; Average occupancy as a percentage of staffed inpatient 

beds[A]: Top 25 percent[B]: 55%; Average occupancy as a percentage of 

staffed inpatient beds[A]: Bottom 25 percent[C]: 9%.



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Average occupancy as a percentage of staffed 

inpatient beds[A]: Top 25 percent[B]: 1.5%; Average occupancy as a 

percentage of staffed inpatient beds[A]: Bottom 25 percent[C]: 1.0%.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] Average daily census reported at midnight for fiscal year 2001 as a 

percentage of the total number of staffed beds reported as of the last 

day of fiscal year 2001. Excludes long-term care, labor and delivery, 

and postpartum beds.



[B] Hospitals in the top 25 percent had an average daily census of more 

than 80.8 percent of staffed inpatient beds.



[C] Hospitals in the bottom 25 percent had an average daily census of 

less than 57 percent of staffed inpatient beds.



[End of table]



The conclusion that the availability of inpatient beds contributes to 

crowding in emergency departments was reiterated at the hospitals we 

visited on our site visits. At 19 of the 24 hospitals we visited, 

hospital officials reported that the lack of inpatient beds and 

subsequent boarding of emergency patients was a key factor contributing 

to crowding. In addition, a 1-week survey conducted in Massachusetts 

found that hospitals’ occupancy rates were higher when hospitals were 

on diversion.[Footnote 24]



Several Reasons Cited for Hospitals Not Always Having Inpatient 

Capacity to Meet Demand for Beds from Emergency Patients:



When we examined why hospitals did not always have the inpatient 

capacity to meet the demand for beds from emergency patients, hospital 

administrators, researchers, and clinicians cited several reasons, 

including (1) financial incentives to control costs and maximize 

revenue by staffing inpatient beds at a point where they will nearly 

always be full--a practice that limits a hospital’s ability to meet 

periodic spikes in demand, and 

(2) competition between emergency department admissions and scheduled 

admissions for available beds.[Footnote 25]



Economic Factors Influence Hospitals’ Capability to Meet Periodic 

Spikes in Demand:



One reason reported for the lack of inpatient beds was the financial 

pressures hospitals face to staff inpatient beds at a level where they 

will nearly always be full. This practice limits a hospital’s ability 

to meet periodic spikes in demand. Hospital administrators, clinicians, 

and health care researchers report that changes in the hospital 

economic climate have contributed to this decline in “surge capacity.” 

For example, in a report prepared for the Massachusetts Health Policy 

Forum, one health policy researcher noted that the lower occupancy 

rates of the 1970s and 1980s became unacceptable in the 1990s when 

hospitals were increasingly driven by market-based factors. In a 

market-based system, successful hospitals run full, attract both 

elective and emergency patients, and are staffed closer to average 

demand than to the peaks.[Footnote 26]



Another factor sometimes cited that is related to insufficient bed 

capacity involves staffing. Officials at some hospitals we visited said 

that they did not staff more of the beds they already had or open new 

beds because they were concerned they would not be able to staff them 

or could not afford the cost of staffing them.[Footnote 27] These 

hospitals cited the costs and difficulties recruiting nurses, 

particularly the cost of hiring nurses from agencies that contract out 

nursing services. For example, officials at a Miami hospital we visited 

that staffed only about two-thirds of the beds for which it was 

licensed in 2001 said that they would lose money if they staffed more 

beds because of the cost of contract nurses.



Emergency Department Admissions Compete with Other Admissions:



For the inpatient beds that are available, many researchers and 

hospital officials we interviewed reported that hospitals often balance 

admissions from emergency departments with scheduled admissions for 

surgical procedures, which are generally considered more profitable. 

One reason that admissions from the emergency department are considered 

to be less profitable is because these admissions tend to be for 

medical conditions, such as pneumonia, rather than for those procedures 

that are considered more profitable. Available data from the Agency for 

Healthcare Research and Quality’s (AHRQ) Healthcare Cost and 

Utilization Project, Nationwide Inpatient Sample, show that of hospital 

admissions from the emergency department in 2000, most were for medical 

conditions (such as pneumonia and heart failure). Further, 19 of the 20 

most prevalent diagnosis related groups (DRG) for these admissions were 

for medical conditions. In contrast, half of the 20 most common DRGs 

for admissions that were not from the emergency department were for 

surgical procedures (such as orthopedic surgery and cardiac pacemaker 

implantation).[Footnote 28]



Many hospital officials and researchers also said that emergency 

department patients are less profitable because a larger proportion of 

emergency admissions are patients for whom the primary payer source is 

self-pay, which includes the uninsured, or Medicaid, which is generally 

considered to provide lower reimbursement. As shown in figure 9, 

available data from AHRQ’s Healthcare Cost and Utilization Project show 

that the proportion of admissions for uninsured patients or patients 

with Medicaid as the primary payer source was higher for admissions 

from the emergency department than for routine admissions in 2000. At 

the same time, the proportion of admissions with private insurance as 

the primary payer source was higher for routine admissions than for 

patients admitted from the emergency department. Because self-pay 

patients and those covered by Medicaid are viewed as providing lower 

reimbursement, hospital officials and health care researchers said that 

hospitals have a financial incentive to fill the limited number of 

available beds with scheduled admissions rather than emergency 

department admissions.



Figure 9: Primary Payer Source of Routine and Emergency Department 

Admissions, 2000:



[See PDF for image]



Note: This analysis is based on national estimates of discharges from 

nonfederal, short-term, general medical and surgical hospitals with 

emergency departments, and excludes neonatal and maternal discharges.



[End of figure]



In addition, some hospital officials reported that surgeons bring in 

business that generates revenues for the hospital and that hospitals 

may not want to cancel or reschedule elective surgeries--and disrupt 

their surgeons and patients--in order to make beds available for 

emergency department patients. This point was supported by our survey 

results--less than one-third of hospitals that went on diversion in 

fiscal year 2001 (29 percent) reported that they had canceled any 

elective procedures to minimize going on diversion.



Additional Factors Cited as Contributing to Crowding:



Hospital officials reported in both our survey and during our site 

visits that other factors contributed to crowding as well, including 

increased demand due to the closure of other hospitals and difficulties 

in accessing physicians and other medical providers in the community. 

For example, officials at one hospital we visited said that when two 

neighboring hospitals closed in 1999 and 2000, their hospital 

experienced a significant increase in emergency department visits and 

subsequent crowding. In addition, officials at some of the hospitals we 

visited said they thought that the availability of physicians and other 

services, such as psychiatric services, in their communities affected 

crowding in one or more instances. For example, in Cleveland, the 

county psychiatric mobile health unit recently stopped taking patients 

in the late evening and on weekends, increasing the amount of time the 

emergency department had to care for psychiatric patients during those 

times. One Cleveland hospital we visited reported that boarding times 

for patients awaiting assessment by this unit had increased for 

patients who arrived late at night.



Another factor that many hospital officials we interviewed and other 

experts reported as contributing to crowding was an increase in the 

amount of time clinicians need to spend with each emergency department 

patient or the amount of time the patient remains in the emergency 

department before a decision is made to admit, transfer, or release him 

or her. Emergency department physicians and researchers report that 

emergency patients are older, have more complex conditions, and have 

more treatment and tests provided in the emergency department than in 

prior years because the standard for admitting them to the hospital has 

been raised and medical practices have changed. For example, one 

emergency department administrator said that patients with asthma are 

now treated and monitored in the emergency department for several hours 

before a decision is made whether to admit them to inpatient beds. 

Similarly, with newer technology available, patients with chest pain 

may remain and be monitored in the emergency department for several 

hours before a decision is made whether to admit them. In addition, 

hospital officials reported that the time it takes to receive 

laboratory and radiology results creates delays in the emergency 

department. While available data from HHS’s National Center for Health 

Statistics indicate significant increases from 1992 through 2000 in the 

proportion of emergency department visits that were for illnesses 

instead of injuries and the proportion of visits in which computed 

axial tomography (CAT) scans and magnetic resonance imaging (MRI) 

screenings were conducted, no

national data are available showing whether the length of time 

emergency patients remain in the emergency department has changed over 

time. [Footnote 29],[Footnote 30]



Wide Range of Activities Under Way to Manage Crowding at Hospitals and 

in Communities, but Problems Persist:



At the six sites we visited, actions to address emergency department 

crowding had been taken at both the hospital and community levels. 

Steps taken by hospitals generally fell into two categories: (1) 

increasing capacity and (2) improving the efficiency with which 

patients are treated--and if necessary, moved to inpatient beds. At the 

community level, EMS agencies, health care associations, and public 

agencies were generally active to some degree in implementing 

communitywide policies and computerized diversion tracking systems to 

help direct the flow of ambulance traffic and keep hospital staff and 

EMS providers informed about which hospitals are on diversion. While 

hospital and community officials reported some positive results for 

their efforts, they generally described these efforts as attempts to 

manage crowding problems rather than to substantially reduce them. The 

effects of these efforts have not been widely studied, though several 

activities are now under way that may help facilitate future 

evaluations.



Hospitals Expand Capacity and Increase Efficiency:



To accommodate increasing demand, a number of hospitals in all six of 

the locations we visited reported having expanded or planning to expand 

their emergency department or hospital inpatient capacity in terms of 

space, staffing, and laboratory capability. For example, 16 of the 24 

hospitals we visited reported having expanded or planning to expand 

their emergency department treatment space. These expansion activities 

ranged from adding stretchers in the hallway to accommodate more 

emergency department patients to building new, larger emergency 

departments.[Footnote 31] Some hospitals added a unit--often referred 

to as a fast-track unit--to the emergency department that is staffed 

with appropriate personnel, such as nurse practitioners and physician 

assistants, to quickly treat nonurgent cases. In addition, officials at 

11 of the 24 hospitals we visited told us that their hospitals had 

expanded or would be expanding inpatient capacity or building new 

hospital facilities, a step that could make it easier to transfer 

patients who need to be admitted as inpatients.[Footnote 32] We found 

expansions or planned expansions at different types of hospitals, 

including not-for-profit, public, and for-profit hospitals. At some 

hospitals that had recently expanded their capacity, hospital officials 

reported that even though the expansion helped, they continue to 

experience very crowded conditions. Table 8 provides examples of the 

kinds of actions taken or planned at the hospitals we visited.



Table 8: Examples of Expansions of Emergency Departments or Inpatient 

Capacity at Hospitals GAO Visited:



Location: Atlanta; Actions taken by hospitals: One hospital has opened 

a new emergency department that increased the number of standard 

treatment spaces from 17 to 33, including a fast-track unit for 

treating nonurgent patients more quickly in the emergency department. 

The hospital also plans to staff 80 additional inpatient hospital 

beds.



Location: Cleveland; Actions taken by hospitals: One hospital increased 

the number of emergency department beds from 15 to 21 and added 12 ICU 

beds.



Location: Miami; Actions taken by hospitals: One hospital is expanding 

its emergency department to double the number of treatment spaces from 

24 to 48 beds. In addition, the hospital recently added (1) a 40-bed 

temporary care unit to handle patients boarding in the emergency 

department and (2) adult and pediatric fast-track units in the 

emergency department to treat nonurgent patients.



Location: Phoenix; Actions taken by hospitals: One hospital added 

additional physician coverage during the busiest time of the day and 

added a fast-track unit to treat nonurgent patients more quickly.



Source: GAO data from site visits to 24 hospitals in selected MSAs, 

2002.



[End of table]



While more than two-thirds of the hospitals we visited were expanding 

or reported having plans to expand their capacity, nearly all of the 24 

hospitals we visited reported taking some type of action to increase 

the flow of patients through the emergency department and to reduce the 

time needed to place admitted emergency department patients into 

hospital beds. When patients cannot be moved efficiently through the 

emergency department and into inpatient hospital beds, they occupy 

emergency department space, staff, and services and reduce the capacity 

that might otherwise be available to treat other patients waiting to be 

seen in the emergency department. As shown in table 9, hospitals’ 

approaches to increase efficiency varied. For example, some hospitals 

focused on increasing the speed of the registration and triage process, 

while others were dependent on actions taken outside of the emergency 

department and on inpatient floors of the hospital, such as having 

coordinating committees or multidisciplinary teams that are directed to 

increase availability of inpatient beds and reduce boarding.



Table 9: Examples of Hospitals’ Increasing Efficiency:



Location: Atlanta; Actions taken by hospitals: One hospital formed a 

“bed briefing group,” which meets three times a day to discuss the 

types of emergency department patients waiting for inpatient beds and 

the types of inpatient beds expected to become available. Attendees 

include representatives from the hospital inpatient units (e.g., 

medical/surgical beds, critical care beds), the emergency department, 

nursing administration, and environmental services.



Location: Boston; Actions taken by hospitals: One hospital developed 

“Code Help ER,” under which all available staff resources are called on 

to expedite admissions and discharges when the hospital’s emergency 

department load is particularly high. Under this policy, priority is 

placed on transporting patients who have been boarding in the emergency 

department to inpatient beds, completing nursing reports, and cleaning 

beds before the hospital goes on diversion. After a “Code Help ER,” 

hospital officials conduct a review to determine the causes leading to 

that situation. Hospital officials recently completed the first 

analysis of the reviews and will be making recommendations for internal 

policy changes later this year. “Code Help ER” has been adopted by the 

state of Massachusetts as a best practice and is being used at other 

hospitals.



Location: Miami; Actions taken by hospitals: One hospital implemented a 

program called, “Think Noon!” to encourage doctors and hospital staff 

to discharge patients from inpatient beds before noon of the discharge 

day. The objective of the program is to make room available for 

patients waiting for inpatient beds, including those boarding in the 

emergency department.



Location: Phoenix; Actions taken by hospitals: One hospital streamlined 

the registration process; changed the process for providing lab and 

radiology services; and implemented “Code Purple,” which is similar to 

“Code Help ER” that is used in Boston.



Source: GAO data from site visits to 24 hospitals in selected MSAs, 

2002.



[End of table]



Community Activities Focus on Systems to Manage Diversion:



At the community level, efforts focused on ways to better manage 

crowding, particularly diversion, through task forces and development 

of diversion policies and tracking systems. At three of the six sites 

we visited, task forces had been formed to address these issues. The 

task forces generally addressed crowding and diversion in three ways: 

assembling stakeholders to examine causes, bringing attention to the 

issue, and developing methods to manage the problem (see table 10).



Table 10: Diversion Task Force Activity:



Location: Boston; Action taken: The Massachusetts Department of Public 

Health started a Diversion Task Force in 1998.; Participants: State 

health officials, health researchers, emergency department physicians, 

hospital administrators, and EMS officials; Result: Major 

accomplishments include the development of uniform guidelines and 

definitions for types of diversion. The Massachusetts Department of 

Public Health conducted a survey for the task force to study the 

reasons for the contributing factors to emergency department crowding 

and ambulance diversion in Massachusetts.



Location: Los Angeles; Action taken: The Healthcare Association of 

Southern California convened a task force in 2001 that focused on 

diversion.; Participants: Hospital administrators and EMS officials; 

Result: The task force developed a list of 12 possible contributing 

factors or underlying causes for diversion and drafted a list of 

potential solution steps.



Location: Phoenix; Action taken: The central Arizona regional EMS 

coordinating agency has a diversion task force that has been meeting 

since 1995.; Participants: Hospital representatives, emergency 

department clinicians, public and private EMS officials, and state and 

county health officials; Result: This group facilitates EMS and 

hospital discussions regarding diversion, developed protocols for 

diversion, and agreed on the use of a diversion tracking system.



Source: GAO data from site visits in selected MSAs, 2002.



[End of table]



Five of the six sites we visited had developed standard policies or 

guidelines regarding diversion and operated or participated in 

electronic systems for tracking ambulance diversion. The sixth site we 

visited--Miami-Dade County--took a different approach. The largest EMS 

provider in the area, the Miami-Dade Fire Rescue Department EMS 

Division, no longer formally honors hospital requests for 

diversion.[Footnote 33] On March 31, 1999, this EMS agency implemented 

a new policy directing ambulances to bring patients to the nearest 

appropriate hospital, citing concerns over the increased number of 

hospital emergency room closures and a compromised ability to deliver 

quality patient care.



For the five sites we visited that allowed diversion, each system 

improved communication among hospital and EMS providers by (1) allowing 

hospitals to request being put on diversion, (2) making hospitals aware 

of other hospitals’ diversion status, and (3) making ambulance 

dispatchers and ambulance drivers aware of which hospitals are on 

diversion. In these locations, diversion systems are used to provide a 

structure to systematically try to spread the ambulance volume during 

times of peak demand by redirecting ambulances to hospitals that are 

presumably less crowded. At three of these sites, EMS agencies produce 

reports on the number of hours each hospital was on diversion each 

month.[Footnote 34] EMS agencies, hospital associations, and government 

agencies use diversion reports to review policies and monitor 

hospitals’ diversion hours.



Current Efforts Unable to Reverse Crowding Trends:



While some sites we visited have experienced limited improvement, 

efforts under way have not made substantial reductions in the current 

extent of crowding. Some officials we interviewed described their 

efforts as attempts to manage the situation to keep it from getting 

worse rather than solving the problem. For example, in Boston, 

officials from the Massachusetts chapter of the American College of 

Emergency Physicians who participate in their state’s diversion task 

force said they see diversion as a Band-Aid for addressing what they 

believe is a crisis. They said that while the task force has taken 

steps to better manage diversion, increased demand for emergency 

department services due to events such as a bad flu season or disaster 

could still tax the system beyond its capacity.



Community-level data tend to support the view that these efforts, while 

perhaps helping to mitigate crowding, are not reversing the recent 

trends in crowding. For example, from 2000 through 2001, the three 

sites we visited that produce regular reports on diversion all 

experienced increases in the percentage of time that their hospitals 

were on diversion. The increase in the hours of diversion in these 

three locations ranged from 39 percent in the Los Angeles region to 73 

percent in the Boston region.



Studies Assessing Impact of Current Efforts Have Been Limited but Other 

Activities Are Under Way:



Despite the number of steps that hospitals and communities have taken, 

few studies have been conducted on the effects of hospitals’ and 

communities’ efforts to address crowding. Only 1 of the 24 hospitals we 

visited reported having completed an evaluation of the impact of its 

activities. This hospital had implemented a program to increase 

efficiency by discharging patients by noon and reported that its 

efforts resulted in earlier placement of admitted emergency department 

patients in inpatient beds. At the community level, while several 

communities monitor the number of hours on diversion, they reported 

that no comprehensive evaluations have been completed on the impact of 

communitywide efforts to address crowding.



Recent initiatives have been started by such organizations as the Joint 

Commission on Accreditation of Healthcare Organizations, AHRQ, and the 

Robert Wood Johnson Foundation that may help in future evaluations of 

crowding. These organizations have initiatives under way to further 

study crowding, develop hospital standards related to crowding, develop 

and test measures of crowding, provide technical assistance to 

hospitals, and evaluate potential steps to ease the problem. However, 

the results of these studies are not anticipated to be available until 

later in 2003 or 2004.



Concluding Observations:



Emergency department crowding is not an issue that can be solved in the 

emergency department alone. Rather, it is a complex issue that reflects 

the broader health care market. It is clear that, as a key part of the 

health care safety net, emergency departments in many of the nation’s 

largest communities are under strain.



Our work suggests that some aspects of the problem are hospital-

specific, such as high numbers of emergency patients, lack of space, 

and delays in obtaining test results. In addition, crowding appears to 

reflect the inability of individual hospitals to meet the demand for 

inpatient beds, particularly critical care and telemetry beds, both 

from emergency patients who need to be admitted to the hospital and 

patients admitted for elective procedures. When hospitals cannot 

accommodate peaks in demand, either because they lack space or because 

they choose to operate at levels that allow little excess capacity, the 

result is that emergency departments will often board patients who are 

waiting for inpatient beds. When they do, the capacity of the emergency 

department to treat additional patients is diminished.



While such issues as concerns about staffing inpatient beds and 

availability of other providers in the community are similar across 

communities, the solutions may differ by community and local health 

care market. For example, one community may face crowding in the 

emergency department largely because people have problems accessing 

physicians and other providers in the community, and potential 

solutions could involve steps to improve access to these other 

providers or establishing fast-track systems to treat nonurgent 

conditions in the emergency department. Another community may face 

crowding primarily because facilities have closed or populations have 

increased and there are too few hospital beds staffed and operated in 

the area. In this situation, the solution could involve reopening beds 

in existing facilities that were not set up and staffed. To address 

communitywide factors contributing to crowding, hospitals may need to 

work collaboratively with other facilities in their communities. 

Communitywide efforts such as task forces and standardized procedures 

and diversion policies have improved communications between hospitals 

and EMS providers and provided some degree of sharing the load when 

multiple hospitals are crowded. However, these efforts appear to only 

manage the problem of crowded conditions in emergency departments, 

rather than eliminate it.



Adding capacity, for both the emergency departments and for inpatient 

beds, has been suggested as a solution, but no one solution is likely 

to fit all circumstances. Crowding is clearly worse in some communities 

and hospitals than in others, and the specific reasons for crowding 

need to be better understood, particularly at the local level.



Comments from External Reviewers:



Representatives from the American College of Emergency Physicians and 

American Hospital Association and an independent reviewer provided 

comments on a draft of this report. The American College of Emergency 

Physicians stated that our methodology was comprehensive and systematic 

and identified and documented the leading causes of emergency 

department crowding. It also stated that while the crowding problems 

may be more pervasive in large metropolitan areas, its members had 

provided recent anecdotal information that indicates that the crowding 

problem is now becoming a concern in rural areas. While it is possible 

that some rural areas are becoming concerned about crowding, our survey 

was limited to hospitals in MSAs because available information and 

contacts with rural health organizations indicated that emergency 

department crowding was not a major problem in these areas.



An independent reviewer who has conducted research on emergency 

department crowding issues stated that the report was well done and 

informative. This reviewer and the American Hospital Association 

provided technical comments that we incorporated as appropriate.



As we agreed with your office, unless you publicly announce the 

contents of this report earlier, we plan no further distribution of it 

until 14 days from the date of this letter. We will then send copies to 

others who are interested and make copies available to others who 

request them. In addition, this report will be available at no charge 

on GAO’s Web site at http://www.gao.gov.



If you or your staff have any questions, please contact me at (202) 

512-7119. An additional GAO contact and the names of other staff 

members who made major contributions to this report are listed in app. 

IV.



Sincerely yours,



Janet Heinrich

Director, Health Care--Public Health Issues:



Signed by Janet Heinrich:



[End of section]



Appendix I: Scope and Methodology:



To accomplish our objectives, we surveyed over 2,000 short-term 

nonfederal, general medical and surgical hospitals with emergency 

departments located in metropolitan statistical areas (MSA). These 

hospitals are located in the 50 states and the District of Columbia. We 

obtained and analyzed data using three indicators of emergency 

department crowding: diversion, boarding, and patients who left before 

receiving a medical evaluation. We also used several hospital and 

community characteristics, including hospital ownership, admissions 

per bed, community population and growth, and the proportion of 

patients in the community without insurance. In addition, we visited 

six metropolitan areas--Atlanta, Boston, Cleveland, Los Angeles, Miami, 

and Phoenix. In these locations, we interviewed emergency medical 

services officials and officials at 4 hospitals in each area, for a 

total of 24 hospitals. We also interviewed (1) federal agency officials 

at the Department of Health and Human Services’ (HHS) National Center 

for Health Statistics, Health Resources and Services Administration, 

and Agency for Healthcare Research and Quality (AHRQ), (2) health care 

researchers at organizations such as the Council on Health Care 

Economics and Policy, the Robert Wood Johnson Foundation, and the Joint 

Commission on Accreditation of Healthcare Organizations, (3) 

representatives of national and local professional associations such as 

the American Ambulance Association, American Hospital Association, 

American College of Emergency Physicians, Emergency Nurses Association, 

National Association of Emergency Medical Services Physicians, and 

American Medical Association, and (4) hospital administrators and 

clinicians. In addition, we reviewed relevant studies and policy 

documents and analyzed information from national databases, including 

HHS’s National Center for Health Statistics’ National Hospital 

Ambulatory Medical Care Survey and AHRQ’s Healthcare Cost and 

Utilization Project, and the Health Resources and Services 

Administration’s Area Resource File. We conducted our work from July 

2001 through February 2003 in accordance with generally accepted 

government auditing standards.



Survey of Hospitals:



Survey Universe and Development:



To address questions about the extent of diversion, boarding, and 

patients leaving before a medical evaluation at hospitals in MSAs, we 

mailed a questionnaire to all 2,041 short-term, nonfederal, general 

medical and surgical care hospitals that reported they had emergency 

departments and were located in MSAs in the 50 states and the District 

of Columbia based on data from the American Hospital Association’s 

Annual Survey Database, Fiscal Year 2000. We mailed the questionnaires 

to the chief administrator of each hospital in May 2002. Each hospital 

was asked to report for the emergency department located at its main 

campus.



The survey included questions on the emergency department, such as 

(1) whether the hospital went on diversion and, if so, the number of 

hours on diversion in the hospital’s fiscal year 2001, (2) whether the 

hospital boarded patients for 2 hours or more in the past 12 months 

and, if so, the percentage of boarded patients who boarded 2 hours or 

more and the average number of hours boarded, and (3) the number of 

emergency department visits and the number of patients who left after 

triage but before a medical evaluation in the hospital’s fiscal year 

2001. It also included questions on the general hospital, including the 

number of staffed beds (excluding long-term care, labor and delivery, 

and postpartum beds) as of the last day of the hospital’s fiscal year 

2001.[Footnote 35] In developing these questions, we reviewed the 

literature and prior surveys related to crowding issues and conducted 

discussions with expert researchers. We also pretested our 

questionnaire in person with officials at 10 hospitals and refined the 

questionnaire as appropriate.



Response Rates:



Of the initial universe of 2,041 hospitals, 18 had closed by 2002 and 2 

did not have emergency departments in fiscal year 2001, resulting in a 

final universe of 2,021 hospitals. We conducted follow-up mailings and 

telephone follow-up calls to nonrespondents. We obtained responses from 

1,489 hospitals, for an overall response rate of about 74 

percent.[Footnote 36]



Survey Analysis:



We analyzed the response rates from various categories of hospitals and 

weighted responses to adjust for a lower response rate from investor-

owned (for-profit) hospitals so that our results would reflect the 

nationwide mix of hospital types. We analyzed the information provided 

by hospitals for three indicators of emergency department crowding--

diversion, boarding, and patients who left before a medical evaluation. 

In many cases, hospitals provided estimates for these indicators. 

Specifically, we estimate that (1) of hospitals that went on diversion, 

about 45 percent provided estimates for the number of hours on 

diversion in fiscal year 2001, (2) of hospitals that boarded patients 

for 2 hours or more in the past 12 months, about 74 percent provided 

estimates for the percentage of patients boarding 2 hours or more and 

about 74 percent provided estimates for the average number of hours 

patients boarded, and (3) about 34 percent of all hospitals provided 

estimates of the number of patients who left after triage but before a 

medical evaluation. For those hospitals that provided estimates, we 

used these estimates in our analyses.



We examined the extent of crowding in hospitals in MSAs, by different 

MSA and hospital characteristics. We grouped MSAs by characteristics 

such as U.S. Census Bureau population in 2000, population growth from 

1996 to 2000, and the percentage of the population without health 

insurance.[Footnote 37] We examined our indicators of crowding by 

hospital characteristics such as the number of staffed beds on the last 

day of fiscal year 2001; whether the hospital was public, private not-

for-profit, or investor-owned (for-profit); the hospital’s teaching 

status; whether it was a certified trauma center; and the proportion of 

emergency department visits covered by Medicare, Medicaid and the State 

Children’s Health Insurance Program, and self-pay as the payer source. 

We compared the medians of our three indicators of crowding across 

these characteristics. In calculating the median number of hours on 

diversion and the median percentage of patients boarding 2 hours or 

more, we considered hospitals that did not go on diversion in fiscal 

year 2001 to have no hours of diversion and hospitals that did not 

board any patients 2 hours or more to have no percentage of patients 

boarding.



We also conducted analyses to determine key factors associated with 

these indicators of crowding. We analyzed hospitals’ responses 

regarding which key factors contributed to our indicators of crowding 

and examined the medians for the crowding indicators grouped by 

admissions per bed in the MSA, percentage of emergency visits resulting 

in hospital inpatient admissions in fiscal year 2001, and the average 

daily census as a percentage of the number of staffed beds in the 

hospitals’ fiscal year 2001. In addition, we analyzed data from AHRQ’s 

Healthcare Cost and Utilization Project, Nationwide Inpatient Sample, 

2000, on the payer source of admissions.



Site Visits:



We conducted site visits in six locations: Atlanta, Georgia; Boston, 

Massachusetts; Los Angeles, California; Cleveland, Ohio; Miami, 

Florida; and Phoenix, Arizona. We selected the six sites judgmentally 

to include locations that varied in geographic location, the proportion 

of people without health insurance, MSA population, and recent 

population growth (see table 11). In addition, media reports and other 

sources had indicated that all six sites had reported problems with 

crowded emergency departments.



Table 11: Characteristics of Locations Selected for Site Visits:



Location: Atlanta; Geographic location 

(Census division): South Atlantic; U.S. Census population--2000 (for 

MSA): 4,112,198; Percentage change in MSA population 1996-2000: 16; 

Level of uninsurance compared to MSA average, 2000-2001 (percentage): 

Not significantly different (14); Admissions per bed, 2000, by quartile 

(1=bottom 25 percent, 4=top 

25 percent): 43.7; (2).



Location: Boston; Geographic location 

(Census division): New England; U.S. Census population--2000 (for MSA): 

3,406,829; Percentage change in MSA population 1996-2000: 5; Level of 

uninsurance compared to MSA average, 2000-2001 (percentage): Below 

(10); Admissions per bed, 2000, by quartile 

(1=bottom 25 percent, 4=top 

25 percent): 50.7; (4).



Location: Cleveland; Geographic location 

(Census division): East North Central; U.S. Census population--2000 

(for MSA): 2,250,871; Percentage change in MSA population 1996-2000: 1; 

Level of uninsurance compared to MSA average, 2000-2001 (percentage): 

Below (12); Admissions per bed, 2000, by quartile 

(1=bottom 25 percent, 4=top 

25 percent): 40.8; (2).



Location: Los Angeles; Geographic location 

(Census division): Pacific; U.S. Census population--2000 (for MSA): 

9,519,338; Percentage change in MSA population 1996-2000: 5; Level of 

uninsurance compared to MSA average, 2000-2001 (percentage): Above 

(25); Admissions per bed, 2000, by quartile 

(1=bottom 25 percent, 4=top 

25 percent): 46.0; (3).



Location: Miami; Geographic location 

(Census division): South Atlantic; U.S. Census population--2000 (for 

MSA): 2,253,362; Percentage change in MSA population 1996-2000: 7; 

Level of uninsurance compared to MSA average, 2000-2001 (percentage): 

Above (27); Admissions per bed, 2000, by quartile 

(1=bottom 25 percent, 4=top 

25 percent): 40.6; (2).



Location: Phoenix; Geographic location 

(Census division): Mountain; U.S. Census population--2000 (for MSA): 

3,251,876; Percentage change in MSA population 1996-2000: 18; Level of 

uninsurance compared to MSA average, 2000-2001 (percentage): Above 

(18); Admissions per bed, 2000, by quartile 

(1=bottom 25 percent, 4=top 

25 percent): 53.4; (4).



Source: U.S. Census Bureau, UCLA Center for Health Policy Research, and 

the American Hospital Association Annual Survey Database, Fiscal Year 

2000.



Note: UCLA Center for Health Policy Research provided analysis of 

uninsurance rates for 96 large MSAs based on the 2000 and 2001 Current 

Population Survey.



[End of table]



At the six locations, we visited four hospitals at each site (including 

public, for-profit, and not-for-profit hospitals), interviewed 

hospital administrators and emergency department clinicians, and 

observed operations in the emergency departments. We also interviewed 

officials from local EMS agencies, hospital associations, and other 

professional associations and experts knowledgeable about emergency 

department crowding.



[End of section]



Appendix II: Diversion Policies at the Six Locations GAO Visited:



While all six locations we visited had local or regional regulations, 

policies, or guidelines on ambulance diversion, these policies varied 

among and within the locations.[Footnote 38] For example, the largest 

emergency medical services (EMS) provider in the Miami area, the Miami-

Dade Fire Rescue Department EMS Division, stopped allowing hospitals to 

go on ambulance diversion as of March 31,1999, though the smaller City 

of Miami Fire-Rescue EMS agency did have policies for 

diversion.[Footnote 39] As shown in table 12, the locations we visited 

illustrate the differences between diversion policies of different 

communities and demonstrate how an episode of diversion in one place 

differs from an occurrence of diversion elsewhere.



* All six locations had defined types of diversion, including 

categories such as overall saturation in the emergency department, 

diversion for trauma cases only, diversion because a neurosurgeon was 

unavailable, diversion because a computed tomography (CT) scanner was 

unavailable, or diversion because of an internal disaster such as a 

power failure.



* Five of the locations had computer-based diversion systems in place 

at the time of our visit that allowed EMS dispatchers and hospital 

officials to check which hospitals, if any, in the EMS region were on 

diversion.



* All six locations had circumstances under which ambulances would take 

patients to the nearest appropriate hospital, regardless of whether the 

hospital was on diversion. For example, all six locations had policies 

to take patients with unstable or critical conditions to the nearest 

hospital, and four had policies that the patient’s request to go to a 

specific hospital could override diversion in certain circumstances.



* Most of the locations had a specific period after which a hospital 

would need to either reconfirm its diversion status or be automatically 

reopened to ambulances. However, the policies regarding the time limits 

varied. For example, 10 major Boston hospitals were automatically taken 

off diversion after 2 hours, while hospitals in Atlanta could go on 

diversion for up to 8 hours before they would automatically be reopened 

to all ambulances. In addition, hospitals in Boston, Phoenix, and 

Cleveland could be taken off of diversion status if too many hospitals 

in their immediate area wanted to go on diversion. For example, when 

two-thirds of hospitals in a given sector in Phoenix are on diversion, 

all of the hospitals are required to reopen.



Table 12: Comparison of EMS Areas and Diversion Policies for Site Visit 

Locations:



[See PDF for image]



Source: GAO data from site visits in selected MSAs, 2002.



Note: Diversion policies as of December 31, 2002.



[A] Miami-Dade Fire Rescue Department EMS Division, the largest EMS 

provider in the area, stopped honoring diversion requests as of March 

31, 1999. The second largest EMS agency, the City of Miami Fire-Rescue 

EMS, continues to honor diversion requests. Information provided is for 

the City of Miami Fire-Rescue EMS.



[B] In Atlanta, diversion categories are guidelines, not policy.



[C] In Georgia, patients are generally permitted to select the hospital 

to which they want to be transported. Ga. Comp. R. & Regs. r. 290-5-30-

.05 (2002).



[End of table]



[End of section]



Appendix III: Select Results of GAO Survey of Hospitals Regarding 

Emergency Department Crowding:



This appendix summarizes the results from questions we asked short-term 

nonfederal, general medical and surgical hospitals in metropolitan 

statistical areas (MSA) in the United States that had emergency 

departments in 2000. We sent the questionnaire to 2,041 hospitals that 

met these criteria--20 did not have emergency departments in fiscal 

year 2001 or were closed, for a total of 2,021 hospitals. We obtained 

responses from 1,489 hospitals, for an overall response rate of about 

74 percent. We weighted responses to adjust for a lower response rate 

from investor-owned (for-profit) hospitals to provide estimates 

representative of the entire universe of 2,021 hospitals in MSAs.



The following tables show select survey information on characteristics 

of the survey universe (table 13), emergency department visits and 

treatment spaces (tables 14 and 15), specialty on-call coverage (tables 

16 and 17), diversion (tables 18 through 27), boarding (tables 28 

through 31), patients who left before a medical evaluation (table 32), 

indicators of crowding by hospital characteristics (tables 33 through 

40), hospitals applying for regulatory approval to increase licensed 

beds (tables 41 and 42), and payer sources for emergency department 

visits (table 43).



Table 13: Characteristics of Hospitals in Survey Universe:



[See PDF for image]



Source: GAO survey of hospitals, 2002; U.S. Census Bureau; UCLA Center 

for Health Policy Research; and American Hospital Association Annual 

Survey Database, Fiscal Year 2000.



[A] Level of uninsurance compared to the average for 96 large MSAs 

based on analysis by the UCLA Center for Health Policy Research using 

data from the 2000 and 2001 Current Population Survey.



[End of table]



Table 14: Hospitals by Volume of Emergency Department Patient Visits, 

Fiscal Years 1997 and 2001:



Less than 25,000: [Empty].



Number of emergency department visits: Less than 25,000; 1997: Number 

of hospitals: 840; 1997: Percentage: 42; [Empty]; 2001: Number of 

hospitals: 696; 2001: Percentage: 34.



Number of emergency department visits: 25,000 to less than 50,000; 

1997: Number of hospitals: 783; 1997: Percentage: 39; [Empty]; 2001: 

Number of hospitals: 924; 2001: Percentage: 46.



Number of emergency department visits: 50,000 to less than 75,000; 

1997: Number of hospitals: 178; 1997: Percentage: 9; [Empty]; 2001: 

Number of hospitals: 276; 2001: Percentage: 14.



Number of emergency department visits: 75,000 or more; 1997: Number of 

hospitals: 43; 1997: Percentage: 2; [Empty]; 2001: Number of hospitals: 

99; 2001: Percentage: 5.



Number of emergency department visits: Data missing; 1997: Number of 

hospitals: 177; 1997: Percentage: 9; [Empty]; 2001: Number of 

hospitals: 26; 2001: Percentage: 1.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages may not add to 100 due to rounding.



[End of table]



Table 15: Mean Number of Emergency Department Standard and Other 

Treatment Spaces and Increase in Treatment Spaces, Last Day of Fiscal 

Years 1997 and 2001:



Standard (e.g., beds or treatment spaces specifically designed for 

emergency patients to receive care): [Empty]; Standard (e.g., beds or 

treatment spaces specifically designed for emergency patients to 

receive care): [Empty].



Type of treatment space: Standard (e.g., beds or treatment spaces 

specifically designed for emergency patients to receive care); 1997: 

Mean: 17.7; 1997: Number of hospitals: 1,927; [Empty]; 2001: Mean: 

20.8; 2001: Number of hospitals: 1,991; [Empty]; 1997 to 2001: Mean 

percentage increase: 21.9; 1997 to 2001: Number of hospitals: 1,919.



Type of treatment space: Other (e.g., stretchers in hallway, chairs); 

1997: Mean: 5.7; 1997: Number of hospitals: 1,718; [Empty]; 2001: Mean: 

7.8; 2001: Number of hospitals: 1,824; [Empty]; 1997 to 2001: Mean 

percentage increase: 35.2; 1997 to 2001: Number of hospitals: 1,295.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[End of table]



Table 16: Hospitals Reporting Problems with On-Call Physician Specialty 

Coverage during Fiscal Year 2001:



Did emergency department encounter any problems with on-call coverage?: 

Yes; Number of hospitals: 1,201; Percentage of hospitals: 59.



Did emergency department encounter any problems with on-call coverage?: 

No; Number of hospitals: 781; Percentage of hospitals: 39.



Did emergency department encounter any problems with on-call coverage?: 

Data missing; Number of hospitals: 39; Percentage of hospitals: 2.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[End of table]



Table 17: Specialty Areas for Which Hospitals Reported Having Problems 

with On-Call Physician Specialty Coverage in the Emergency Department 

during Fiscal Year 2001:



Specialty area: Anesthesiology; Number of hospitals: 64; Percentage: 5.



Specialty area: Cardiology; Number of hospitals: 127; Percentage: 11.



Specialty area: Cardio/thoracic surgery; Number of hospitals: 106; 

Percentage: 9.



Specialty area: Ear, nose, and throat; Number of hospitals: 332; 

Percentage: 28.



Specialty area: General surgery; Number of hospitals: 164; Percentage: 

14.



Specialty area: Neurology; Number of hospitals: 239; Percentage: 20.



Specialty area: Neurosurgery; Number of hospitals: 504; Percentage: 42.



Specialty area: Orthopedics; Number of hospitals: 401; Percentage: 33.



Specialty area: Pediatrics; Number of hospitals: 110; Percentage: 9.



Specialty area: Plastic surgery; Number of hospitals: 505; Percentage: 

42.



Specialty area: Psychiatry; Number of hospitals: 381; Percentage: 32.



Specialty area: Other (1); Number of hospitals: 340; Percentage: 28.



Specialty area: Other (2); Number of hospitals: 52; Percentage: 4.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages are based on an estimated 1,201 hospitals that 

reported problems with on-call coverage. Some hospitals reported 

multiple specialties.



[End of table]



Table 18: Hospitals on Diversion, Fiscal Year 2001:



Hospital on diversion in fiscal year 2001?: Yes; Number of hospitals: 

1,389; Percentage: 69.



Hospital on diversion in fiscal year 2001?: No; Number of hospitals: 

614; Percentage: 30.



Hospital on diversion in fiscal year 2001?: Data missing; Number of 

hospitals: 18; Percentage: 1.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[End of table]



Table 19: Hospitals by Percentage of Time on Diversion, Fiscal Year 

2001:



Percentage of time on diversion: Greater than 20 percent; Number of 

hospitals: 179; Percentage of hospitals: 9.



Percentage of time on diversion: More than 10 and up to 20 percent; 

Number of hospitals: 146; Percentage of hospitals: 7.



Percentage of time on diversion: More than 5 and up to 10 percent; 

Number of hospitals: 157; Percentage of hospitals: 8.



Percentage of time on diversion: Up to 5 percent; Number of hospitals: 

839; Percentage of hospitals: 42.



Percentage of time on diversion: Did not go on diversion; Number of 

hospitals: 614; Percentage of hospitals: 30.



Percentage of time on diversion: Data missing; Number of hospitals: 85; 

Percentage of hospitals: 4.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[End of table]



Table 20: Reasons Contributing to the Hospital Not Going on Diversion 

in Fiscal Year 2001:



Reason for not going on diversion: Adequate hospital capacity made 

diversion unnecessary; Yes: Number of hospitals: 358; Yes: Percentage: 

58; [Empty]; No: Number of hospitals: 177; No: Percentage: 29; [Empty]; 

Data missing: Number of hospitals: 80; Data missing: Percentage: 13.



Reason for not going on diversion: Only hospital serving large 

geographic area; Yes: Number of hospitals: 178; Yes: Percentage: 29; 

[Empty]; No: Number of hospitals: 332; No: Percentage: 54; [Empty]; 

Data missing: Number of hospitals: 104; Data missing: Percentage: 17.



Reason for not going on diversion: Other hospitals on diversion; Yes: 

Number of hospitals: 104; Yes: Percentage: 17; [Empty]; No: Number of 

hospitals: 383; No: Percentage: 62; [Empty]; Data missing: Number of 

hospitals: 127; Data missing: Percentage: 21.



Reason for not going on diversion: Administrative decision by emergency 

department to accept all ambulances; Yes: Number of hospitals: 346; 

Yes: Percentage: 56; [Empty]; No: Number of hospitals: 170; No: 

Percentage: 28; [Empty]; Data missing: Number of hospitals: 99; Data 

missing: Percentage: 16.



Reason for not going on diversion: Administrative decision by hospital 

to accept all ambulances; Yes: Number of hospitals: 398; Yes: 

Percentage: 65; [Empty]; No: Number of hospitals: 122; No: Percentage: 

20; [Empty]; Data missing: Number of hospitals: 94; Data missing: 

Percentage: 15.



Reason for not going on diversion: Diversion requires approval from 

outside the hospital and the request was denied; Yes: Number of 

hospitals: 10; Yes: Percentage: 2; [Empty]; No: Number of hospitals: 

469; No: Percentage: 76; [Empty]; Data missing: Number of hospitals: 

135; Data missing: Percentage: 22.



Reason for not going on diversion: Diversion requires approval from 

outside the hospital and was not worth requesting--it would have been 

denied anyway; Yes: Number of hospitals: 9; Yes: Percentage: 1; 

[Empty]; No: Number of hospitals: 473; No: Percentage: 77; [Empty]; 

Data missing: Number of hospitals: 132; Data missing: Number of 

hospitals: 80: 104: 127: 99: 94: 135: 22.



Reason for not going on diversion: State or local law or regulation 

prohibits diversion; Yes: Number of hospitals: 24; Yes: Percentage: 4; 

[Empty]; No: Number of hospitals: 465; No: Percentage: 76; [Empty]; 

Data missing: Number of hospitals: 125; Data missing: Number of 

hospitals: 80: 104: 127: 99: 94: 135: 20.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages are based on an estimated 614 hospitals that did 

not go on diversion in fiscal year 2001. Some hospitals reported 

multiple reasons.



[End of table]



Table 21: Trauma Center Status and Diversion, Fiscal Year 2001:



Was hospital that went on diversion designated as a certified trauma 

center?: Yes; Number of hospitals: 426; Percentage: 31.



Was hospital that went on diversion designated as a certified trauma 

center?: No; Number of hospitals: 929; Percentage: 67.



Was hospital that went on diversion designated as a certified trauma 

center?: Data missing; Number of hospitals: 33; Percentage: 2.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages are based on an estimated 1,389 hospitals that 

went on diversion in fiscal year 2001.



[End of table]



Table 22: Conditions Contributing to Hospitals Going on Diversion, 

Fiscal Year 2001:



(Percentages in parentheses)



Condition: Inability to transfer to intensive care unit/critical care 

unit (ICU/CCU beds); Number of hospitals (Percentage): Very great 

extent: 527; (38); Number of hospitals (Percentage): Great extent: 353; 

(25); Number of hospitals (Percentage): Moderate extent: 189; (14); 

Number of hospitals (Percentage): Some extent: 134; (10); Number of 

hospitals (Percentage): Little or 

no extent: 73; (5); Number of hospitals (Percentage): Not applicable: 

77; (6); Number of hospitals (Percentage): Missing data: 38; (3).



Condition: Inability to transfer to telemetry beds; Number of hospitals 

(Percentage): Very great extent: 476; (34); Number of hospitals 

(Percentage): Great extent: 329; (24); Number of hospitals 

(Percentage): Moderate extent: 185; (13); Number of hospitals 

(Percentage): Some extent: 101; (7); Number of hospitals (Percentage): 

Little or 

no extent: 99; (7); Number of hospitals (Percentage): Not applicable: 

138; (10); Number of hospitals (Percentage): Missing data: 59; (4).



Condition: Emergency department capacity exceeded; Number of hospitals 

(Percentage): Very great extent: 523; (38); Number of hospitals 

(Percentage): Great extent: 276; (20); Number of hospitals 

(Percentage): Moderate extent: 141; (10); Number of hospitals 

(Percentage): Some extent: 130; (9); Number of hospitals (Percentage): 

Little or 

no extent: 164; (12); Number of hospitals (Percentage): Not applicable: 

104; (7); Number of hospitals (Percentage): Missing data: 52; (4).



Condition: Inability to transfer to other inpatient beds; Number of 

hospitals (Percentage): Very great extent: 229; (17); Number of 

hospitals (Percentage): Great extent: 235; (17); Number of hospitals 

(Percentage): Moderate extent: 253; (18); Number of hospitals 

(Percentage): Some extent: 187; (13); Number of hospitals (Percentage): 

Little or 

no extent: 202; (15); Number of hospitals (Percentage): Not applicable: 

193; (14); Number of hospitals (Percentage): Missing data: 89; (6).



Condition: Inability to transfer to other facilities; Number of 

hospitals (Percentage): Very great extent: 43; (3); Number of hospitals 

(Percentage): Great extent: 44; (3); Number of hospitals (Percentage): 

Moderate extent: 105; (8); Number of hospitals (Percentage): Some 

extent: 176; (13); Number of hospitals (Percentage): Little or 

no extent: 535; (39); Number of hospitals (Percentage): Not applicable: 

397; (29); Number of hospitals (Percentage): Missing data: 88; (6).



Condition: Inability to transfer to pediatric beds; Number of hospitals 

(Percentage): Very great extent: 45; (3); Number of hospitals 

(Percentage): Great extent: 33; (2); Number of hospitals (Percentage): 

Moderate extent: 49; (4); Number of hospitals (Percentage): Some 

extent: 86; (6); Number of hospitals (Percentage): Little or 

no extent: 499; (36); Number of hospitals (Percentage): Not applicable: 

531; (38); Number of hospitals (Percentage): Missing data: 146; (10).



Condition: Concern emergency department would be overloaded due to 

other hospitals’ diversion; Number of hospitals (Percentage): Very 

great extent: 15; (1); Number of hospitals (Percentage): Great extent: 

36; (3); Number of hospitals (Percentage): Moderate extent: 48; (3); 

Number of hospitals (Percentage): Some extent: 70; (5); Number of 

hospitals (Percentage): Little or 

no extent: 571; (41); Number of hospitals (Percentage): Not applicable: 

541; (39); Number of hospitals (Percentage): Missing data: 108; (8).



Condition: Lack of on-call physician specialty coverage for emergency 

department; Number of hospitals (Percentage): Very great extent: 24; 

(2); Number of hospitals (Percentage): Great extent: 18; (1); Number of 

hospitals (Percentage): Moderate extent: 26; (2); Number of hospitals 

(Percentage): Some extent: 58; (4); Number of hospitals (Percentage): 

Little or 

no extent: 614; (44); Number of hospitals (Percentage): Not applicable: 

551; (40); Number of hospitals (Percentage): Missing data: 99; (7).



Condition: Internal disaster (e.g., power failure); Number of hospitals 

(Percentage): Very great extent: 21; (2); Number of hospitals 

(Percentage): Great extent: 4; (0.3); Number of hospitals (Percentage): 

Moderate extent: 5; (0.4); Number of hospitals (Percentage): Some 

extent: 27; (2); Number of hospitals (Percentage): Little or 

no extent: 615; (44); Number of hospitals (Percentage): Not applicable: 

604; (44); Number of hospitals (Percentage): Missing data: 111; (8).



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages are based on an estimated 1,389 hospitals going 

on diversion in fiscal year 2001 and may not add to 100 due to 

rounding. Some hospitals reported multiple conditions.



[End of table]



Table 23: Methods Hospitals Used to Minimize Diversion, Hospitals That 

Diverted in Fiscal Year 2001:



Methods used to minimize going on diversion: Staff worked overtime; 

Number of hospitals using this method: 1,142; Percentage: 82.



Methods used to minimize going on diversion: Opened inpatient beds in 

other areas of emergency department or hospital; Number of hospitals 

using this method: 823; Percentage: 59.



Methods used to minimize going on diversion: Canceled elective 

procedures; Number of hospitals using this method: 403; Percentage: 29.



Methods used to minimize going on diversion: Used on-call system for 

additional staff; Number of hospitals using this method: 652; 

Percentage: 47.



Methods used to minimize going on diversion: Moved patients to other 

facilities; Number of hospitals using this method: 358; Percentage: 26.



Methods used to minimize going on diversion: Used hospital float pool 

for additional staff; Number of hospitals using this method: 732; 

Percentage: 53.



Methods used to minimize going on diversion: Used overflow or holding 

areas for patients; Number of hospitals using this method: 905; 

Percentage: 65.



Methods used to minimize going on diversion: Other; Number of hospitals 

using this method: 221; Percentage: 16.



Methods used to minimize going on diversion: No particular method was 

used; Number of hospitals using this method: 80; Percentage: 6.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages are based on an estimated 1,389 hospitals going 

on diversion in fiscal year 2001. Some hospitals reported multiple 

methods.



[End of table]



Table 24: Hospitals Reporting State or Local Laws or Rules That 

Restrict When Hospitals Can Go on Diversion:



State or local laws or rules restricting when the emergency department/

hospital can go on diversion?: Yes; Number of hospitals: 624; 

Percentage: 45.



State or local laws or rules restricting when the emergency department/

hospital can go on diversion?: No; Number of hospitals: 733; 

Percentage: 53.



State or local laws or rules restricting when the emergency department/

hospital can go on diversion?: Data missing; Number of hospitals: 32; 

Percentage: 2.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages are based on an estimated 1,389 hospitals going 

on diversion in fiscal year 2001.



[End of table]



Table 25: Hospitals’ Knowledge of When Other Hospitals Are on 

Diversion:



Emergency department or hospital knows when other area hospitals are on 

diversion?: Yes; Number of hospitals: 1,328; Percentage: 96.



Emergency department or hospital knows when other area hospitals are on 

diversion?: No; Number of hospitals: 48; Percentage: 3.



Emergency department or hospital knows when other area hospitals are on 

diversion?: Data missing; Number of hospitals: 13; Percentage: 1.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages are based on an estimated 1,389 hospitals going 

on diversion in fiscal year 2001.



[End of table]



Table 26: Methods for Learning about Other Hospitals’ Diversion:



How emergency department or hospital knows when other area hospitals 

are on diversion: Internet site; Number of hospitals: 415; Percentage: 

31.



How emergency department or hospital knows when other area hospitals 

are on diversion: Telephone or radio alert from other hospitals; Number 

of hospitals: 570; Percentage: 43.



How emergency department or hospital knows when other area hospitals 

are on diversion: Telephone or radio alert from emergency medical 

services; Number of hospitals: 519; Percentage: 39.



How emergency department or hospital knows when other area hospitals 

are on diversion: Word of mouth (e.g., ambulance drivers); Number of 

hospitals: 458; Percentage: 35.



How emergency department or hospital knows when other area hospitals 

are on diversion: Other; Number of hospitals: 195; Percentage: 15.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages are based on an estimated 1,328 hospitals going 

on diversion in fiscal year 2001 that knew when other hospitals were on 

diversion. Some hospitals reported multiple methods.



[End of table]



Table 27: Type of Care the Hospital Was Unable to Receive or Accept for 

the Most Recent Episode of Diversion:



Type of care unable to accept: Acute care (medical/surgical); Number of 

hospitals: 626; Percentage: 45.



Type of care unable to accept: Telemetry; Number of hospitals: 719; 

Percentage: 52.



Type of care unable to accept: Intermediate (step-down); Number of 

hospitals: 471; Percentage: 34.



Type of care unable to accept: Critical (ICU/CCU); Number of hospitals: 

914; Percentage: 66.



Type of care unable to accept: Trauma; Number of hospitals: 434; 

Percentage: 31.



Type of care unable to accept: Pediatric; Number of hospitals: 313; 

Percentage: 23.



Type of care unable to accept: Psychiatric; Number of hospitals: 313; 

Percentage: 23.



Type of care unable to accept: Other; Number of hospitals: 240; 

Percentage: 17.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages are based on an estimated 1,389 hospitals going 

on diversion in fiscal year 2001. Some hospitals reported multiple 

types of care.



[End of table]



Table 28: Hospitals Boarding Patients 2 Hours or More, Past 12 Months:



Boarded patients for 2 hours or 

more in the past 12 months?: Yes; Number of hospitals: 1,822; 

Percentage: 90.



Boarded patients for 2 hours or 

more in the past 12 months?: No; Number of hospitals: 173; Percentage: 

9.



Boarded patients for 2 hours or 

more in the past 12 months?: Data missing; Number of hospitals: 26; 

Percentage: 1.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[End of table]



Table 29: Hospitals by Percentage of Patients Boarded 2 Hours or More, 

Past 12 Months:



Percentage of patients boarded 

2 hours or more: 75 percent or more; Number of hospitals: 630; 

Percentage of hospitals: 31.



Percentage of patients boarded 

2 hours or more: 50 percent to less than 75 percent; Number of 

hospitals: 260; Percentage of hospitals: 13.



Percentage of patients boarded 

2 hours or more: 25 percent to less than 50 percent; Number of 

hospitals: 200; Percentage of hospitals: 10.



Percentage of patients boarded 

2 hours or more: Less than 25 percent; Number of hospitals: 651; 

Percentage of hospitals: 32.



Percentage of patients boarded 

2 hours or more: Did not board any patients 2 hours 

or more; Number of hospitals: 173; Percentage of hospitals: 9.



Percentage of patients boarded 

2 hours or more: Data missing; Number of hospitals: 107; Percentage of 

hospitals: 5.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Hospitals were asked what percentage of all the patients 

boarded in the past 12 months boarded for 2 hours or more.



[End of table]



Table 30: Hospitals by Average Hours of Patients Boarding, Past 12 

Months:



Average number of hours patients boarded: 8 hours or more; Number of 

hospitals: 399; Percentage of hospitals: 20.



Average number of hours patients boarded: 6 to less than 8 hours; 

Number of hospitals: 266; Percentage of hospitals: 13.



Average number of hours patients boarded: 4 to less than 6 hours; 

Number of hospitals: 371; Percentage of hospitals: 18.



Average number of hours patients boarded: Less than 4 hours; Number of 

hospitals: 593; Percentage of hospitals: 29.



Average number of hours patients boarded: Did not board any patients 

2 hours or more; Number of hospitals: 173; Percentage of hospitals: 9.



Average number of hours patients boarded: Data missing; Number of 

hospitals: 219; Percentage of hospitals: 11.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Hospitals that boarded patients 2 hours or more in the past 

12 months were asked the average number of hours that a patient was 

boarded, including those patients boarded for less than 2 hours.



[End of table]



Table 31: Conditions Contributing to Boarding Patients 2 Hours or More 

in Past 12 Months:



Condition: Inability to transfer to telemetry beds; Number of hospitals 

(percentage): Very great extent: 855; (47); Number of hospitals 

(percentage): Great extent: 427; (23); Number of hospitals 

(percentage): Moderate extent: 193; (11); Number of hospitals 

(percentage): Some extent: 103; (6); Number of hospitals (percentage): 

Little or no extent: 64; (3); Number of hospitals (percentage): Not 

applicable: 98; (5); Number of hospitals (percentage): Data missing: 

82; (5).



Condition: Inability to transfer to critical care (ICU/CCU beds); 

Number of hospitals (percentage): Very great extent: 775; (43); Number 

of hospitals (percentage): Great extent: 418; (23); Number of hospitals 

(percentage): Moderate extent: 279; (15); Number of hospitals 

(percentage): Some extent: 178; (10); Number of hospitals (percentage): 

Little or no extent: 63; (3); Number of hospitals (percentage): Not 

applicable: 37; (2); Number of hospitals (percentage): Data missing: 

72; (4).



Condition: Inability to transfer to other inpatient beds; Number of 

hospitals (percentage): Very great extent: 494; (27); Number of 

hospitals (percentage): Great extent: 383; (21); Number of hospitals 

(percentage): Moderate extent: 341; (19); Number of hospitals 

(percentage): Some extent: 219; (12); Number of hospitals (percentage): 

Little or no extent: 135; (7); Number of hospitals (percentage): Not 

applicable: 124; (7); Number of hospitals (percentage): Data missing: 

126; (7).



Condition: Emergency department capacity exceeded; Number of hospitals 

(percentage): Very great extent: 276; (15); Number of hospitals 

(percentage): Great extent: 189; (10); Number of hospitals 

(percentage): Moderate extent: 179; (10); Number of hospitals 

(percentage): Some extent: 207; (11); Number of hospitals (percentage): 

Little or no extent: 490; (27); Number of hospitals (percentage): Not 

applicable: 277; (15); Number of hospitals (percentage): Data missing: 

204; (11).



Condition: Inability to transfer to other facilities; Number of 

hospitals (percentage): Very great extent: 149; (8); Number of 

hospitals (percentage): Great extent: 157; (9); Number of hospitals 

(percentage): Moderate extent: 213; (12); Number of hospitals 

(percentage): Some extent: 297; (16); Number of hospitals (percentage): 

Little or no extent: 489; (27); Number of hospitals (percentage): Not 

applicable: 224; (12); Number of hospitals (percentage): Data missing: 

293; (16).



Condition: Inability to transfer to pediatric beds; Number of hospitals 

(percentage): Very great extent: 114; (6); Number of hospitals 

(percentage): Great extent: 61; (3); Number of hospitals (percentage): 

Moderate extent: 108; (6); Number of hospitals (percentage): Some 

extent: 188; (10); Number of hospitals (percentage): Little or no 

extent: 643; (35); Number of hospitals (percentage): Not applicable: 

510; (28); Number of hospitals (percentage): Data missing: 198; (11).



Condition: Lack of on-call physician specialty coverage for emergency 

department; Number of hospitals (percentage): Very great extent: 27; 

(2); Number of hospitals (percentage): Great extent: 26; (1); Number of 

hospitals (percentage): Moderate extent: 71; (4); Number of hospitals 

(percentage): Some extent: 189; (10); Number of hospitals (percentage): 

Little or no extent: 784; (43); Number of hospitals (percentage): Not 

applicable: 524; (29); Number of hospitals (percentage): Data missing: 

202; (11).



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages are based on an estimated 1,822 hospitals 

boarding patients for 2 hours or more in the past 12 months and may not 

add to 100 due to rounding.



[End of table]



Table 32: Hospitals by Percentage of Patients Who Left after Triage but 

before a Medical Evaluation, Fiscal Year 2001:



Percentage of patients who left after triage but before a medical 

evaluation: 5 percent or more; Number of hospitals: 133; Percentage of 

hospitals: 7.



Percentage of patients who left after triage but before a medical 

evaluation: More than 3 to less than 5 percent; Number of hospitals: 

244; Percentage of hospitals: 12.



Percentage of patients who left after triage but before a medical 

evaluation: 1 to 3 percent; Number of hospitals: 780; Percentage of 

hospitals: 39.



Percentage of patients who left after triage but before a medical 

evaluation: Less than 1 percent; Number of hospitals: 730; Percentage 

of hospitals: 36.



Percentage of patients who left after triage but before a medical 

evaluation: Data missing; Number of hospitals: 134; Percentage of 

hospitals: 7.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals. Percentages may not add to 100 due to rounding.



[End of table]



Table 33: Indicators of Crowding, by Number of Staffed Inpatient Beds, 

Last Day of Fiscal Year 2001:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Number of staffed inpatient beds[A]: Top 25 percent[B]: 196 

hours; Number of staffed inpatient beds[A]: Bottom 25 percent[C]: <7 

hours.



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 12 months; Number of staffed inpatient beds[A]: Top 25 

percent[B]: 66%; Number of staffed inpatient beds[A]: Bottom 25 

percent[C]: 8%.



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Number of staffed inpatient beds[A]: Top 25 

percent[B]: 2.0%; Number of staffed inpatient beds[A]: Bottom 25 

percent[C]: 1.0%.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] Excludes long-term care, labor and delivery, and postpartum beds.



[B] Hospitals in the top 25 percent had more than 294 staffed inpatient 

beds.



[C] Hospitals in the bottom 25 percent had 107 or fewer staffed 

inpatient beds.



[End of table]



Table 34: Indicators of Crowding, by Number of Emergency Department 

Visits per Standard Treatment Space, Fiscal Year 2001:



Crowding indicators.



Median number of hours on diversion in fiscal year 2001; Visits per 

standard treatment space[A]: Top 25 percent[B]: 35 hours[D]; Visits per 

standard treatment space[A]: Bottom 25 percent[C]: 22 hours[D].



Median percentage of patients boarded 2 hours or more in past 12 

months; Visits per standard treatment space[A]: Top 25 percent[B]: 

25%[D]; Visits per standard treatment space[A]: Bottom 25 percent[C]: 

24%[D].



Median percentage of patients who left before a medical evaluation; 

Visits per standard treatment space[A]: Top 25 percent[B]: 1.6%; Visits 

per standard treatment space[A]: Bottom 25 percent[C]: 1.2%.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] Number of visits in fiscal year 2001 and number of standard 

treatment spaces as of the last day of fiscal year 2001.



[B] Hospitals in the top 25 percent had more than 1,993 visits per 

standard treatment space.



[C] Hospitals in the bottom 25 percent had 1,426 or fewer visits per 

standard treatment space.



[D] No statistically significant difference between the medians for 

hospitals in the top and bottom quartiles.



[End of table]



Table 35: Indicators of Crowding, by Number of Emergency Department 

Standard Treatment Spaces per Staffed Inpatient Hospital Bed, Last Day 

of Fiscal Year 2001:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Standard treatment space per staffed 

inpatient bed[A]: Top 25 percent[B]: 19 hours; Standard treatment space 

per staffed 

inpatient bed[A]: Bottom 25 percent[C]: 97 hours.



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 

12 months; Standard treatment space per staffed 

inpatient bed[A]: Top 25 percent[B]: 22%[D]; Standard treatment space 

per staffed 

inpatient bed[A]: Bottom 25 percent[C]: 37%[D].



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Standard treatment space per staffed 

inpatient bed[A]: Top 25 percent[B]: 1.2%[D]; Standard treatment space 

per staffed 

inpatient bed[A]: Bottom 25 percent[C]: 1.6%[D].



Source: GAO survey of hospitals, 2002.



Note: Number of standard treatments spaces and staffed inpatient beds 

as of the last day of fiscal year 2001. Responses were weighted to 

provide estimates for the universe of hospitals.



[A] Excludes labor and delivery, postpartum, and long-term care beds.



[B] Hospitals in the top 25 percent had more than 0.15 standard 

treatment spaces in the emergency department per staffed inpatient bed.



[C] Hospitals in the bottom 25 percent had less than 0.07 standard 

treatment spaces in the emergency department per staffed inpatient bed.



[D] No statistically significant difference in the medians for 

hospitals in the top and bottom quartiles.



[End of table]



Table 36: Indicators of Crowding, by Emergency Department Admissions 

per Staffed Inpatient Bed, Fiscal Year 2001:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Emergency department admissions per staffed inpatient bed[A]: Top 

25 percent[B]: 86 hours[D]; Emergency department admissions per staffed 

inpatient bed[A]: Bottom 25 percent[C]: 22 hours[D].



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 12 months; Emergency department admissions per staffed 

inpatient bed[A]: Top 25 percent[B]: 46%; Emergency department 

admissions per staffed inpatient bed[A]: Bottom 25 percent[C]: 19%.



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Emergency department admissions per staffed 

inpatient bed[A]: Top 25 percent[B]: 1.5%[D]; Emergency department 

admissions per staffed inpatient bed[A]: Bottom 25 percent[C]: 1.2%[D].



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] Number of staffed inpatient beds as of the last day of fiscal year 

2001. Excludes long-term care, labor and delivery, and postpartum beds.



[B] Hospitals in the top 25 percent had more than 35 emergency 

department admissions per staffed inpatient bed.



[C] Hospitals in the bottom 25 percent had less than 21 emergency 

department admissions per staffed inpatient bed.



[D] No statistically significant difference in the medians for 

hospitals in the top and bottom quartiles.



[End of table]



Table 37: Indicators of Crowding, by Hospital Ownership:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Type of ownership: Private, not-

for-profit: 52 hours[A]; Type of ownership: Investor-owned

(for-profit): 40 hours[A]; Type of ownership: Public (nonfederal): 11 

hours[A].



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 12 months; Type of ownership: Private, not-

for-profit: 40%; Type of ownership: Investor-owned

(for-profit): 22%; Type of ownership: Public (nonfederal): 23%[A].



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Type of ownership: Private, not-

for-profit: 1.3%; Type of ownership: Investor-owned

(for-profit): 1.6%[A]; Type of ownership: Public (nonfederal): 1.7%.



Source: GAO survey of hospitals, 2002, and American Hospital 

Association Annual Survey Database, Fiscal Year 2000.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] No statistically significant difference between the medians for 

hospitals with this type of ownership compared with other types of 

ownership.



[End of table]



Table 38: Indicators of Crowding, by Trauma Center Status:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Certified trauma center: 75 hours; Not a certified 

trauma center: 32 hours.



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 

12 months; Certified trauma center: 46%[A]; Not a certified 

trauma center: 28%[A].



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Certified trauma center: 1.5%[A]; Not a certified 

trauma center: 1.3%[A].



Source: GAO survey of hospitals, 2002, and American Hospital 

Association Annual Survey Database, Fiscal Year 2000.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] No statistically significant difference between the medians for 

hospitals that are certified trauma centers and those that are not.



[End of table]



Table 39: Indicators of Crowding, by Teaching Status:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Teaching hospital: 148 hours; Not a teaching hospital: 19 hours.



Crowding indicators: Median percentage of patients boarded 2 hours or 

more in past 12 months; Teaching hospital: 59%; Not a teaching 

hospital: 20%.



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Teaching hospital: 1.7%; Not a teaching hospital: 

1.2%.



Source: GAO survey of hospitals, 2002, and American Hospital 

Association Annual Survey Database, Fiscal Year 2000.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[End of table]



Table 40: Indicators of Crowding, by Select Payer Sources for Emergency 

Department Visits, Fiscal Year 2001:



Crowding indicators: Median number of hours on diversion in fiscal year 

2001; Medicare: Top 25 percent[A]: 25 hours; Medicare: Bottom 25 

percent[A]: 76 hours; [Empty]; Medicaid and State Children’s Health 

Insurance Program (SCHIP): Top 25 percent[B]: 32 hours[D]; Medicaid and 

State Children’s Health Insurance Program (SCHIP): Bottom 25 

percent[B]: 100 hours[D]; [Empty]; Self-pay: Top 25 percent[C]: 41 

hours[D]; Self-pay: Bottom 25 percent[C]: 50 hours[D].



Crowding indicators: Median percentage of patients boarded 

2 hours or more in past 12 months; Medicare: Top 25 percent[A]: 31%[D]; 

Medicare: Bottom 25 percent[A]: 38%[D]; [Empty]; Medicaid and State 

Children’s Health Insurance Program (SCHIP): Top 25 percent[B]: 42%[D]; 

Medicaid and State Children’s Health Insurance Program (SCHIP): Bottom 

25 percent[B]: 49%[D]; [Empty]; Self-pay: Top 25 percent[C]: 38%[D]; 

Self-pay: Bottom 25 percent[C]: 49%[D].



Crowding indicators: Median percentage of patients who left before a 

medical evaluation; Medicare: Top 25 percent[A]: 1.1%; Medicare: Bottom 

25 percent[A]: 1.7%; [Empty]; Medicaid and State Children’s Health 

Insurance Program (SCHIP): Top 25 percent[B]: 1.8%; Medicaid and State 

Children’s Health Insurance Program (SCHIP): Bottom 25 percent[B]: 

1.0%; [Empty]; Self-pay: Top 25 percent[C]: 2.3%; Self-pay: Bottom 25 

percent[C]: 1.0%.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[A] The top 25 percent of hospitals had more than 30 percent of visits 

covered by Medicare, while the bottom 25 percent had 14 percent or 

fewer visits covered by Medicare.



[B] The top 25 percent of hospitals had more than 21 percent of visits 

covered by Medicaid or SCHIP, while the bottom 25 percent had 8 percent 

or fewer visits covered by Medicaid or SCHIP.



[C] The top 25 percent had more than 20 percent of visits covered by 

self-pay patients, while the bottom 25 percent had 8 percent or fewer 

visits covered by self-pay patients.



[D] No statistically significant difference between the medians for 

hospitals in the top and bottom quartiles.



[End of table]



Table 41: Hospitals Applying for Regulatory Approval to Increase 

Licensed Beds, since Start of Fiscal Year 2001:



Requested approval to increase licensed beds?[A]: Yes; Number of 

hospitals: 296; Percentage: 15.



Requested approval to increase licensed beds?[A]: No; Number of 

hospitals: 1,639; Percentage: 81.



Requested approval to increase licensed beds?[A]: Data missing; Number 

of hospitals: 86; Percentage: 4.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals and include responses from all hospitals, regardless of 

whether states had certificate of need processes.



[A] According to the American Health Planning Association, 26 states 

and the District of Columbia required hospitals to apply for regulatory 

approval to increase licensed inpatient beds in 2002, a process known 

as the certificate of need process, while 24 states had no such 

requirement.



[End of table]



Table 42: Types of Beds Requested since Start of Fiscal Year 2001:



Type of bed: Acute care (medical/surgical); Number of hospitals: 129; 

Percentage: 44.



Type of bed: Telemetry; Number of hospitals: 44; Percentage: 15.



Type of bed: Intermediate; Number of hospitals: 17; Percentage: 6.



Type of bed: Critical care (ICU/CCU); Number of hospitals: 97; 

Percentage: 33.



Type of bed: Pediatric; Number of hospitals: 11; Percentage: 4.



Type of bed: Psychiatric; Number of hospitals: 36; Percentage: 12.



Type of bed: Other; Number of hospitals: 97; Percentage: 33.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals and include responses from all hospitals, regardless of 

whether states had certificate of need processes. Percentages are based 

on an estimated 296 hospitals that applied for regulatory approval to 

increase licensed beds since start of fiscal year 2001. Some hospitals 

reported multiple types of beds.



[End of table]



Table 43: Average Proportion of Emergency Visits Covered by Medicare, 

Medicaid and SCHIP, and Self-Pay, Fiscal Year 2001:



Payer source: Medicare; Mean percentage: 24; Number of hospitals 

reporting: 1,892.



Payer source: Medicaid and SCHIP; Mean percentage: 16; Number of 

hospitals reporting: 1,884.



Payer source: Self-pay; Mean percentage: 15; Number of hospitals 

reporting: 1,860.



Source: GAO survey of hospitals, 2002.



Note: Responses were weighted to provide estimates for the universe of 

hospitals.



[End of table]



[End of section]



Appendix IV: GAO Contact and Staff Acknowledgments:



GAO Contact:



Kim Yamane, (206) 287-4772:



Acknowledgments:



Other major contributors to this report were Diana Birkett, Jennifer 

Cohen, Bruce Greenstein, Katherine Iritani, Susan Lawes, Lisa A. Lusk, 

Behn Miller, Dae Park, Tina Schwien, and Stan Stenersen.



FOOTNOTES



[1] We focused on hospitals located in metropolitan areas designated as 

MSAs and primary metropolitan statistical areas by the U.S. Census 

Bureau. For purposes of this report, we will refer to both types of 

areas as MSAs. In 2000, MSAs accounted for about 80 percent of the 

nation’s population.



[2] The hospitals that met our criteria were located in 321 MSAs. We 

also excluded federal hospitals, specialty hospitals, long-term care 

facilities, and hospitals located outside the 50 states or the District 

of Columbia.



[3] Many hospitals provided estimates for the three indicators. These 

estimates were used for our analyses.



[4] L.F. McCaig and N. Ly, “National Hospital Ambulatory Medical Care 

Survey: 2000 Emergency Department Summary,” Advance Data from Vital and 

Health Statistics, no. 326 (Hyattsville, Md.: National Center for 

Health Statistics, 2002).



[5] The Lewin Group, Emergency Department Overload: A Growing Crisis; 

The Results of the AHA Survey of Emergency Department (ED) and Hospital 

Capacity, April 2002.



[6] EMS officials also report that in addition to longer ambulance 

transport times when hospitals are on diversion, crowded emergency 

departments also tie up ambulance providers while they wait to transfer 

their patients to the emergency department staff.



[7] R. Derlet and others, “Frequent Overcrowding in U.S. Emergency 

Departments,” Academic Emergency Medicine, vol. 8, no. 2 (2001), and 

S.K. Epstein and D. Slate, “The Massachusetts College of Emergency 

Physicians Ambulance Diversion Survey” (abstract), Academic Emergency 

Medicine, vol. 8, no. 5 (2001).



[8] 42 U.S.C. 1395dd(a) (2000). Under certain circumstances, a hospital 

may also transfer an emergency patient to another hospital. See U.S. 

General Accounting Office, Emergency Care: EMTALA Implementation and 

Enforcement Issues, GAO-01-747 (Washington, D.C.: June 22, 2001) for 

more information on this federal law.



[9] See 42 C.F.R. § 489.24(b) (2002). Under federal regulation, a 

hospital may only deny access to non-hospital-owned ambulances.



[10] McCaig and Ly.



[11] S. Baker and others, “Patients Who Leave a Public Hospital 

Emergency Department Without Being Seen by a Physician,” Journal of the 

American Medical Association, vol. 266, no. 8 (1991).



[12] If data were not available, we asked hospitals to provide their 

best estimates. We estimate that about 45 percent of hospitals that 

went on diversion in fiscal year 2001 provided estimates for the total 

number of hours that their emergency departments were on diversion.



[13] The 248 MSAs include those MSAs for which (1) more than half of 

hospitals in the MSA returned surveys and (2) of those hospitals that 

returned surveys, more than half provided data on diversion hours.



[14] If data were not available, we asked hospitals to provide their 

best estimates. We estimate that about 74 percent of hospitals that 

boarded patients for 2 hours or more in the past 12 months estimated 

the percentage of patients boarding, and about 74 percent estimated the 

average number of hours patients boarded.



[15] The 206 MSAs include those MSAs for which (1) more than half of 

hospitals in the MSA returned surveys and (2) of those hospitals that 

returned surveys, 50 percent or more provided data on the percentage of 

boarded patients boarding for 2 hours or more and the average number of 

hours boarded.



[16] If data were not available, we asked hospitals to provide their 

best estimates. We estimate that about 34 percent of hospitals provided 

estimates of the number of patients who completed triage in the 

emergency department but left before a medical evaluation during fiscal 

year 2001.



[17] The 243 MSAs include those MSAs for which (1) more than half of 

hospitals in the MSA returned surveys and (2) of those hospitals that 

returned surveys, 50 percent or more provided data on patients who left 

before a medical evaluation.



[18] These characteristics may be associated with other MSA or hospital 

characteristics. Our analysis was limited to examining the independent 

associations of MSA and hospital characteristics and our three 

indicators of crowding.



[19] In looking at those hospitals on diversion for more than 10 

percent of the time, 41 percent of hospitals were located in MSAs with 

populations of 2.5 million or more people compared to 27 percent in 

MSAs of less than 1 million people.



[20] Our analysis was limited to 96 large MSAs for which data on the 

level of uninsurance were available and used data on waiting times from 

the National Hospital Ambulatory Medical Care Survey.



[21] In addition, the amount of time hospitals were on diversion varied 

between different parts of the Los Angeles MSA. Of nine areas 

designated by the Los Angeles County EMS agency, the percentage of time 

that hospitals were on diversion in 2001 ranged from 12 percent in one 

area to more than 46 percent of the time in another area.



[22] These characteristics may be associated with other hospital or MSA 

characteristics. Our analysis was limited to examining the independent 

associations of hospital and MSA characteristics and our three 

indicators of crowding.



[23] For the average occupancy in fiscal year 2001, our analysis used 

information that hospitals reported on their average daily census at 

midnight. While the census at midday may be higher than at midnight, 

only an estimated 13 percent of hospitals provided data on midday 

census.



[24] Massachusetts Department of Public Health, The DPH Ambulance 

Diversion Survey: February 1-7, 2001.



[25] A third reason cited by some hospital officials was that low 

profit margins make it difficult to access capital to expand. However, 

we did not find any significant difference in our three crowding 

indicators between those surveyed hospitals with the highest and lowest 

average hospital margins reported for fiscal years 1997 to 1999.



[26] M. McManus, “Emergency Department Overcrowding in Massachusetts: 

Making Room in our Hospitals,” The Massachusetts Health Policy Forum, 

no. 12 (2001).



[27] While our survey asked hospital officials to provide data on (1) 

the hours of emergency department physician and other clinician patient 

care coverage in the emergency department on a typical day in fiscal 

year 2001 and (2) data on the hospital and agency (contract) nursing 

full-time equivalent staff in both the emergency department and the 

general hospital on the last day of fiscal year 2001, a large 

proportion of missing data prevented us from examining our three 

crowding indicators by hospital staffing levels.



[28] This analysis is based on national estimates of discharges from 

nonfederal, short-term, general medical and surgical hospitals with 

emergency departments and excluded neonatal and maternal discharges.



[29] Data from the National Center for Health Statistics for 1992 to 

2000 also showed that the percentage of emergency department visits 

admitted to the hospital had not changed significantly--about 12 

percent of visits resulted in admissions in 2000. However, the same 

data found that the percentages of emergency department visits referred 

to another physician or clinic or with no follow-up planned had 

increased significantly to about 47 percent and 10 percent of visits, 

respectively, in 2000.



[30] Although officials at several hospitals we visited reported that 

difficulty getting specialty coverage for the emergency department may 

contribute to longer patient stays in the emergency department while 

waiting for specialists to evaluate their condition, most hospitals we 

surveyed did not believe that this problem contributed to crowding to a 

great extent. While our survey found that 59 percent of hospitals 

reported problems with on-call specialty coverage, only about 5 percent 

of hospitals that went on diversion reported that lack of on-call 

specialty coverage contributed to diversion to a moderate, great, or 

very great extent, and only 7 percent of hospitals that boarded 

patients reported that problems with on-call coverage contributed to 

boarding to a moderate, great, or very great extent.



[31] Many hospitals we visited and surveyed reported using nonstandard 

treatment spaces such as stretchers in the hallway or chairs for 

treating emergency patients. Nineteen of the 24 hospitals we visited 

reported using nonstandard treatment spaces, and 78 percent of 

hospitals in our survey reported having hallway treatment spaces and 

other nonstandard treatment spaces at the end of fiscal year 2001.



[32] While 11 hospitals reported having expanded or planning to expand, 

1 hospital in Los Angeles reported plans to build a new, but smaller 

hospital inpatient facility to replace the older, larger one. A 

hospital official cited financial pressures as the primary reason for 

smaller capacity. Of the hospitals we surveyed, about 296 (15 percent) 

reported having applied for an increased number of licensed beds since 

the beginning of fiscal year 2001. In 2002, 26 states and the District 

of Columbia required hospitals to apply for regulatory approval to 

increase the number of hospital inpatient beds, according to the 

American Health Planning Association.



[33] The second largest EMS agency in Miami-Dade County, the City of 

Miami Fire-Rescue EMS, does have policies for diversion that govern its 

service area. See app. II for additional information on the diversion 

policies of each site visited.



[34] These three sites included Boston’s EMS region, Los Angeles County 

EMS, and Phoenix’s EMS region. While Atlanta’s EMS region has a system 

to notify hospitals and EMS providers when hospitals are on diversion, 

data reports that track diversions over time are not yet available. 

Cleveland has 55 local EMS agencies and does not report on 

communitywide diversion data regularly.



[35] A third section included questions on emergency preparedness for 

mass casualty incidents, which will be reported separately.



[36] Questionnaires received after September 3, 2002, and those of 

hospitals that only returned the section on emergency preparedness were 

not included in calculating our response rate and were excluded from 

our analyses.



[37] Our analysis of uninsurance rates in MSAs was limited to data from 

the UCLA Center for Health Policy Research for 96 large MSAs based on 

the 2000 and 2001 Current Population Survey. 



[38] We will refer to regulations, policies, and guidelines as policies 

in this appendix.



[39] For purposes of this appendix, we will refer to the policies of 

the City of Miami in discussing the diversion policies in the Miami 

area.



GAO’s Mission:



The General Accounting Office, the investigative arm of Congress, 

exists to support Congress in meeting its constitutional 

responsibilities and to help improve the performance and accountability 

of the federal government for the American people. GAO examines the use 

of public funds; evaluates federal programs and policies; and provides 

analyses, recommendations, and other assistance to help Congress make 

informed oversight, policy, and funding decisions. GAO’s commitment to 

good government is reflected in its core values of accountability, 

integrity, and reliability.



Obtaining Copies of GAO Reports and Testimony:



The fastest and easiest way to obtain copies of GAO documents at no 

cost is through the Internet. GAO’s Web site ( www.gao.gov ) contains 

abstracts and full-text files of current reports and testimony and an 

expanding archive of older products. The Web site features a search 

engine to help you locate documents using key words and phrases. You 

can print these documents in their entirety, including charts and other 

graphics.



Each day, GAO issues a list of newly released reports, testimony, and 

correspondence. GAO posts this list, known as “Today’s Reports,” on its 

Web site daily. The list contains links to the full-text document 

files. To have GAO e-mail this list to you every afternoon, go to 

www.gao.gov and select “Subscribe to daily E-mail alert for newly 

released products” under the GAO Reports heading.



Order by Mail or Phone:



The first copy of each printed report is free. Additional copies are $2 

each. A check or money order should be made out to the Superintendent 

of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or 

more copies mailed to a single address are discounted 25 percent. 

Orders should be sent to:



U.S. General Accounting Office



441 G Street NW,



Room LM Washington,



D.C. 20548:



To order by Phone: 	



	Voice: (202) 512-6000:



	TDD: (202) 512-2537:



	Fax: (202) 512-6061:



To Report Fraud, Waste, and Abuse in Federal Programs:



Contact:



Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: fraudnet@gao.gov



Automated answering system: (800) 424-5454 or (202) 512-7470:



Public Affairs:



Jeff Nelligan, managing director, NelliganJ@gao.gov (202) 512-4800 U.S.



General Accounting Office, 441 G Street NW, Room 7149 Washington, D.C.



20548: