a b s t r a c t

Study objective: We describe emergency physician staffing, capabilities, and academic practices in US Veterans Health Administration emergency departments (EDs).

Methods: As part of an ongoing process improvement effort for the VHA emergency care system, VHA-wide sur- veys are conducted among ED medical directors every 3 years. Web-based surveys of VHA ED directors were con- ducted in 2013 on clinical operations and academic program development. We describe the results from the 2013 survey. When available, we compare responses with the previously administered survey from 2010.

Results: A total of 118 of 118 ED directors filled out the survey in 2013 (100% response rate). Respondents reported that 45.5% of VHA emergency physicians are board certified in emergency medicine, and 95% spend most their time in direct patient care. Clinical care is also provided by part-time (b 0.5 full-time employee equivalent) emergency physicians in 59.3% of EDs. More than half of EDs (57%) provide on-site tissue plasminogen activator for acute ische- mic stroke patients, and only 39% can administer tissue plasminogen activator 24 hours per day,7 days per week. Less than half (48.3%) of EDs have emergency Obstetrics and Gynecology consultation availability. Most VHA EDs (78.8%) have a university affiliation, but only 21.5% participated in the respective academic emergency medicine program. Conclusions: Veterans Health Administration emergency physicians have primarily clinical responsibilities, and less than half have formal emergency medicine board certification. Despite most VHA EDs having university affiliations, traditional academic activities (eg, teaching and research) are performed in only 1 in 3 VHA EDs. Less than half of VHA EDs have availability of consulting services, including advanced stroke care and women’s health.

Introduction

The Veterans Health Administration (VHA) is the largest integrated health care system in the United States. In 2014, the VHA spent $46 billion for veteran medical care across the United States in 23 regional networks, called Veterans Integrated Service Networks (VISNs) [1]. Throughout the United States, there are 118 VHA hospitals that had EDs providing

? Conflict of interest: none.

?? Grants/financial support: The project described was supported by award number K12

HL109019 from the National Heart, Lung, and Blood Institute. The content is solely the re- sponsibility of the authors and does not necessarily represent the official views of the Na- tional Heart, Lung, and Blood Institute or the National Institutes of Health.

? Authors report no potential, perceived, or real conflict of interest and do not have any

additional disclosures.

* Corresponding author. Tel.:+1 615 936 8379; fax: +1 615 936 3754.

E-mail address: [email protected] (M.J. Ward).

around-the-clock care for patients presenting with emergencies. The VHA defines an emergency department (ED) as providing 24 hours per day, 7 days per week resuscitation and stabilization of life-threatening emergencies and should be staffed and equipped to provide initial evalu- ation, treatment, and disposition of a broad spectrum of medical condi- tions regardless of severity [2]. For fiscal year 2008, there were

1.8 million ED visits increasing by 33% to 2.3 million ED visits in fiscal year 2013.

The quality of acute care delivery in VHA hospitals has been a source of discussion for more than 2 decades. In 1989, an American College of Emergency Physicians Task Force on Military Emergency Medicine identified the quality and access to emergency care for veterans as a critical issue [3]. This Task Force recommended that standards for veter- an emergency care should be commensurate with civilian standards. Specifically, they recommended that VHA EDs be staffed with board- certified/prepared emergency physicians [3-5]. However, because of the deemphasized role of emergency care within the VHA [6], these

http://dx.doi.org/10.1016/j.ajem.2015.03.062 0735-6757/

recommendations were not fully implemented. In 2006, VHA leadership changed this approach and issued a directive mandating that VHA facil- ities with medical-surgical beds, and an intensive care unit must pro- vide ED availability comparable with civilian standards [7]. Coinciding with this effort, the VHA increased the hiring of board-certified emer- gency physicians [8]. In this article, we describe VHA EDs in 2013, with a focus on service availability, staffing, and how VHA EDs partici- pate in academic medicine, including emergency medicine research and education.

Methods

Study design, setting, and population

We describe cross-sectional data from the VHA’s emergency medi- cine national survey of medical directors from 2013. Surveys of all VHA ED medical directors are conducted approximately every 3 years. These surveys are conducted by the VHA’s emergency medicine leader- ship, which consists of the National Emergency Medicine Field Advisory Committee, in conjunction with the VHA’s National Director of Emer- gency Medicine and the Healthcare Analysis and Information Group. The survey covers many topics; however, for this project, we focused on qualifications of the emergency medicine physicians, service capabil- ities, and academic practices of VHA EDs. The prior survey was adminis- tered in 2010 and formed the basis for the 2013 survey with many of the same questions. A full description of the development and dissemina- tion of prior surveys is described elsewhere [8].

To develop the 2013 survey, VHA’s Emergency Medicine Field Advi- sory Committee worked with the VHA’s National Director of Emergency Medicine and the VHA’s Healthcare Analysis and Information Group to adapt the 2010 survey and to identify priority areas for the 2013 survey. Compared with the 2010 survey, the 2013 survey added new questions about the care of female veterans and those with emergent mental health needs. Questions also were added about emergency care physi- cian certification and the academic activities of each department.

The Healthcare Analysis and Information Group is responsible for in- ternal VHA program reviews and consists of data analysts experienced in survey design and development. The Healthcare Analysis and Infor- mation Group met weekly with the VHA Field Advisory Committee to refine survey questions and generated a preliminary list of questions for the 2013 survey. The 2013 survey was pilot tested by 6 volunteer sites to ensure clarity of questions and to determine the time to com- plete the survey. Minor revisions were made after pilot testing. The final survey was distributed by Inquisite (Allegiance Software Inc, South Jordan, UT), a Web-based online survey. An e-mail link was sent to the regional VISN offices, which then disseminated the surveys to each of the facility directors. Each VISN office ensured completion by the ED director or individuals with equivalent responsibility (eg, chief of staff, nurse manager, or chief of medicine). The survey was open for response between July 18, 2013 and July 31, 2013.

All facilities were mandated by the National Emergency Medicine Director’s Office to complete this survey. Responses were reviewed by the National VHA Emergency Medicine Field Advisory Committee to en- sure they appropriately matched the questions. When potential errors were identified (eg, responses not congruent were earlier answers), an- alysts from the Healthcare Analysis and Information Group contacted survey respondents to verify responses.

All facilities that provide Emergency medical care through an ED were included in the online survey. Although these EDs could have an urgent care center colocated with the ED, if no ED was available (ie, they were solely an urgent care center), the facility was excluded from our analysis. For facilities with both an ED and urgent care center, respondents were asked to provide answers for their ED only. Re- sponses were self-reported and unless there was incongruence with prior responses were not independently verified as to their accuracy.

The full survey was a 104-item, 46-question cross-sectional survey of VHA medical facilities offering acute, unscheduled care (ie, both EDs and urgent care centers). The 2013 survey focused on 8 content areas:

organizational structure of the ED, staffing, and shift assignments;

recruitment and compensation of physicians to staff the ED;

(3) ED practices including structural elements of the ED, policies, and privileges; (4) care of the female veteran including equipment and ser- vices available; (5) care of the veteran with emergent mental health needs; (6) ED security; (7) ED support services including pharmacist and social worker staffing; and (8) programmatic and academic devel- opment. The authors selected specific topic areas and questions from the survey relevant to emergency physician staffing, service capabilities, and academic program development. Not all results are reported (eg, ED security). The full survey and all questions can be seen in the Appendix. This study was approved by the VA Tennessee Valley healthcare services Nashville campus Institutional Review Board.

Data analysis

These data were collected and processed by the VHA Office of Healthcare Analysis and Information Group. Standard descriptive statis- tics were used. Where identical questions were asked in both the 2010 and 2013 surveys, we compared differences in responses between the 2 periods. We calculated the difference in sample proportions between 2 populations using 2-tailed probabilities and report the 95% confidence interval of the difference using difference in proportion calculators [9].

Results

In the 2013 survey, 118 facilities met criteria to be included as an ED. An additional 23 urgent care centers were surveyed but excluded from our analysis because they did not have an ED. There was a response rate of 100% (118/118) for facilities with any type of ED (ie, ED alone or ED plus urgent care center). There was also a 100% (117/117) response rate for the 2010 survey. Details of survey responses are summarized in the Table.

Organizational structure

Compared with 2010, more VHA EDs had separate departmental sta- tus, but this result was not significant and still represented less than one-quarter of all VHA EDs. Most VHA EDs had an assigned ED medical director who was likely to be board certified in emergency medicine. Emergency department bed size remained largely unchanged, and 4 fa- cilities nationally provided pediatric care (<=16 years old).

Physician staffing, qualifications, and recruitment

Of the 118 EDs, most used part-time emergency physicians (b 0.5 full-time employee equivalent), an increase compared with 2010. Most VHA EDs also used nurse practitioners and physician assistants. Among 737 full-time VHA emergency physicians (>=0.5 full-time em- ployee equivalent), survey respondents identified less than half as board certified in emergency medicine with a mean annual salary of

$208891 for board certified/eligible emergency medicine physicians. Most full-time VHA emergency physicians provided direct patient care with little time spent on research, education, and administration.

When asked about barriers to physician recruitment, ED directors identified inadequate salary as the most important barrier to recruit competent and qualified emergency physicians for their ED. The num- ber of working hours required for full-time emergency physicians as well as geographic location was identified as the next most important barriers to effective recruitment.

Table

Summary of 2013 survey responses for the categories of organization structure, emergency physician staffing and qualifications, service availability, and academic and programmatic development

2010 (n = 117)			2013 (n = 118)	Difference (95% CI)		P
Organizational structure Departmental status	17	14.5%	27	22.9%	8.4% (-1.70%, 18.2%)	.10
Assigned medical director	–		104	88.1%
Board certified	–		97	82.2%
Mean no. of ED beds	12		13
Provides pediatric care Emergency physician staffing, qualifications, and recruitment Use part-time physicians	– 49	41.9%	4 70	3.4% 59.3%	17.4% (4.70%, 29.5%)	.008
Use NP or PAs	–		80	67.8%
No. of full-time physicians	–		737	–
Board certified in EM	–		335	45.5%
Mean salary for BC/BE physicians Medical directors identified barriers to physician recruitment Inadequate salary	– –		$208891 108	91.5%
No. of working hours	–		70	59.3%
Geographic location Service availability Can provide tPA on-site	– –		70 67	59.3% 56.8%
Can provide tPA 24/7 Emergency OB/GYN care Emergent GYN consultations available	– –		47 57	39.8% 48.3%
GYN consultations provided on site	–		47	39.8%
Emergent OB consultations available	–		28	23.7%
OB consultations provided on site	–		14	11.9%

Emergency mental health care Involuntary patients

Admit to same facility	–	90		76.3%
Transfer within VISN	–	7		5.9%
Transfer to non-VHA community hospitals	–	21		17.8%
Voluntary patients
Admit to same facility	–		97	82.2%
Transfer within VISN	–		18	15.3%
Transfer to non-VHA community hospitals	–		3	2.5%
Programmatic and academic development
University affiliation	–		93	78.8%
Participate in affiliated academic EM program	–		20	16.9%
Academic affiliates declined participation	–		23	19.5%
Emergency medicine residents rotate	12	10.3%	16	13.6%	3.3% (-5.17%, 11.8%)	.43
Any physician with peer-reviewed publication (past 5 y)	–		44	37.3%

Abbreviations: CI, confidence interval; NP, nurse practitioner; PA, physician assistant; EM, emergency medicine; BC/BE, board certified/board eligible; 24/7, 24 hours per day, 7 days per week; OB, obstetrics; GYN, gynecology.

Service availability

More than half of EDs could provide tissue plasminogen activator on site for the treatment of acute ischemic stroke. Among these fa- cilities, most could provide tPA 24 hours per day, 7 days per week. Over- all, slightly more than one-third of VHA EDs could provide 24 hours per day, 7 days per week tPA administration for acute ischemic stroke pa- tients who presented to VHA EDs requiring the remainder required transfer to another hospital for care (ie, VHA and non-VHA facilities).

Emergent gynecologist consultations were available in nearly half of VHA EDs, with less than a quarter of these consultations occurring in person 24 hours per day, 7 days per week. For EDs without such capabil- ities, nearly all transferred locally 24 hours per day, 7 days per week. Fewer emergent obstetrics consultations were available in VHA EDs, and less than a quarter provided on-site care 24 hours per day, 7 days per week. The remainder provided telephone support for patients 24 hours per day, 7 days per week.

For patients with emergent mental health needs, more than three- quarters of VHA EDs admit involuntary mental health patients to the same VHA facility. For those patients not admitted to the same facility, approximately one-quarter are transferred to non-VHA community fa- cilities. Similarly, for voluntary patients, most VHA EDs admit to the same facility, and approximately one-quarter are transferred to either VHA or non-VHA community facilities.

Programmatic and academic development

A university affiliation was present in most VHA EDs. Among the university-affiliated EDs, slightly more than 1 in 5 participated in the affili- ated academic emergency medicine program. However, nearly a quarter of academic affiliates declined to participate with their respective VHA EDs. The proportion of VHA ED sites that rotate emergency medicine residents minimally changed between 2010 and 2013. Finally, in the preceding 5 years, few VHA EDs have had any staff publish peer-reviewed literature.

Discussion

The results of this survey including every VHA ED in the United States provide a comprehensive description of the staffing, capabilities, and academic practices in EDs across the VHA in 2013. At a time when improving veteran access to VHA services is one of the VHA agency’s strategic goals [10], high-quality emergency care can play an instru- mental role in implementing this strategy. Our analysis of this survey has 5 main findings. First, we found an underlying theme that VHA emergency medicine is in the process of developing a model for the de- livery of emergency care where patients with certain diseases can be treated by VHA facilities, whereas some patients are transferred to other facilities (both VHA and non-VHA) for specialty care. Second, al- though VHA emergency medicine resembles civilian emergency

medicine in many ways, their approach also has features of an academic ED. Despite most facilities having academic affiliations, clinicians pri- marily provide clinical care and perform virtually no teaching or re- search. Third, many VHA EDs are staffed with part-time physicians, and less than half of full-time physicians are board certified in emergen- cy medicine. Fourth, when compared with non-VHA ED physicians with similar nonacademic responsibilities using market surveys, those staffing VHA EDs are paid substantially less than non-VHA emergency physicians. Finally, if challenges in the existing relationships can be overcome, the VHA academic affiliations represent a substantial oppor- tunity for growth and development through the identification of quali- fied physicians and enrichment of provider responsibilities (ie, teaching and research).

This survey has identified several areas of change over time in the in- frastructure and makeup of VHA EDs. Specific changes include more VHA EDs having departmental status; more part-time emergency phy- sicians; and among those who are full time, more board certification. Each of these is discussed below.

More frequent departmental status not only recognizes the in- creased prominence of emergency medicine within the VHA, but it pro- vides more flexibility and administrative independence to make Operational decisions. Interestingly, the size of VHA EDs, as measured by the number of ED beds, has changed minimally despite the rising number of ED visits from 2010 to 2013. Downstream consequences of this growth should be investigated to determine its impact on common ED operational metrics.

We found that most VHA EDs do not have immediate, around- the-clock access to tPA for acute ischemic stroke, radiologic services for obstetric and gynecologic ultrasound, or on-site mental health beds. The lack of these services often results in interfacility transfers. Although immediate access to tPA is one of the core strategies to re- duce treatment times for patients with acute ischemic stroke [11] as identified in this survey, VHA EDs frequently do not have tPA avail- able, which may prevent timely treatment. This lack of immediate tPA access may help explain why veterans with acute ischemic stroke have lower treatment rates than their civilian counterparts [12]. Also noteworthy is that despite the reliance on transfers, there are no existing VHA process measures to monitor interfacility transfer performance.

Beyond infrastructure, the qualifications of the emergency physi- cians who staff VHA EDs can influence the quality of health care de- livery and access to care. We found that there is an increased reliance upon part-time emergency physicians to provide clinical care. Although human resource flexibility can be an important oper- ational strategy to respond to the changing operational demands of the ED [13], they receive little training and may have less experience and familiarity with a facility’s operational policies, culture, and practice [14,15]. Temporary ED staff have also been found to affect patient safety in the ED. Temporary staff resulted result in more harmful Medication errors than those caused by permanent ED staff [16]. Thus, this approach may have long-term drawbacks for pa- tient flow and operational improvement.

Full-time emergency physicians, on the other hand, were more fre- quently board certified when compared with 2010. A prior VHA study from 1993 identified only 21 (18.9%) of 111 of VHA EDs having “any” board-certified/board-eligible emergency physicians [17]. Although this highlights substantial progress, a 46% rate of board certification is still lower than the 57% seen in civilian settings [18]. The clinical and di- agnostic skills of a residency-trained, board-certified emergency physi- cian have been found to be associated with higher quality care and protective against adverse event outcomes [19]. Specifically, incorporat- ing point-of-care ultrasound into VHA emergency medicine clinical practice, which is now standard at many institutions and in residency training [20], could be a strategy to decrease the need for transfers for formal emergent obstetric and gynecologic ultrasounds [21]. Thus, credentialing emergency physicians to perform point-of-care

ultrasound in VHA EDs could be a strategy, which potentially decreases the need for transfers for formal emergent obstetric and gynecologic ultrasounds.

The lower rates of board certification may be a result of barriers to recruiting qualified physicians. The 2013 VHA survey identified salary and compensation as a primary barrier to recruiting qualified emergen- cy physicians. Compared with results from a national physician com- pensation survey, VHA emergency physicians had mean salaries in our survey ($208891) comparable with academic organizations from the compensation survey ($208000) rather than those employed by health care organizations ($283000) [22]. However, the clinical responsibili- ties of a VHA emergency physician are more consistent with a nonaca- demic role, which traditionally has higher compensation, rather than an academic emergency physician. Yet, VHA emergency physician com- pensation is more consistent with the lower academic rate. Since this VHA survey was completed, the VHA increased salaries for emergency physicians in September 2014.

The existing academic affiliate relationships at more than three- quarters of VHA EDs provide rich infrastructure for increasing the quality of emergency physicians and enhancing the academic mis- sion of the VHA. Considering the primarily clinical role of VHA emer- gency physicians, using existing academic affiliate relationships could provide opportunities for teaching and research. Diversifica- tion of physician responsibilities may not only attract a broader set of emergency physicians but may also increase employee satisfac- tion and, subsequently, retention [23]. The existing infrastructure of the academic affiliates could also be used to increase the number of emergency medicine residents rotating through VHA EDs. More rotating residents would provide them with exposure to VHA emer- gency medicine as a possible career path potentially increasing the pool of qualified physicians for VHA EDs. However, considering that a quarter of VHA ED academic affiliates declined participation with their VHA counterparts, these relationships warrant further exami- nation to understand why this has occurred and what opportunities exist to improve these relationships.

Limitations

Our study has several limitations. This survey was conducted as part of an ongoing process improvement effort and was not designed for re- search. Furthermore, although these results reflect facility-reported data, there was no independent verification of accuracy beyond the evaluation of incongruent results with 2013 survey data. In addition, this survey only included VHA EDs, and similar data were not collected from non-VHA EDs as a comparison. Finally, our survey only considered physicians and did not consider nurse and technician staffing or other aspects of ED care and structure, which are vital for the efficient opera- tion of an ED.

Conclusions

Veterans Health Administration EDs incorporate components of both academic and community models, with a reliance upon interfacility transfer to access capabilities not available within its system. Comparing 2010 with 2013, VHA EDs are more likely to have departmental status, use part-time emergency physicians, and have more board-certified emergency physicians. However, the presence of board-certified emergency physicians still lags civilian counterparts.

Acknowledgments

None.

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ajem.2015.03.062.

References

1. Office of Budget. 2014 VA performance and accountability reportIn: Department of Veterans Affairs, editor. ; 2014.
2. Veterans Health Administration. VHA directive 2010-010: standards for emergency department and urgent care clinic staffing needs in VHA facilitiesIn: Department of Veterans Affairs, editor. ; 2010.
3. American College of Emergency Physicians. Military emergency medical systems. Ann Emerg Med 1989;18(2):214-21.
4. Hamilton G. Military sponsorship of resident training in civilian programs. Academic News and Views; 1989.
5. Ocrant I, Mallory D, Moore R. The relationship between emergency room patient di- agnoses and medical officer of the day training in a small Military hospital. Mil Med 1984;149(7):366-8.
6. Millard WB. Emergency medicine in the VA: the battleship is turning. Ann Emerg Med 2008;51(5):632-5.
7. Veterans Health Administration. VHA directive 2006-051, standards for nomencla- ture and operations in VHA facility emergency departmentsIn: Department of Veterans Affairs, editor. ; 2006.
8. Kessler C, Chen J, Dill C, Tyndall G, Olszyk MD. State of affairs of emergency medicine in the Veterans Health Administration. Am J Emerg Med 2010;28(8):947-51.
9. VassarStats. Website for statistical computation, the confidence interval for the dif- ference between two independent proportions. http://vassarstats.net/prop2_ind. html. (Accessed September 26, 2014).
10. Department of Veterans Affairs FY 2014-2020 strategic plan[Online]. Accessed November 11, 2014 ; 2014.
11. Xian Y, Smith EE, Zhao X, Peterson ED, Olson DM, Hernandez AF, et al. Strategies used by hospitals to improve speed of tissue-type plasminogen activator treatment in acute ischemic stroke. Stroke 2014;45(5):1387-95.
12. Keyhani S, Arling G, Williams LS, Ross JS, Ordin DL, Myers J, et al. The use and misuse of thrombolytic therapy within the Veterans Health Administration. Med Care 2012; 50(1):66-73.
13. Ward MJ, Ferrand YB, Laker LF, Froehle CM, Vogus TJ, Dittus RS, et al. The nature and necessity of operational flexibility in the emergency department. Ann Emerg Med 2015;65(2):156-61.
14. Needleman J, Buerhaus P, Mattke S, Stewart M, Zelevinsky K. Nurse-staffing levels and the quality of care in hospitals. N Engl J Med 2002;346(22):1715-22.
15. May JH, Bazzoli GJ, Gerland AM. Hospitals’ responses to nurse Staffing shortages. Health Aff (Millwood) 2006;25(4):W316-23.
16. Pham JC, Andrawis M, Shore AD, Fahey M, Morlock L, Pronovost PJ. Are temporary staff associated with more severe emergency department medication errors? J Healthc Qual 2011;33(4):9-18.
17. Young GP. Status of clinical and academic emergency medicine at 111 veterans af- fairs medical centers. Ann Emerg Med 1993;22(8):1304-9.
18. Ginde AA, Sullivan AF, Camargo Jr CA. National study of the emergency physician workforce, 2008. Ann Emerg Med 2009;54(3):349-59.
19. Wilson M, Welch J, Schuur J, O’Laughlin K, Cutler D. Hospital and emergency depart- ment factors associated with variations in missed diagnosis and costs for patients age 65 years and older with acute myocardial infarction who present to emergency departments. Acad Emerg Med 2014;21(10):1101-8.
20. American College of Emergency Physicians. emergency ultrasound guidelines. Ann Emerg Med 2009;53(4):550-70.
21. Moore C, Todd WM, O’Brien E, Lin H. Free fluid in Morison’s pouch on bedside ultra- sound predicts need for operative intervention in suspected Ectopic pregnancy. Acad Emerg Med 2007;14(8):755-8.
22. Medscape. Emergency medicine physician compensation report: 2013. http://www. medscape.com/features/slideshow/compensation/2013/emergencymedicine; 2013. [Accessed September 26, 2014].

    Hackman JR, Oldham GR. Motivation through the design of work: test of a theory. Organ Behav Hum Perform 1976;16(2):250-79.