a b s t r a c t

Purpose: Our objectives were to determine the frequency of patient transfers to a tertiary care emergency depart- ment (Tertiary ED) due to a lack of radiology services in Rural hospital EDs (rural EDs), and examine the commu- nity and patient attributes that are associated with these transfers.

Methods: This was a retrospective chart review of patients transferred to a Tertiary ED from Rural EDs. Transfers ex- cluded from the study included pediatric patients (age b 18 years old) and patients transferred for trauma surgeon evaluation. Only those patients who were transferred for radiology services were included in the final analysis.

Results: Over a 12-month period, 1445 patients were transferred to the Tertiary ED with 73.8% (n = 1066) of this population being transferred from a Rural ED. Excluding 381 trauma and pediatric patients, 64.3% (n = 685) of pa- tients were transferred from a Rural ED and were included in the study. Of these 685 transfers, 24.5% (n = 168) were determined to be due primarily to a lack of a radiology service.

Discussion: Lack of radiology services in Rural EDs leads to numerous patient transfers to the Tertiary ED each year. A disproportionate number of these transfer patients are African American. These transfers place additional financial and social burdens on patients and their families. This study discusses these findings and alternative diagnostic op- tions (ie, telemedicine and ultrasound video transfer) to address the lack of radiology services available in Rural EDs. The use of these alternate diagnostic options will likely reduce the number of patient transfers to Tertiary EDs.

(C) 2015

Introduction

Many Rural emergency departments (EDs) do not have radiology services available 24 hours a day to assist in the diagnosis of emergent conditions. This can be attributed to a lack of equipment, qualified per- sonnel to run the equipment, or qualified personnel to interpret images. (A Rural ED was defined as an ED located outside of a Metropolitan Sta- tistical Area (MSA) [1].) The imaging modalities in question most com- monly are computed tomography (CT) or ultrasound (US). When CT or US are needed to diagnose or refute an emergent condition, patients are often transferred from a Rural ED to a Tertiary ED to obtain the needed imaging. This approach delays diagnosis and treatment while consum- ing valuable resources [2]. Additionally, many patients transferred for diagnostic radiology services ultimately do not need further treatment (ie, a negative result from the diagnostic test) [3]. Thus, the lack of real-time radiology interpretation, technologist support, or equipment availability in Rural EDs can result in higher medical and social costs for the patient [3]. Determining the frequency, reason(s), and ultimate disposition of transfers can help us to better understand this issue and the role that alternative methods, such as telemedicine, may play in

? Presented: Abstract Presentation ACEP Scientific Assembly Research Forum. San Francisco 2011.

* Corresponding author at: 1120 15th St, Augusta GA 30912.

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

reducing the financial and social costs passed on to these patients, many of whom do not have the financial or social capital to pay them. A number of patients are transferred from a Rural ED to the Tertiary ED due to the lack of radiology services every year. What is unknown is the frequency at which these transfers occur, the type of radiology service patients are transferred for, and the characteristics associated

with these patients such as poverty level, insurance status, and race. In this study we sought to determine the frequency of adult, non-

trauma patient transfers from Rural EDs to a Tertiary ED due to a lack of radiology services by conducting a retrospective chart review. Secondary objectives included examining transfers according to payer source, race, and poverty level compared to their respective county and state averages. In addition, we sought to determine how community attributes such as poverty and minority Population rates correlated with the rate of transfers.

Materials and methods

A retrospective chart review was completed on all patients trans- ferred to a Tertiary ED from Rural EDs over a 12-month period (June 1, 2008, through May 31, 2009). The study period was chosen to corre- late with the most recent complete census data. The study was conduc- ted at a high-acuity, urban, Tertiary ED, designated as both a pediatric and adult Level 1 trauma center with multiple residency training pro- grams. Radiology services are available 24 hours a day at the Tertiary

http://dx.doi.org/10.1016/j.ajem.2015.07.050

0735-6757/(C) 2015

Table 1

Percentage of patients transferred by race compared to state and county averages

Race	Transferred	Transferred for lack of a radiological service	County average?	Georgia average??	South Carolina average??
African American	711 (49.2%)	105 (64.0%)	44.8%	38.1%	32.5%
White	691 (47.8%)	59 (35.9%)	53%	61.9%	67.5%
Other	43 (3%)	4 (not included)	2.2% (not included)	(not included)	(not included)
Total	1445	164	100%	100%	100%

* P b 4.298 x 10^-6 comparing Transferred patients to county averages.

?? P b 2.2 x 10^-16 comparing transferred patients to GA and SC averages.

ED. The Tertiary ED volume during the study period was approximately 80,000 patient visits per year from Georgia and South Carolina commu- nities. (Due to its geographic location on the border of Georgia and South Carolina, the Tertiary ED receives patients from Georgia and neighboring South Carolina communities.)

In terms of racial classification the sample consists primarily of whites and African Americans, with small numbers of patients of “other” racial backgrounds. Racial identification was self-reported in a free response for- mat and recorded in patients’ medical records at the Rural ED. This data was also recorded in the medical record at the Tertiary ED. This retrospec- tive study uses the same racial classification categories of the U.S. Census Bureau [4]. In this study there were four patients who self-identified as other than African American or white. Because of the extremely small number, patients in the “other” category were excluded from the analysis and the results include only African American and white patients.

All patient transfers to the Tertiary ED were managed through a cen- tralized communications center and recorded in real time in a hospital database. Data elements in the database include: the reason for the transfer stipulated by the treating physician at the Rural ED, the insur- ance type, the services performed at the Rural ED, and some demo- graphic information about the patient. The initial cohort of patients was derived using a standardized data sheet by a single abstractor who was blinded to the study purpose. A transfer to the Tertiary ED solely for radiology services was determined by examining both the rea- son stipulated for transfer by the treating physician at the Rural ED, and the treatment received at the Tertiary ED. If any procedure or specialty consultation was performed either before or after a radiology evaluation as part of the Tertiary ED evaluation, the transfer was classified as not being primarily for radiology services and excluded from the study.

Patient transfers were grouped by the counties where their respec- tive transferring hospitals were located. Data were collected on all the counties with greater than five patient transfers to the Tertiary ED during the study period. Using the U.S. Census Bureau database, racial demographics, insurance rates, and poverty level for each county where the transferring hospitals are located were collected. These values were then compared to the respective state averages. Data on the patient’s Place of residence was not collected. Because the time pe- riod for this study was from June 1, 2008, through May 31, 2009, the data from the U.S. Census Bureau was collected for 2008 and 2009 and the numbers averaged. Data on the racial makeup of the study counties was derived from the U.S. Census Bureau 2005-2009 American Commu- nity Survey [5]. Insurance status information was collected from the U.S. Census Bureau Small Area Health Insurance Estimates interactive data tool [6], and poverty level data were gathered from the Small Area Income and Poverty Estimates (SAIPE) [7]. The study period was deter- mined to allow access to the best comparative data, as the Small Area Health Insurance Estimates and the Small Area Income and Poverty Estimates are several years old when published. The institutional review board at the Tertiary ED approved the study.

Results

Over the 12-month study period, a total of 1445 patients were trans- ferred to the Tertiary ED. Of these, 1066 (73.8%) were transferred from a Rural ED. After excluding 381 trauma and pediatric patients, a total of 685 patients transferred from a Rural ED remained. Of these transfers, 168 (24.5%) were determined to be due primarily to a lack of a radiology

Fig. 1. Proportion of the population that is minority for each county with the highest percentage of transfers for radiological purposes. The data is normalized to the respective state minority populations. Source: 2005-2009 U.S. Census Bureau American Community Survey.

Fig. 2. Proportion of the population whose Household income is below the national poverty level for each county with the highest percentage of transfers for radiological purposes. The data are normalized to the respective state averages. Source: 2008-2009 U.S. Census Bureau Small Area Income and Poverty Estimates.

service. The leading cause for transfer in this group (71.2%) was the lack of a radiologist to interpret the radiology information in real time, gen- erally a CT scan, at the Rural ED. Lack of US services and/or equipment was the next most common reason for transfer (24.5%). Following transfer to the Tertiary ED, the most common radiology exams performed were CT of the abdomen and pelvis (30.7%), CT of the head (28.2%), CT angiography for pulmonary embolism (9.2%), first- trimester US (9.2%), and US for deep Venous thrombus (6.1%).

A second reviewer, blinded to the purpose of the study, examined a random sample of approximately 50% of the initial sample for accuracy. In 3 cases, there was disagreement between the primary abstractor and the second reviewer. The Cohen ? coefficient for inter-rater agreement was .96, indicating a high level of inter-rater reliability. For the 3 cases

in which the raters’ disagreed, the disagreement was resolved through consensus, and the cases were not removed from the study.

Table 1 presents the racial distribution of patients in the study. The initial sample of transferred patients (1445) was divided nearly equally between African-American (49.2%) and white (47.8%) patients. The ra- cial breakdown of the 168 patients transferred for lack of radiology ser- vices was 105 African Americans, 59 whites, and 4 “others” (3 Hispanics and 1 person of mixed race). The percentage of African American and white patients transferred for a lack of radiology services from the Rural ED was 64.2% and 35.8%, respectively. Ninety-nine patients (60.4%) transferred primarily for lack of a radiology service at a Rural ED were subsequently discharged after radiology evaluation at the Tertiary ED.

Fig. 3. Proportion of the population that is uninsured for each county with the highest percentage of transfers for radiology purposes. The data is normalized to the respective state averages. Source: 2008-2009 U.S. Census Bureau Small Area Health Insurance Estimates.

The county population averages for race (African American ~44.8%, white ~ 53%) were compared to the racial percentages for the sample using a ?² test of significance. The proportion of African Americans in the transferred patient population is significantly higher than the rate of African Americans in the counties they were transferred from (Pb.00001). It is also significantly higher than the proportion of African Americans in Georgia (Pb.00001 and South Carolina (Pb.00001). African-American patients were significantly more likely to be transferred from a Rural ED to the Tertiary ED than white patients. Fig. 1 presents the proportion of minority populations for each coun-

ty included in the study. Of the eleven counties included in the study, 8 have a statistically significant higher proportion of minorities than the state average (Pb .005).

Fig. 2 shows the poverty rate for the sample. All counties included in the study had a statistically significant higher proportion of poverty than the respective state averages (Pb .005).

Insurance rates for the counties included in the study generally approximated insurance rates for the states (Fig. 3).

Patients who were admitted or discharged from the Tertiary ED were analyzed by race and insurance status. The difference in the num- ber of patients admitted or discharged by race was not statistically significant (Table 2).

However, when the difference in the rate at which patients were ad-

mitted or discharged was examined by insurance status, a statistically significant difference was found. Uninsured patients (both African American and white) were significantly more likely to be discharged (Tables 3 and 4).

Discussion

The results of this study both support and expand on the existing literature regarding the issue of patient transfer due to the lack of radio- logical services. Although it is known that transfers due to the lack of ra- diological services occur, this study details the percentage of patients that are affected by the lack of 3 specific resources - machinery, techni- cians, and radiologists. In addition, this study analyses the disposition of transfers and suggests possible ways to reduce the frequency of trans- fers from Rural EDs to a Tertiary ED due to the lack of a radiological ser- vice. Finally, this research provides preliminary data about the demographic makeup of affected patients and suggests avenues for future research.

This retrospective chart review was designed to measure the fre- quency of patients transferred from Rural EDs to a Tertiary ED due to a lack of radiology services. It should be noted that the study has several limitations. First, although the demographic data on the patients who were transferred to the Tertiary ED were collected, demographic data on all patients transferred from the Rural EDs were not. It is possible that patients were transferred to other EDs besides the Tertiary ED. In addition, only the variables of interest rather than all demographic data were collected for the patients who were transferred to the Tertiary ED. It should be noted that additional factors not accounted for in this study such as patient preference could have affected the distribution of the patient population that presented to the Rural EDs subsequently affecting the makeup of the patient population transferred to the Tertiary ED.

Table 3

?² Table, race vs insurance-P = .578

Insurance status	White	African American	Total admitted
Uninsured	23	35	58
Insured	36	70	106
Total	59	105	164

This study has an additional limitation. Of the patients that were transferred to the Tertiary ED due to the lack of radiology services at a Rural ED, a greater percentage were African American. The difference between this percentage and the racial percentages for the counties in- cluded in the study was statistically significant based on a ?² test. There were more African Americans transferred due to the lack of radiology services in a Rural ED then would be expected. While the relationship between the census data and associated Rural EDs patient populations is likely similar, the actual relationship remains unknown. Therefore the study data can only speak to the transferred patient population and cannot be generalized to the larger patient population.

We observed that there was an unexplained Racial disparity in the patients affected by the lack of radiology services in these Rural EDs. Although the study was not designed to provide explanation for the ob- served disparity, we believe some of the differences can be attributed to the racial make-up of the counties where the Rural EDs are located. The counties with the most transfers due to a lack of radiology services had a significantly higher proportion of minorities and impoverished persons when compared to their respective state averages (Figs. 1 and 2). How- ever, this does not provide a full explanation as the percentages for African Americans, although high in these counties, are still below the percentage of white residents in 8 of the 11 counties surveyed.

The question of whether or not African Americans are disproportion- ately impacted by transfers as compared to white patients was one that developed as the data analysis was being conducted rather than one of our original research questions. While the relationship between the census data and associated Rural EDs patient populations is likely simi- lar, the actual relationship remains unknown. This study only examines the disproportionality found based on the patients that were transferred to the Tertiary ED and therefore is only generalizable to this population rather than the larger patient population utilizing the services of these Rural EDs. A study with the primary goal of assessing whether or not a racial disparity exists within the general population of patients that seek care at a Rural ED and are ultimately transferred to a Tertiary ED is needed to conclude whether or not this finding is generalizable to a larger patient population.

While the absolute numbers of patients affected by the lack of re- sources in these Rural EDs is small, each one of these transfers can have a large social and Financial impact on patients and their families. Additional burdens are placed on patients when a Rural ED is unable to provide radiology services. These burdens include the increased cost of care due to a second ED visit, the ambulance services needed to transport the patient to the Tertiary ED and the transportation costs required to return to their home community. The additional costs asso- ciated with these transfers can be significant. These results describe a pattern of health care disparity. Many of these patients and their families are challenged with socioeconomic and other issues, and a transfer to a Tertiary ED for a radiology service not locally available creates additional financial and social burdens [8].

Table 2

Percentage of patients transferred due to lack of radiology services that were admitted compared to the proportion of patients discharged, by race. The differences were not sta- tistically significant (P N .70)

Table 4

?² Table, insurance status vs admission rate-P = .012

	Admitted	Discharged	Admitted	15	50	65
African American	66.2%	62.7%	Discharged	43	56	99
White	33.8%	37.4%	Total	58	106	164

Disposition Uninsured Insured Total

While it is unreasonable and duplicative to require Rural EDs to maintain services similar to Tertiary EDs, some of these disparities can be addressed by using technology.

Telemedicine is one solution to address the lack of resources avail- able in Rural EDs and a possible way to reduce the number of transfers due to a lack of radiology services. The results of this study show that 71.2% of the 168 patient transfers due to lack of radiology services were the result of the unavailability of a qualified professional to read and interpret CT scans. When available, telemedicine provides remote access to radiology reading services in real time. Transmission of real- time US video via commercially available video mobile phones to an available ED physician is a possible way to reduce transfers [9]. This method would allow for a US technician from a Rural ED to be directly in contact with a remote physician who would be able to see the image in real-time and provide guidance to the technician. While Ultra- sound training of rural ED physicians is another option, telemedicine would provide a diagnostic and cost effective alternative. The use of telemedicine for CT scans and Ultrasound video transfer could poten- tially prevent up to 96.2% of all transfers due to lack of radiology ser- vices. Further research is needed to determine if this potential solution will reduce the frequency of transfers from Rural EDs to a Tertiary ED due to a lack of radiology services.

Technology in Rural hospitals has been shown to affect the scope of services provided, in radiology and emergency departments [10]. Though the technology is available, rural hospitals have to justify the cost of the equipment and personnel to provide these services [11]. The demand for imaging has increased over the last three decades and new technology allows one radiologist to offer interpretation coverage for numerous hospitals [10]. This is particularly advantageous for rural

hospitals. This study found that over a 1-year period a number of poten- tially avoidable transfers from Rural EDs to the Tertiary ED occurred. Further research is needed to determine if technological solutions, such as telemedicine, could reduce the frequency of potentially avoid- able transfers from Rural EDs to the Tertiary ED, which in turn could re- duce the additional financial and social burdens associated with these transfers for patients.

References

1. U.S. Department of Health and Human Services. Health information technology, how is rural defined? http://www.hrsa.gov/healthit/toolbox/RuralHealthITtoolbox/Introduction/ defined.html. [Accessed 7/1/2015].
2. Gupta R, Greer S, Martin E. Inefficiencies in a rural Trauma system: the burden of repeat imaging in interfacility transfers. J Trauma Inj Infect Crit Care 2010;69(2):253-5.
3. Goh K, Lam C, Poon K. The impact of teleradiology on the inter-hospital transfer of neurosurgical patients. Br J Neurosurg 1997;11(1):52-6.
4. U.S. Census Bureau Race. http://www.census.gov/topics/population/race.html. [Accessed 7/1/15].
5. U.S. Census Bureau American Community Survey website. http://factfinder2.census. gov/faces/nav/jsf/pages/searchresults.xhtml?refresh=t. [Accessed December 3, 2012].
6. 2008-2009 U.S. Census Bureau Small Area Health Insurance Estimates. http://www. census.gov/did/www/sahie/data/interactive/. [Accessed December 3, 2012].
7. 2008-2009 U.S. Census Bureau Small Area Income and Poverty Estimates. http:// www.census.gov/did/www/saipe/data/interactive/#. [Accessed December 3, 2012].
8. Rego J, Tan K. Advances in imaging-the changing environment for the imaging specialist. Permaente J 2006;10(1):26-8.
9. Osen H, Bass R, Abdullah F, Chang D. Rapid discharge after transfer: risk factors, incidence, and implications for trauma systems. J Trauma 2010;69(3):602-6.
10. Adhikari S, Blaivas M, Lyon M, Shiver S. Transfer of real-time ultrasound video of FAST examinations from a simulated disaster scene via a mobile phone. Prehosp Disaster Med 2014;29(3):290-3.
11. Culler S, Atherly A, Walczak S, Davis A, Hawley JN, Rask KJ, et al. Urban-rural differences in the availability of hospital information technology applications: a survey of Georgia hospitals. J Rural Health 2006;22(3):242-7.