Disparities in Access to Emergency General Surgery Care in the United States

Abstract

Background

As fewer surgeons take emergency general surgery (EGS) call and hospitals decrease emergency services, a crisis in access looms in the US. We examined national EGS capacity and county-level determinants of access to EGS care with special attention to disparities.

Methods

To identify potential EGS hospitals, we queried the database of the American Hospital Association (AHA) for “acute care general hospital,” with “surgical services,” and “emergency department,” and ≥1 “operating room.” Internet search and direct contact confirmed EGS services that covered the emergency roon 7 days a week, 24 hours a day (24/7). Geographic and population-level EGS access was derived from Geographic Information Systems and US Census.

Results

Of the 6,356 hospitals in the 2013 AHA database, only 2,811 were EGS hospitals. Counties with greater percentages of black, Hispanic, uninsured, and low-education individuals and rural counties disproportionately lacked access to EGS care. For example, counties above the 75^th percentile of African American population (10.2%) had 80% greater odds of not having an EGS hospital compared to counties below the 25^th percentile of African American population (0.6%).

Conclusion

Areas without EGS services exist across the US, disproportionately affecting underserved, rural communities. Policy initiatives need to increase EGS capacity nationwide.

Introduction

In 2006, the Institute of Medicine reported a national crisis in emergency care.¹ Over the preceding decade, emergency rooms (ERs) in the United States (US) experienced a 26% increase in visits and were often operating at full capacity.² Many ERs and hospitals closed their doors resulting in a net loss of 425 ERs (9%) and 198,000 ER beds.^1,2 Given that many ER visits result in need for urgent surgical evaluation and possible emergency surgery, surgeons play a critical role in our nation’s emergency health system. General surgeons care routinely for patients with a range of emergency conditions from common and typically straightforward diseases, such as appendicitis, cholecystitis, and soft tissue abscess, to less common but potentially lifethreatening diseases, such as perforated viscus, ischemic enteritis, and necrotizing soft tissue infection; despite this need, owever, there is a well documented nationwide shortage of surgeons willing to provide emergency general surgery (EGS) coverage that is worsening, placing an estimated 324 million Americans³ at risk of inadequate access to emergency surgical care.^4–6

The evidence of a deficit in general surgeons is substantial.^4–6 Over the years, the number of practicing general surgeons has remained stagnant at approximately 17,000, while at the same time, the US population has increased about 1% per year resulting in a 26% decline in the number of general surgeons per 100,000 US population between 1981 to 2005.^7,8 The inability to regenerate the pool of surgeons for EGS coverage has been attributed to multiple factors.

Surgery trainees are showing decreasing interest in general surgery, often opting for other specialties, largely due to the unpredictable hours, increased liability of providing emergency care, and lesser salary of general surgeons.^7,9,10 While the proposed size of the general surgeon workforce per 100,000 population is 7.5 in the US overall,^12,13 nationally this ratio is only 6.4/100,000, with rural areas suffering an even greater deficit at 4.7/100,000.⁷

While the general surgery workforce crisis has been an area of substantial interest in the surgical literature,^4–11 EGS care as an essential component of our national emergency health system has been poorly described. Little is known about access to EGS care in the US based on geography and sociodemographic characteristics of the population. We sought to analyze disparities in access to EGS care at the hospital level. We hypothesized that vulnerable populations, including rural residents, minorities, and the socioeconomically disadvantaged, would have less access to EGS care.

Methods

Reported EGS Capability

The Annual Survey of Hospitals of the American Hospital Association (AHA), with a historic response rate of greater than 75%, is a comprehensive database of US hospitals considered the gold standard for data on hospital-level characteristics including clinical services provided.¹⁴ We queried the database of the 2013 AHA Annual Survey for the classifications of “acute care hospital,” “surgical services,” “ER,” and at least one “operating room (OR)” to identify acute care general hospitals that in theory provide 24/7 EGS care throughout the entire year. Academic medical centers or other teaching hospitals were included if they were nonfederal, short-term hospitals. Nonfederal, short-term specialty hospitals (e.g., orthopedic, cardiac, strictly pediatric), hospitals not accessible to the general public (e.g., Veteran’s Affairs Hospitals, prison hospitals, college infirmaries), and hospitals located in non-state US territories (e.g., Guam and Puerto Rico) were excluded.

Investigation of True EGS Capacity

To confirm EGS capacity at these potential EGS hospitals, we first undertook a grassroots examination of EGS capacity at all hospitals that appeared to be EGS capable based on the AHA query. We researched hospital websites for evidence of 24/7 EGS services. If the website did not confirm EGS services, Chief Medical Officers and/or ERs were called directly to ask if the hospital provided EGS coverage 24 hours a day. Hospitals that stabilized patients but transferred them elsewhere for all operations, conducted elective operations only, or did not provide 24/7 OR access were eliminated. Second, as part of a larger survey, all hospitals presumed to be offering 24/7 EGS access (through AHA data combined with grassroots investigation) were contacted for further details on their specific EGS practices via the surgeons responsible for EGS services at those hospitals. Hospitals that responded that they did not provide 24/7 EGS care in response to the survey were also eliminated. The remaining hospitals were labeled “EGS hospitals”. With a final survey response of 60%, fewer than 2.0% of responders noted that they did not provide 24/7 EGS care. Because our initial approach resulted in a misclassification rate of <2.0% for responders, we classified all remaining hospitals thought to be EGS capable based on our initial AHA query, internet searches, and direct hospital contact as “EGS hospitals” for the purposes of the analysis presented here (Figure 1).

Figure 1.

Flow chart of the exclusion criteria used to craft final list of hospitals capable of providing 24/7 access to emergency general surgery care

Analysis

In order to measure access to the EGS care based on geography and county sociodemographic makeup, locations of EGS hospitals were overlaid on US maps using Maptitude Geographic Information Systems software (Caliper Corp, Newton, MA, 2016). The 2010 US Census data on race, ethnicity, and socioeconomic variables, including insurance status, poverty level, and education level, were overlaid on maps of hospital distribution to identify population-level disparities in access to EGS care. Poverty was mapped as the percentage of county population living <200% (EDITOR ASK THE AUTHORS IF THEY REALLY MEAN 2005) below poverty level as defined by the US Department of Health and Human Services. Insurance coverage was mapped as percentage of county population without health insurance. Conversely, owing to the relatively low overall rate of Bachelor’s level education in the US, the percentage of the county population greater than age 25 with a Bachelor’s degree was mapped. The percentage categories for black race, Hispanic ethnicity, education-level, insurance status, and poverty were generated using the Fisher-Jenks “Optimal Breaks,” or “Natural Breaks,” algorithm. This method parses data into classes, where each class is a cluster of values that minimizes within-group variance and maximizes any single group’s mean from the mean of other groups.¹⁵

We measured lack of access to EGS hospitals at the county-level based on these same sociodemographic characteristics as well as for county population. First, we calculated the 25th and 75th percentile among all US counties for each variable of interest. We then calculated odds of having no EGS hospital in any given county based on whether or not the county is greater than or less than the 75^th and 25^th percentile, respectively, for any given variable.

Results

Distribution of EGS Capacity in the US

The AHA Annual Survey reported data on 6,356 hospitals. After querying AHA data, 1,546 hospitals were eliminated, because they did not provide acute care (e.g., long-term care hospitals, rehabilitation hospitals, chronic psychiatric facilities). Subsequently, based on a combination of querying AHA data, reviewing hospital websites, calling hospitals, and mailing a survey to presumed EGS-capable hospitals, we eliminated nine hospitals not located in a US territory, seven hospitals that were closed, 1,023 specialty hospitals (e.g., cardiac, orthopedic, pediatric), 79 hospitals with no ER, 412 hospitals with no OR, 226 hospitals with no available general surgeon, and 243 hospitals performing only outpatient or elective surgery (Figure 1). The remaining 2,811 hospitals where an adult with a general surgery emergency might be able to receive care were then mapped for the geographic analyses presented below. Figure 2 shows the distribution of EGS hospitals in the US by county population density.

Figure 2. Distribution of Emergency General Surgery Across the United States.

This map shows the distribution of EGS (2,811) hospitals in the continental US by state and county population density. Each dot represents a hospital reported by the Annual Survey of Hospitals by the American Hospital Association to be an “acute care hospital” providing “surgical services” with both an “emergency room” and at least one “operating room.” Maroon dots represent EGS hospitals (EGS hospital = hospital providing emergency general surgery care 24 hours daily)

In general, EGS hospitals were more likely to be in the more densely populated of the 3,242 counties; 90.7% of the counties without a single EGS hospital (EGS “deserts”) had <50,000 population, and all counties with >500,000 population had at least one EGS hospital. When comparing county-level access to at least one EGS hospital, with a median US county population of 25,939, counties with a population greater than the 75^th percentile of the overall US county population (65,854) had nearly 82% lesser odds of not having an EGS hospital compared to counties with a population below the 25^th percentile of the overall US county population (11,096) (Table 1).

Table 1.

Odds of Lacking Access to an EGS Hospital According to the 3,242 US Counties’ Population Characteristics

	Aggregate County Population Statistics				Odds of >75th Percentile County having No EGS hospital vs <25th Percentile County (95% CI)
	Mean (SD)	Median	25th Percentile	75th Percentile
Overall County Population	98,125 (315,241.4)	25,939	11,096	65,854	0.18 (0.14–0.22)
Percentage Non- Hispanic Black Population	8.9% (14.4)	2.2%	0.6%	10.2%	1.80 (1.48–2.18)
Percentage Hispanic Population	10.8% (19.1)	3.7%	1.8%	9.6%	1.93 (1.58–2.34)
Percentage Population Living in Poverty	14.7% (8.5)	13.2%	9.3%	17.8%	2.31 (1.89–2.83)
Percentage Population Without Health Insurance	14.4% (5.9)	13.9%	10.2%	17.5%	1.85 (1.52–2.25)
Percentage Population With College Degree	20.2% (8.9)	17.95%	13.9%	23.6%	0.15 (0.11, 0.18)

Racial and Ethnic Disparities in Access to EGS Care

Figure 3 shows EGS hospitals by county-level percentage of African American and Hispanic population. Table 1 shows that with a median county-level percentage of African Americans of 2.2%, counties gfeater than the 75^th percentile of the county-level African American population (10.2%) had 80% greater odds of not having an EGS hospital compared to counties less than the 25^th percentile of the county-level African American population (0.6%). Similar disparities are apparent for the distribution of Hispanic populations and EGS hospitals across counties (Table 1).

Figure 3. Access to Hospitals Providing Emergency General Surgery Care Based on Race and Ethnicity.

Figure 3a. EGS Hospitals (2,811) Mapped Over US County Percent African American Population. Figure 3b. EGS Hospitals (2,811) Mapped Over US County Percent Hispanic Population (EGS hospital = hospital providing emergency general surgery care 24 hours daily)

Socio-economic Disparities in Access to EGS Care

Figure 4 also shows EGS hospitals by county-level rates of uninsurance, persons living in poverty, and adults with a college degree. Again, as shown in Table 1, with a median county-level percentage of uninsurance of 13.9%, counties greater than the 75^th percentile of the county-level uninsured population (17.5%) had 85% greater odds of not having an EGS hospital compared to counties less than the 25^th percentile of the county-level uninsured population (10.2%). Similarly, disparities are apparent for poverty level and for college education and EGS hospital distribution across counties.

Figure 4. Access to Hospitals Providing Emergency General Surgery Care Based on Socioeconomic Characteristics.

Figure 4a. EGS Hospitals (2,811) Mapped Over US County Percent Uninsured Population. Figure 4b. EGS Hospitals (2,811) Mapped Over US County Percent Population Living less than 200% below Federal Poverty Level Figure 4c. EGS Hospitals (2,811) Mapped Over US County Percent of Adults >25 Years Old with at Least a Bachelor’s Degree. (EGS hospital = hospital providing emergency general surgery care 24 hours daily)

Discussion

Nearly a decade after the Institute of Medicine declared our nation’s emergency care system at a breaking point, and both surgeons and the lay press brought national attention to a crisis in EGS care,^1,4,8,16 we found a nation unprepared to deliver such care, in particular to the most vulnerable Americans. We found that the 2,811 hospitals that do provide EGS care are distributed unevenly across the US Rural residents, minorities, and the socioeconomically disadvantaged are at high risk of living in an EGS desert with little or no access to EGS care compared to their white, affluent, and urban counterparts. These disparities in access to EGS care substantiate the findings that poor, black, and Hispanic patients are affected disproportionately by ER closures in both urban and rural communities and that rural communities have been particularly hard hit by a declining general surgery workforce.^4–11,17,18 Supply and demand issues in less populated areas will result necessarily in fewer hospitals or perhaps fewer hospitals providing EGS care. Moreover, given that all populations are susceptible to surgical emergencies, our findings highlight the need for a national imperative to achieve EGS equity even if geographic equality is not practical. Thus, while not feasible to have all Americans in close proximity to a hospital well-equipped to provide round-the-clock EGS care (equality), we must take steps to assure high quality outcomes for all EGS patients, regardless of race/ethnicity, socioeconomic status, or place of residence (equity).

While general surgery emergencies can and do impact anyone, minorities, those living in poverty, the un/under-insured, and those with lower educational attainment have been shown to experience worse EGS outcomes, in particular if they reside in rural areas.^19–25 Our findings that these same factors are associated with living in an EGS desert might explain why some patients are particularly vulnerable to lesser than acceptable standards of care (e.g., laparoscopic procedures for appendicitis), greater complication rates, and increased mortality. ^{19–21,26,27} Those living in an EGS desert—whether due to distance, poverty, lack of insurance, education level, or race/ethnicitymight delay care for elective needs only to present emergently in need of a general surgeon who may not be available or may be unable to provide emergency care due to insufficient hospital resources (e.g., anesthesia staff, intensive care unit). In fact, hospitals with greater proportions of socioeconomically disadvantaged patients provide more emergency surgery for conditions typically treated electively.^26,27 Meanwhile, rural residents transferred for emergency operations, including common procedures, such as inguinal hernia repairs and cholecystectomies, traveled an average of 67 miles, with the longest distance traveled greater than 300 miles and were found to require higher level resources compared to those who were not transferred in.²⁸ The time needed to arrange transfer or travel to an EGS-capable hospital may be a contributing factor to these poorer outcomes among socio-economically and geographically vulnerable EGS patients. Ideally, our national public health goal would be to prevent all Americans from reaching a point at which they require EGS care; however, that minorities, the socioeconomically disadvantaged, and rural residents experience worse outcomes suggests a “separate and unequal” emergency care system in our nation today despite efforts—such as the governmental subsidies to critical access (fewer than 25 beds, greater than 35 miles by road from nearest hospital) and safety net hospitals (provide disproportionate care to indigent populations)—to assure access to care for our most vulnerable populations. ^28–34 Yet, in our nation’s Veterans’ Affairs (VA) health system, geographically and sociodemographically diverse veterans with EGS conditions experienced no differences in treatment or outcomes.³⁵ This finding may be attributable to the fact that the VA is a single, integrated health system with uniform oversight despite the variable patient population. The US health care system, conversely, consists of multiple competing interests across vastly different regional economies. Ultimately, the disparate treatment and outcomes of EGS patients from these underserved communities across the US are multifactorial, including unconscious bias among providers,¹⁴baseline socio-economic depravation,³⁶ culturally-rooted, health seeking behaviors,^37–41 access to primary/preventative care facilitated typically via health insurance in our nation,^41–44 and disparities in access to EGS care. Therefore, while it is impractical to develop health care facilities in all of the vulnerable communities we have described, achieving equity in EGS care will require a multipronged approach through system-wide alterations aimed at decreasing the number of EGS deserts across the US through a shared public health approach targeting improved quality, outcomes, and costs for the more than 2.4 million Americans a year who will need care for a general surgery emergency.¹⁴

At the approximately 1,300 critical access hospitals across the US and the other 38% hospitals classified as rural according to the census bureau, general surgeons often function as advanced endoscopists and elective general surgeons while providing emergency urologic, obstetric, gynecologic, facial trauma, hand trauma, and general surgery care.^45–48 Having a general surgeon on staff is estimated to generate approximately $2.7 million in annual revenue and $1.4 million in annual payroll, to and create 26 jobs at rural hospitals.¹⁴ Therefore, without a pool of new graduates to replace retiring older surgeons, many rural hospitals face closure.^7,11Augmenting the pool of general surgeons to provide EGS coverage at rural hospitals is an ongoing challenge that may improve EGS access nationally. Positive experiences in rural based clerkships in medical school and enthusiastic role models have been shown to increase interest in general surgery among medical students.^49–52 Therefore, educators should gear medical school curriculum toward increasing student interest in surgery, while surgeons and residents should understand the profound influence the tone they set has on their students’ career choices. Still, once committed to training in surgery, current trends suggest low likelihood residents choosing broad-based general surgery careers.^{4,9,10,53–55} State and federal health care legislation that includes provisions for financial incentives similar to loan repayment plans for primary care physicians utilized for the National Health Service Corps^56–58 may increase resident interest in non-specialized general surgery. Where current training paradigms are possibly not preparing residents to enter into general surgery practices,^54,55strong mentorship, innovations in surgical training (e.g., apprenticeships in rural surgery), and transition to practice programs may be warranted.^59–61 Finally, given increasingly hospital-based models of employment for surgeons, hospital administrators should create more lifestyle-friendly rural practice environments with ample assistance in the form of mid-level practitioners and surgical assistants as well as cross-coverage opportunities.^9,10 These myriad changes may increase the number of surgeons who choose to make their professional homes at rural hospitals, thereby increasing access to EGS in more remote areas of the US.

Even with the aforementioned recruitment strategies, it is unlikely all rural hospitals, in particular CAHs with fewer than 25 beds, will be fully staffed by general surgeons year round, 24 hours a day.⁶² Still, there appears to be a long held belief among various provider and not just surgeons that even basic EGS diseases cannot be managed without the immediate expertise of a general surgeon. Anecdotally, to this day most general surgeons are required to provide ED coverage for staff privileges at most hospitals across the US, even if the facility cannot provide round the clock OR services or the critical care for often needed for EGS patients. Interestingly,, a study of rural EGS patients who lacked access to surgical evaluation found that nearly one third did not require any intervention at the tertiary care center.²⁸ Therefore, empowering the existing non-surgeon workforce to identify rapidly and appropriately which patients truly need to be treated by a surgeon may alleviate local deficits in general surgery coverage, decrease the travel burden on EGS patients not requiring immediate surgery, and ameliorate overcrowding of tertiary care ERs. This approach has been successful for injured patients through the wide dissemination of Advanced Trauma Life Support (ATLS), a curriculum that trains first line emergency room providers, irrespective of their specialty to identify and manage immediate lifethreats rapidly and appropriately triage injured patients.⁶³ Just as not all injured patients require transfer to a regional trauma center, not all those presenting with common abdominal and skin/soft tissue complaints will need to be transferred to be seen immediately by a general surgeon when one is not available locally. Importantly, implementing a clear, evidence-based, algorithm-guided training paradigm for the ‘optimal care of the EGS patient’ will only be possible with: acceptance by surgeons, primary care providers, emergency room physicians, and hospital administrators alike; the involvement of surgeons nationally in generating and disseminating a robust curriculum, availability of telemedicine consult services as have been utilized for stroke care,^64,65 and the enthusiastic participation of internists/hospitalists in managing non-operative surgical disease when transfer is not warranted. This approach would represent a transformative change in initial EGS management and triage and would greatly enhance EGS equity while containing costs.

Addressing geographic disparities alone, however, will do little to address the poor, uninsured, and minorities who reside in our nation’s urban/suburban areas. Receiving care at an approved stroke center, typically in metropolitan and non-rural areas, has improved both processes and outcomes for some minority patients experiencing a similarly time-sensitive health condition.^66–68 Thus, adopting a similar ‘centers of excellence model’ for EGS care may both narrow the urban-rural divide in EGS care and augment EGS capacity across vulnerable populations irrespective of their county socio-demographic profiles. To do so, how ever, would require re-envisioning the Acute Care Surgery paradigm with a new scope of equity across the arena of EGS care . The paradigm of acute care surgery was proposed by leading trauma surgeons as a care delivery model incorporating trauma surgery, general surgery, and surgical critical care just over a decade ago and has been shown to improve outcomes.^69,70 This model,however has been viewed largely from the perspective of the proponets of the acute care paradigm as an isolated, hospital-based service line; but, our prior research suggests that the resource intensiveness of a round-the-clock acute care surgery service may not be practical at all hospitals, in particular, the smaller hospitals lacking a trauma center caable of providing 24/7 coverge.^71,72 Therefore, this paradigm of acute care surgery should be understood, by surgeon leaders, hospital administrators, and regional public health officials alike, as a network for patients presenting for EGS care centered around an EGS Center of Excellence (i.e., the hospitals with an acute care surgery service).

These various stakeholders should come together with the goal of providing monetary support, infrastructure, and human capital to develop evidence-based triage criteria (to supplement education and training cited above), strong referral networks (with telemedicine components) for immediate transfer/acceptance of patients in need of EGS care from a critical care-certified surgeon, compensation schemes for the care of uninsured patients similar to trauma activation payments, and surgical home models (with e-medicine components) for socio-economically vulnerable and geographically distant patients to ensure follow-up and decreasee complications. The overall costs to the health care system along with human suffering due to delay in access to care and substandard care would be decreased if the paradigm of acute care surgery were to re-envisioned and implemented as a tiered level of capability (e.g. Level 1, 2, 3), and regionalized with a hub and spoke Center of Excellence model with verification criteria as are in place currently for trauma care, stroke care, and multiple other high risk diseases.

The existence of unconscious bias has been documented among a number of health care providers across specialties, including general and acute care surgeons.^73,74 Unconscious bias has been implicated in disparate clinical decision-making and poor health outcomes in a number of surgical and unexpected conditions,^75–78 but this finding has been inconsistent;^79,80 however, in light of the disparities in access to EGS care we describe, we believe that the goal of equiity in EGS care cannot be achieved without combining broad measures for increasing the general surgery workforce, training non-surgeons in the optimal evaluation of the EGS patient, and regionalization of EGS care without unconscious bias training for all providers who participate in the care of patients who present with abdominal and skin/soft tissue complaints that may require emergency surgery. Self-awareness training, education in ‘blind spot’ recognition, and implementation of ‘physician interaction’ checklists are all approaches that have been explored in health care settings to attempts to decreae the negative effects of unconscious bias; these initiatives may play an important role in improving EGS equity across all domains of emergency health care delivery where a person with a potential general surgery may present.^81–83

There are some important limitations to our work. The US Census data were used to characterize EGS deserts and the most recent data was five years older than our study period. In particular, insurance data were prior to implementation of the Affordable Care Act. Our own findings revealed weaknesses of the AHA Annual Survey, and it is possible that we misclassified the characteristics pf the EGS hospitals due to other discrepancies in data collection, in particular those 1,154 hospitals that did not reply to our survey. Additionally, we used having at least one OR as our search criteria; while this may not represent EGS capacity accurately, it is a starting point and the minimum necessary. Therefore, EGS capacity is likely even less than we documented. Large numbers of missing data on trauma center verification limited our ability to correlate EGS capacity with trauma capacity. Closures of health systems are dynamic; in the time it took us to analyze and map our data, the EGS capacity may have changed. Importantly, while EGS is generally accepted to include a range of diseases from appendicitis to necrotizing soft tissue infection, we did not have data on hospital case mix to verify or establish the actual EGS care rendered in our sample. To the best of our our knowledge, however, our study is the first analysis at the national level of access to EGS care at the hospital level with a focus on socio-demographic disparities at a time when health care access for our nation’s most vulnerable populations may well worsen. We maintain that our detailed, sequential efforts to confirm 24/7 EGS services are an acceptable proxy to begin examining such a vexing public health concern that is expected to worsen as our population ages, and more patients require EGS care.

Conclusion

Our analysis of access to EGS care is timely and brings attention to one important but neglected facet of our emergency health system in the US. More than 2 million Americans year will need EGS care annually; in comparison, approximately 1.4 million will require treatment for myocardial infarction, 1.2 million will need management for serious injuries, and 900,000 will need care for strokes.^84–87 Still, achieving timely access to high quality and cost-effective EGS care has been lacking in the national policy and public health focus of these other conditions. General surgery emergencies are equal opportunity diseases that require urgent surgical evaluation and potential operative intervention, butrural, poor, uneducated, and minority patients disproportionately lack access to EGS care. That our current system of emergency care, already at its breaking point at baseline, is more ill-equipped to meet the basic EGS needs of some Americans versus others should be alarming to surgeons, public health advocates, and policy makers. Efforts should be directed toward strengthening EGS capacity in the US and revitalizing the general surgery workforce to aligning hospital priorities with population level health needs that are vital to ensure EGS equity that is timely, high quality EGS care for all patients irrespective of place of residence, socioeconomic status, or race/ethnicity.