a b s t r a c t

Objective: The objective of the study is to examine the effect of the opening of a freestanding emergency department (FED) on the surrounding emergency medical services (EMS) system through an examination of EMS system metrics such as ambulance call volume, ambulance Response times, and turnaround times.

Methods: This study is based on data from the county’s computer-aided dispatch center, the FED, and the Maryland Health Services Cost Review Commission. The analysis involved a pre/post design, with a 6-month washout period. The preintervention period was April to October 2010, and the postintervention period was April to October 2011. Data were analyzed using standard t tests.

Results: The average daily number of EMS-related calls received in the computer-aided dispatch center was lower after the FED opened (16.3 [95% confidence interval {CI}, 15.7-16.9] vs 15.8 [95% CI, 14.9-16.9]). One-fourth of all patients were transported by ambulance to the FED after it opened. Use of the FED and adjacent hospitals increased by 8647 visits (15.8%) during the study period. Turnaround time for the county’s ALS units decreased from 26.8 (95% CI, 26.2-27.5) to 25.1 (95% CI, 24.3-25.8) minutes. The ambulance out-of-service interval decreased from 87.3 (95% CI, 86.0-88.5) to 81.1 (95% CI, 79.7-82.4) minutes. Based on change in out-of-service this study had a small effect size (Cohen’s d = 0.33).

Conclusions: The opening of an FED was associated with a modest improvement in time-specific EMS system metrics: a decrease in ambulance turnaround time and shorter out-of-service intervals.

(C) 2015

Introduction

Emergency departments (EDs) play a critical role in the US health care system, as primary locations for unscheduled hospital admissions

? Meetings: This work was presented as a poster at the 2013 Scientific Assembly of the American College of Emergency Physicians, held in Seattle, Washington.

?? Conflicts of interest: The authors have no potential conflicts of interest to report.

? Author contributions: BJL, JMH, and BJB conceived the study. BJL, JMH, LP, and RA de-

signed the study. JK and MAV provided access to data and conceptual input for the manuscript. AC analyzed the data. BJL, JMH, and CT drafted the manuscript, and all authors contributed substantially to its revision. BJL takes responsibility for the manuscript as a whole.

* Corresponding author at: Department of Emergency Medicine, University of Maryland School of Medicine, Baltimore, MD.

E-mail addresses: [email protected], [email protected] (B.J. Lawner).

and complex diagnostic evaluations [1,2]. In 2010, approximately 130 million visits to EDs were documented, and this number continues to climb [3]. Within the current health care environment, EDs are strug- gling to meet the demands of increasing patient age, volume, and acuity [4]. Emergency department crowding interferes with the timely trans- fer of patients from prehospital crews to hospital personnel, which could delay access to stabilization and Diagnostic procedures [4].

A new model designed to address the increasing demand for acute care is the creation of freestanding emergency departments (FEDs), which are being integrated into emergency care systems across the United States [5,6]. In theory and in practice, these facili- ties constitute a valuable addition to emergency care resources; however, their role and function within health care systems in gen- eral and emergency medical services (EMS) systems specifically have not yet been defined.

http://dx.doi.org/10.1016/j.ajem.2015.11.042

0735-6757/(C) 2015

Although there is no agreed-upon national definition for FEDs, they can be defined operationally as health care facilities that pro- vide emergency care and are not physically connected to an acute care hospital. In Maryland, where this study was conducted, state law defines a freestanding medical facility as facility (a) in which medical and health services are provided, (b) that is physically sepa- rated from a hospital or hospital grounds, and (c) that is an adminis- trative part of a hospital or related institution [7]. Also according to the Maryland statute, these facilities must be open 24 hours per day, 7 days per week.

The purpose of FEDs is to provide initial diagnostic and stabilizing interventions to victims of acute illness or injury, in accordance with the requirements outlined in the American College of Emergency Physicians’ Emergency Department Planning and Resource Guide[7]. freestanding EDs generally do not have the ability to admit patients or perform complex procedures. Because FEDs lack inpatient beds, they accommodate sicker patients by transferring them to hospitals with appropriate resources. These transfers typically occur through a collaborative arrangement with a local ambulance company or commercial EMS provider [6,8].

Theoretically, the opening of an FED should improve the metrics of the surrounding EMS system and ambulance availability while enhanc- ing patient access to acute care services. This study examined the impact of a FED through an examination of EMS-specific time intervals such as ambulance call volume, ambulance response time, and ambu- lance out-of-service interval. We hypothesized that construction of the FED would be linked to an increase in overall EMS system efficiency as characterized in terms of system response metrics.

Methods

Study population

The study population includes patients who presented or were transported by ambulance to the University of Maryland Shore Emer- gency Center at Queenstown (SEC) on the Eastern Shore of Maryland FED between April 2010 and October 2011. The FED provides emergen- cy health care services to several counties. The FED is physically located in Queen Anne’s County which covers 372 square miles and has a pop- ulation of 48517 [9]. No other EDs are located within 20 miles of the FED (Fig. 1). Eight volunteer ambulance squads operate in the county, and the local government funds 4 full-time advanced life support units. emergency medical services system metrics, such as response time, are examined for units that transported to the FED as well as other area hospitals.

History of the Queen Anne’s County FED

Efforts to enhance access to Emergency medical care in Queen Anne’s County intensified in 2007, after several major traffic crashes and extensive construction projects on the Chesapeake Bay Bridge closed the west-bound span of the bridge for extended periods. As a demonstration project for freestanding emergency centers in rural areas of the state, the Maryland General Assembly adopted legislation that regulates and licenses the FED.

The FED opened to receive patients on October 4, 2010. It is a full- service emergency facility open 24 hours a day, 7 days a week, 365

Fig. 1. Location of SEC and surrounding EDs.

days a year, with full diagnostic imaging and laboratory services and the ability to care for all levels of Medical emergencies. According to the 2010 edition of the Maryland Medical Protocols for EMS Providers, FEDs can receive patients at priority 3 or 4 (less acute) as well as priority 1 pa- tients who have an unsecured airway or are in extremis, requiring stabi- lization beyond the capability of the EMS crew [10]. When a patient’s condition requires a higher level of care, the center requests EMS trans- port to a hospital via ambulance or medical helicopter. A commercial ambulance is stationed at the FED for the sole purpose of expediting patient transfers to other health care facilities; this vehicle does not respond to community requests for emergency medical services.

Data management

Several sources were tapped for data used to assess EMS efficiency and hospital utilization. Data from the county’s EMS computer-aided dis- patch (CAD) center were used to calculate EMS metrics. Intervals obtain- ed from the CAD system included time between receipt of a 9-1-1 call and the arrival of the first transport vehicle at the scene, transport time, and turnaround time (defined below).

Hospital utilization metrics included ED volumes in the 3 hospitals within the general vicinity of the FED: University of Maryland Shore Medical Center at Easton, University of Maryland Shore Medical Center at Chestertown, and Anne Arundel Medical Center, and in the FED as well as patient disposition from the FED. The sources were University of Maryland Medical System’s Shore Health Data for the FED and the Maryland Health Services Cost Review Commission for the hospitals.

Ethics approval

This research was exempted by the institutional review board at the University of Maryland, Baltimore.

Data analysis

A pre/post design was used for this study. The preintervention peri- od was April to October 2010. A washout period lasted from October 2010 to April 2011; data from this time were not included in the analy- sis to allow for system-wide penetration of the new FED transport pro- tocol. The postintervention period, therefore, began in April 2011 and ended in October 2011. The primary outcome of interest was change in mean ambulance turnaround interval (time from arrival at a hospital to departure from the hospital). Secondary outcomes included changes

in other mean ambulance time metrics, such as average ambulance re- sponse interval (time from ambulance dispatch to arrival at the scene) and overall out-of-service interval (the amount of time that an ambu- lance is not available to respond to another incident). The mean preintervention and postintervention time intervals were compared using a standard t test. The overall annual ED census for the FED was documented, as were the dispositions of patients from the FED. Specific information of interest included the number of patients transferred to other facilities and their original Mode of transport to the FED (ambula- tory versus ambulance). The preintervention and postintervention ED volumes for the surrounding hospitals were also tallied.

The CAD data did not reliably identify ambulances that transported directly to the FED-individual patient priority (1, 2, 3, or 4) was not re- corded and the destination was not reported consistently. Computer- aided dispatch data were matched probabilistically with FED patient data by date and time of hospital arrival to determine time intervals for ambulances that arrived at the FED. An ambulance arriving within 30 minutes before or 30 minutes after the arrival time documented by the FED was considered a potential match. From potential matches, the ambulance with the closest time and no other hospital noted as the destination was designated a definitive match. The mean time inter- vals for this group of ambulances arriving at the FED were compared to the time intervals of ambulances determined to not have traveled to the FED using the Student t test. The CONSORT diagram in Fig. 2 provides a visual reference for how Ambulance transports were recorded.

Results

There were a total of 10233 EMS incidents and 1897 hospital trans- ports logged into the CAD system to during the preintervention period of the study. During the postintervention period, 11022 incidents and 1722 hospital transports were logged. The average number of daily EMS incidents decreased from 16.3 (95% confidence interval [CI], 15.7- 16.9) in the preintervention period to 15.8 (95% CI, 14.9-16.9) per day in the postintervention period. Paralleling this trend was a concurrent decrease in the number of daily incidents that resulted in patient trans- port (5.9 [95% CI, 5.7-6.0] vs 5.7 [95% CI, 5.5-5.9]). Response intervals, that is, the time between receipt of a 9-1-1 call and the arrival of the first EMS vehicle and the first EMS transport vehicle on the scene, did not differ significantly. The transport interval, defined as the time from scene departure to arrival at a hospital, was significantly different following the opening of the FED (22.2 minutes in the preintervention period [95% CI, 21.7-22.7] vs 19.9 minutes in the postintervention

Fig. 2. Consort diagram.

Table 1

Queen Anne’s County EMS system ambulance metrics before and after opening of SEC

	Response interval	Transport interval	Turnaround time	Back-in-service interval	Out-of-service interval
Preintervention average time (min)	7.8	22.2	26.8	18.6	87.3
Postintervention average time (min)	8.3	19.9	25.1	14.0	81.1
P	.81	b.01	b.01	b.01	b.01

period [95% CI, 19.4-20.4]) (Table 1 and Fig. 3). The turnaround time was lower in the postintervention period (26.8 [95% CI, 26.2-27.5] vs 25.1 [95% CI, 24.3-25.8] minutes). Finally, the out-of-service interval (in- corporating the response, turnaround, and transport intervals) de- creased from 87.3 (95% CI, 86.0-88.5) to 81.1 (95% CI, 79.7-82.4)

minutes.

Time intervals were affected more significantly for ambulances transporting patients to the FED as opposed to area hospitals. Of the 1722 ambulance transports in Queen Anne’s County during the postin- tervention period, 419 (24.3%) went to the FED. The matching analysis identified 333 (79.5%) of 419 ambulance transports arriving at the FED between April 4, 2010, and October 10, 2011. Transport intervals aver- aged 12.0 minutes (95% CI, 11.2-12.7) for ambulances arriving at the FED compared with 21.9 minutes (95% CI, 21.3-22.5) for other desti- nations (Table 2 and Fig. 3). The out-of-service interval for ambu- lances arriving at the FED was 59.4 minutes (95% CI, 56.5-62.4), compared with 86.4 minutes (95% CI, 85.0-88.0) for all ambulances in the preintervention period. The turnaround time lasted 18.6 mi- nutes (95% CI, 16.4-20.8) for ambulances transporting patients to the FED vs 26.7 minutes (95% CI, 25.9-27.5) for ambulances transporting to alternative destinations. Response intervals between the 2 ambulance destinations did not differ significantly (8.7 [95% CI, 7.9-9.4] vs 8.5 [95% CI, 8.3-8.8] minutes).

Overall use of the FED and of the EDs at the adjacent hospitals in- creased by 8647 visits (15.8%) between the preintervention and postin- tervention periods. The changes at the individual facilities were as follows: +7.3% (by 2338) at Anne Arundel Medical Center, -3.5% (by 542) at University of Maryland Shore Medical center at Easton, and

+ 1.7% (122) at the University of Maryland Shore Medical Center at Chestertown.

Discussion

Despite the growing interest in FEDs as a means of increasing access to emergency medical care in isolated and remote communities, little is known about the interface between these facilities and the local EMS sys- tem. In 1999, Avery [11] published a descriptive, hypothetical study about a rural FED. She examined the costs associated with rural emergency care and mentioned the need for a proximally located “mobile intensive nurse” to effect patient transfers [11]. The study focused

Pre

explicitly on cost modeling and did not examine the impact of the FED on existing demands for EMS. Fulks et al [8] published a descriptive study about the experience of starting a FED to meet “community need.” State regulations mandated the facility to have an on-site ambu- lance capable of providing advanced cardiac life support. The costs of staffing such a unit were found to exceed $320 000. This study did not describe the role of this facility, or its ambulance, in the existing EMS system [8]. Gillespie [12] studied the effects of an FED’s partner- ship with local EMS personnel. The incorporation of EMS personnel (who were not expected to respond to EMS calls from the communi- ty) into FED staffing resulted in a negligible increase in ED length of stay and an 8-minute decrease in “response time.” The response time decrease applied only to pending interfacility transports [12]. Therefore, it is not possible to formulate conclusions about the FED’s impact on system efficiency.

To the authors’ knowledge, our study is the first to link the operation

of an FED to prehospital system metrics. We expected that the availabil- ity of an emergency care center in proximity to a community’s preHospital resources would positively influence system efficiency. We designed the study in an attempt to quantify efficiency through the use of time intervals. Turnaround time and out-of-service intervals decreased after the opening of a community-based FED, but response time was longer than before the FED was available.

Limitations

Several limitations of this SEC impact study, in addition to its retro- spective design, can be cited. First, there are no well-validated metrics related to EMS system efficiency. Although unit utilization hours and response intervals have been cited as Surrogate markers for system per- formance, the link between better response times and patient outcomes is less than clear [13-15]. The authors do not dispute that shortened response times might matter more in the context of time-sensitive illness but reject the metric as a suitable measure of system efficiency. A compre- hensive discussion about the utility of a shortened response interval is beyond the article’s scope. Furthermore, the data reported in this study were not linked to hospital-based patient outcome data. The results focus solely on the association between the emergency center’s opening and commonly reported time intervals. It was technically difficult to match the CAD data with the FED data. Although the strategy of probabi- listic matching should have captured most Patient transports to the FED, it is highly likely that some transports were missed. In addition, the out-of- service intervals were undoubtedly affected by ambulance crews’ deci- sions to occasionally remain at the FED to assist with the provision of care, particularly for critically ill patients. Finally, after the opening of

Response On-Scene Transport

Table 2

Queen Anne’s County EMS system ambulance metrics for ambulances transporting pa- tients to the SEC vs other hospitals

Turnaround Time

Post

Back-in-Service

Response interval

Transport interval

Turnaround time

Back-in- service interval

Out-of- service interval

0 20 40 60 80 100

Minutes

Shore Emergency Center

Other emergency centers

8.7 min 12.0 min 18.6 min 5.4 min 59.4 min

8.5 min 21.9 min 26.7 min 16.1 min 86.4 min

Fig. 3. Queen Anne’s County EMS system interval metrics before and after opening of the SEC.

P .74 b.01 b.01 b.01 b.01

the FED, people in the community might have become more likely to drive themselves or a sick person directly to the medical center, thereby bypassing the EMS system. The number of daily EMS incidents decreased after the FED opened, as did the number of EMS transports. These chang- es, however, should not have affected variables such as Turnaround times and out-of-service intervals.

Sicker patients, in accordance with established medical protocols, were preferentially routed to area hospitals as opposed to the FED. The transport of a less acutely ill cohort of patients may have indeed influenced the abbreviated EMS turnaround times associated with units arriving at the FED. Despite this acknowledged source of bias, any com- prehensive analysis of EMS system efficiency must incorporate institu- tions with the potential to increase ambulance availability. It is understood that the unique demographics of the EMS system under review limits the generalizability of the findings. However, high-performing EMS systems must adapt, conform, and evolve to address community-specific needs. In that regard, the study highlights a collaborative relationship be- tween health care facilities and the surrounding EMS system.

Finally, a study of EMS efficiency does little to highlight the benefits of proximate access to emergency health care. Historically, residents of Queen Anne’s County would have to cross the Bay Bridge to reach an ED. With span in excess of four miles, delays due to weather or traffic could pose significant challenges to a patient in extremis. Anecdotal reports from the local EMS agency, county officials, and patients attest to the FED’s qualitative value.

Conclusion

The construction ofa FED in Queen Anne’s County, Maryland, was asso- ciated with a modest improvement in time-specific EMS system metrics. Specifically, FED operation was concurrent with a decrease in ambulance turnaround and out-of-service intervals. Further research is needed to elu-

cidate the effect of an FED on its surrounding EMS system and to better de-

fine which patients are appropriate for transfer to this type of facility.

References

1. Hirshon J, Alson R, Blunk D, Brosnan D, Epstein S, Gardner A. America’s emergency care environment, a state-by-state report card: 2014 edition. Ann Emerg Med 2014;63(2):97-242. http://dx.doi.org/10.1016/j.annemergmed.2013.11.024.
2. Morganti KG, Bauhoff S, Blanchard JC, Abir M, Iyer N, Smith A, et al. The evolving role of emergency departments in the United States. Santa Monica, CA: Rand Corpora- tion; 2013 http://www.rand.org/pubs/research_reports/RR280.html.
3. CDC. Emergency department visits; 2014.
4. Olshaker JS, Rathlev NK. Emergency department overcrowding and Ambulance diversion: the impact and potential solutions of extendED boarding of admitted patients in the emergency department. J Emerg Med 2006;30(3):351-6. http://dx. doi.org/10.1016/j.jemermed.2005.05.023.
5. Sullivan AF, Bachireddy C, Steptoe AP, Oldfield J, Wilson T, Camargo CA. A profile of freestanding emergency departments in the United States, 2007. J Emerg Med 2012; 43(6):1175-80. http://dx.doi.org/10.1016/j.jemermed.2012.02.078.
6. California Healthcare Foundation. Freestanding emergency departments: do they have a role in California?; 2009 1-23.
7. State of Maryland. Code of Maryland Regulations. http://www.dsd.state.md.us/ comar/getfile.aspx?file=10.07.08.02.htm.
8. Fulks C, McFarlin S, Stolley J. Starting a freestanding emergency center. Nurs Manage

   2000;31(8):41-2.

   United States Census Bureau. State and County Quick Facts. http://quickfacts.census. gov/qfd/states/24/24035.html; 2014.

9. Maryland Institute for Emergency Medical Services. The Maryland Medical Protocols for Emergency Medical Services Providers. http://www.miemss.org/home/LinkClick. aspx?fileticket=_ZlEmf6DuYI=&tabid=106&mid=875; 2010.
10. Avery S. A limited-service Rural hospital model: the freestanding emergency department. J Rural Health 1999;15(2):170-9 [http://www.ncbi.nlm.nih.gov/pubmed/10511753].
11. Gillespie GL, Yap TL, Singleton M, Elam M. A summative evaluation of an EMS part- nership aimed at reducing ED length of stay. J Emerg Nurs 2009;35(1):5-10. http:// dx.doi.org/10.1016/j.jen.2007.10.002.
12. Blancard IE, Doing CJ, Hagel BE, Anton AR, Zygun DA, Kortbeek JB, et al. Emer- gency medical services response time and mortality in an urban setting. Prehosp Emerg Care 2012;16(1):142-51.
13. Blackwell TH, Kline J, Willis J, Hicks J. Lack of association between prehospital response times and patient outcomes. Prehosp Emerg Care 2009;13(4):444-50.
14. Pons PT, Markovchick VJ. Eight minutes or less: does the ambulance response time impact trauma patient outcome? J Emerg Med 2002;23(1):43-8.