EGS regionalization: pro (HS)

Despite the universal susceptibility to EGS disease, access to EGS care is not uniformly available across the USA. With a median US county population of ~26 000, 91% of US counties with <50 000 population lack access to a hospital within the county where EGS care could be provided.21 US counties with <25th percentile population had 82% higher odds of lacking access to EGS care compared with those with >75th percentile population. This likely explains why rural residents transferred for EGS, including common procedures such as incarcerated inguinal hernia repairs and cholecystectomies for acute cholecystitis, travel an average of 108km, with the longest distance being > 482km.22 In addition to rural areas, socioeconomically vulnerable areas, including those with higher concentration of minorities, uninsured residents, and residents lacking college degrees, are most likely to lack access to EGS care nearby.21 Therefore, regionalization is an approach to provide EGS care to these millions of Americans who would have otherwise been treated in their own communities by disappearing general surgeons.

Even if a smaller or more rural hospital has a general surgeon who takes ED call and operating rooms, the inherent surgical and perioperative needs of critically ill or complex EGS patients may outstrip locally available resources. EGS patients are clinically complex compared with elective counterparts. They have more comorbidities (eg, diabetes, hypertension, chronic obstructive pulmonary disease, immunosuppression), more acute physiologic derangement (eg, respiratory failure, ascites, sepsis), and worse baseline function and cognition.1 23–25 Of all EGS patients, 13% to 50% require intensive care, and 12% to 14% have septic shock.23 26–31 Among all patients in the USA requiring intensive care for sepsis, approximately 16% are septic due to EGS disease.32 Across all EGS diagnoses, the median length of stay (LOS) is 3 days (IQR 2–6), 29% to 34% of patients require operation, 10% experience complication(s), and 1.6% to 3% die.1 5 33 As with other complex surgical diseases, EGS patients should benefit from a regionalized system that enables access to the highest quality care.34–37

In fact, as many as 20% of EGS patients presenting to EDs are already transferred from elsewhere for definitive care.38 The need for specialized critical care resources and added risk of baseline comorbidities, along with operative availability, is theorized as the leading reason for transfers of care for EGS patients.17 19 20 39–43 However, these patients lack the benefit of evidence-based triage criteria and verified centers of excellence that benefit patients with trauma, acute myocardial infarction (AMI), and ischemic stroke. That Americans experiencing acute chest pain, neurologic deficits, and injury benefit from robust systems of care targeting rapid evaluation and prompt triage for prompt intervention when needed has been shown to repeatedly reduce socioeconomic and demographic disparities in access to care and outcomes (HP Santry, unpublished data).44–54 Investing resources to provide this kind of life-saving infrastructure and human capital at all hospitals is not feasible in the current landscape of general surgery.

There is no rational argument to treat EGS patients who need prompt diagnosis of a life threat and referral for emergency operation differently from patients with AMI or ischemic stroke who need prompt diagnosis and referral for emergent interventions. Regionalization for EGS patients to tertiary care centers that are already investing in acute care surgery (ACS) teams to provide robust 24/7 EGS coverage is an ideal pathway to ensure timely and high-quality access to EGS care.

EGS regionalization: con (LSK)

Regionalization of EGS has been proposed to be concentrated at (1) high-volume centers, (2) trauma centers, or (3) centers that have adopted the ACS model. There are arguments both for and against regionalization at each of these types of centers.

Proponents of regionalization argue that high-volume centers perform better. The relationship between operative volumes and outcomes, primarily mortality, has been described for multiple procedures and replicated by numerous investigators.55–57 Such a volume–outcome relationship has also been demonstrated for EGS procedures. Ogola et al58 identified a significantly lower mortality of EGS patients with a predicted risk of death greater than 4% who were treated at high-volume versus low-volume EGS hospitals (7.7% vs. 10.2%, p<0.001). However, there are additional issues that should be considered beyond the volume–mortality association.

First, mortality should not be considered the sole measure of surgical quality. Metrics such as minor and major complications, functional outcomes, quality of life, and other patient-reported outcomes should also be considered. Second, EGS differs from other surgical subspecialties due to the high proportion of non-operative patients, who were not accounted for in the Ogola et al58 analysis. When non-operative EGS patients are included in risk-adjusted analyses, hospital rankings based on performance are significantly altered.59 Third, the relationship between volume and mortality varies based on risk, suggesting a greater benefit for treatment of high-risk patients in a high-volume center. However, current strategies for identifying high-risk patients (ie, such as using risk calculators) may be insufficient to identify those patients who would most benefit from transfer.60

Another argument is that standards of practice at level 1 trauma centers enhance care of EGS patients. On the one hand, level 1 trauma centers have multiple requirements which would complement EGS care, such as 24-hour, inhouse surgical coverage; existence of a robust quality assurance program; collaborative clinical services such as anesthesiology and radiology; and availability of subspecialists. On the other hand, EGS outcomes have not been shown to be better at trauma centers.61–63 In fact, studies have suggested increased morbidity,61 mortality,63 LOS,63 and mean charges at trauma centers.63

As with the volume–outcome relationship, comparisons of outcomes at trauma versus non-trauma centers are more complex than they may at first appear. To start, not all surgeons have equivalent skills and outcomes. Surgeon-driven variations in care and outcomes may have a greater impact on outcome than center-driven variations.64 Second, not all trauma centers are created equal. When stratified based on mortality among trauma patients, high-quality trauma care hospitals had lower EGS mortality rates.65 Similarly, trauma centers with a certified ACS program have improved EGS outcomes.66

Finally, there is the argument that EGS care should be provided at centers with an ACS model. In general, the ACS model, first proposed in 2005, combines trauma, EGS, and critical care in a centralized unit.67 Since then, multiple studies have demonstrated that the ACS model is associated with reduced morbidity and mortality.68 However, the devil is in the details. There is no consensus regarding the essential components of the ACS model, and there are wide variations in the processes of care for EGS patients among hospitals with and without an ACS model.68 69

Despite the proposed benefits, adoption of the ACS model across the USA has been slow. In a national survey, only 272 hospitals out of 1690 (16%) had adopted an ACS model as of 2015.70 Qualitative studies reveal multiple surgeon concerns.71 One such concern is that the ACS model may be surgeon-centered rather than patient-centered. The ACS model is intended to benefit surgeons in revitalizing trauma surgery, which is becoming increasingly non-operative; this may punish patients by worsening continuity of care. A second concern was voiced that the ACS model may not be compatible with institutional resources, such as for a dedicated operating room for emergency cases. Lastly, some surgeons thought that the ACS model may not be financially viable.

The unintended consequences of regionalization must also be scrutinized. Proponents of EGS regionalization suggest that it would improve safety, decrease costs, and reduce resource utilization. Regionalization has been investigated for improving outcomes in other surgical subspecialties such as surgical oncology and bariatric surgery. However, opponents of regionalization have cited negative consequences such as decreased access to care for vulnerable patient populations, increased disparities in quality of care between high-volume and low-volume hospitals, increased financial and travel burden for patients, and decreased coordination of care due to multidisciplinary team members being located at multiple hospitals.72 Furthermore, data on whether outcomes are truly improved after regionalization, such as after establishment of centers of excellence for bariatric surgery, are conflicting.73–77 The concerns that regionalization might unfairly restrict access to care without improving outcomes or reducing costs hold true for EGS surgery as well.

Data regarding the potential effects of regionalization for EGS patients are limited. Studies comparing EGS patients who were transferred versus those who were directly admitted have demonstrated that transfer is an independent predictor of increased mortality.78 79 Additionally, there are reports of small groups of hospitals in other countries such as Canada and Ireland where EGS care was regionalized. In Canada, regionalization of ACS to three out of six hospitals did not improve morbidity, mortality, LOS, or readmissions.80 Furthermore, it resulted in an increase in wait times from referral hospitals and from the ED to inpatient admission. In Ireland, regionalization of complex and emergency cases to a hospital with a dedicated operating room around the clock resulted in an increase in night-time cases and an increased delay in the time to the operating room and LOS for weekday EGS cases.81

There are several alternatives to regionalization to improve EGS care. First, participation in regional quality collaboratives that use a common data registry such as the Michigan Surgical Quality Collaborative can move the needle in improving care.82 Such collaboratives must collect data on both operative and non-operative EGS patients.59 Second, standardization of care, based on high-quality evidence when available, can also improve outcomes. Huddart et al83 found that a five-component bundle reduced mortality after emergency laparotomy. Last but not least, telementoring may provide a method for more experienced surgeons to coach those who are less experienced or have less comfort with EGS care.84

In conclusion, there is no clear evidence that regionalization of care would improve safety, decrease costs, or reduce resource utilization. Furthermore, regionalization may have unintended consequences such as reducing the timeliness, efficiency, and patient-centeredness of EGS care or increasing disparities in access to EGS care. To determine the best strategy for improving EGS care across the USA, acute care surgeons need to participate in robust data collection in a common data registry, to develop evidence-based pathways and bundles to improve outcomes, and to develop regional or virtual quality collaboratives to share best practices. Should acute care surgeons continue to strive for regionalization, more work needs to be done in terms of understanding what constitute the core components of the ACS model, why implementation of the ACS model has not been more widespread, and how to best stratify risk (for both operative and non-operative management) of EGS patients so as to identify those most likely to benefit from transfer.

In favor of regionalization: a rebuttal (SS)

Variations in risk-adjusted patient outcomes are well known. Two decades ago, data from the Veteran Affairs system demonstrated significant variations in outcomes of surgical patients, with certain hospitals achieving significantly lower mortality than others.85 These findings formulated the basis for the National Surgical Quality Improvement Project. About a decade later, similar findings were demonstrated for trauma centers, which formulated the basis for Trauma Quality Improvement Project.86 Similar variations in outcomes of EGS patients across various hospitals have also been shown to occur, strengthening the case to establish the EGS Quality Improvement Program.87 Mortality ratio at high mortality hospitals was over three times higher than low mortality hospitals, indicating a significant gap in quality. It is estimated that if all patients treated at average and high mortality hospitals experienced the same low risk of death as the low mortality hospitals, then there could be 16 812 fewer deaths per years.

The relationship between volume of surgical patients and their outcomes is well known across other diseases. Similarly, a sharp drop in mortality for EGS patients has been found as the hospital volume of EGS patients increased over 500, and became stable at a volume of about 660 EGS patients a year (figure 2 in Ogola et al).58 The mortality rate in low-volume hospitals that treated less than 660 patients a year was found to be 3.7%, whereas the mortality rate in high-volume hospitals that treated 660 or more patients was 1.6%.

The key is determining which EGS patients would benefit the most from transfer to a high-volume center. Figure 3 in Ogola et al shows observed versus expected mortality rates at high-volume versus low-volume hospitals.87 Mortality rates are similar at high-volume and low-volume hospitals when the risk of death is low. However, when the risk of death approached 4% and higher, the observed mortality rates start diverging, with lower mortality rates at high-volume hospitals, suggesting that high-volume hospitals achieved better outcomes for high-risk EGS patients. As expected, high-risk patients include those presenting with NSTIs, bowel perforation, bowel ischemia, vascular emergencies, gastrointestinal hemorrhage, and any emergent conditions associated with shock. These high-risk patients constituted less than 10% of all EGS patients but are most likely to benefit from transfer to high-volume centers. The remaining less severely ill EGS patients at low risk of death can be, and should be, treated locally. However, without a plan for regionalization in place, similar to regionalization of care for trauma patients, depending on severity of injury and institutional resources to provide adequate care, high-risk EGS patients will continue to suffer preventable morbidity and mortality. High-risk EGS patients should be treated at hospitals that care for a high volume of EGS patients, and this can only be done with a regionalized system of care, with transfer agreements and evidence-based triage guidelines.

Against regionalization: a rebuttal (LL)

ACS has been proposed as the model of care to serve as the hub in a hub-and-spoke regionalized model of EGS care. There are numerous aspects of an ACS service at tertiary centers that have established trauma programs, with robust trauma and acute care surgeons directing all aspects of injured and critically ill patients that might benefit EGS patients. Many of these centers serve as regional referral centers for hospitals that do not have the ability to perform complex surgical procedures or advanced surgical critical care. Data are abundant that such ACS services improve time to treatment, reduce LOS, and improve outcomes for EGS patients as previously noted. However, this model may not have the promise that has been lauded by the highly resourced single centers that also publish their successful outcomes.

In 2016 the American College of Surgeons Governors conducted a survey examining how the ACS paradigm might change the surgical landscape.88 Most survey respondents (69%) worked in a facility with an ACS service, and about half were in academic institutions. Of the survey respondents 26% reported a decrease in the number of surgical cases done by general surgeons who were not involved in the ACS program. Although overall, 70% of survey respondents could foresee no significant change in their income if their hospital developed an ACS program, 65% of the responders preferred not to develop such a program. When analyzed by type of surgeon, a substantial majority—73% of general surgeons and 54% of other surgeons—are opposed to the development of an ACS program. If the main operators within a system of care are opposed to adopting it, how then will patient outcomes be improved by regionalizing care with such a system, namely a hospital with an ACS service, at the core? Among the deterrents to high-quality care cited by respondents were fragmented care/poor continuity of care, operating room access barriers, and inexperience among those working as acute care surgeons.88

Overly prescriptive regionalization plans may also have the unintended consequence of reducing access to EGS care. Kreindler and colleagues,80 who examined the effect on EGS patients overall in a regionalized system within Canada of a hospital with ACS and six surrounding hospitals, found that wait times increased primarily for patients presenting at non-referral hospitals who were likely to require transfer to a referral hospital. They surmised from their data that those who presented to a non-system hospital had increased wait times due to those from the insystem hospitals being given priority when previously they may have been able to get adequate care at the originating hospital.

Finally, as with any practice model, the consequences for the practitioners historically providing such care must be considered. Unfortunately, the emergence of ACS has been viewed by some, including the Governors of the American College of Surgeons, as negatively impacting the profession of general surgery with overspecialized care.88 All surgeons trained in general surgery are capable of providing EGS care, and with ACS and regionalization there may be a number of deleterious effects, including an untenable call schedule; conflicts/controversy between acute care surgeons and community surgeons; and devaluation of the general surgeon. Financially, general surgeons are also possibly hurt by ACS models of care. Miller and colleagues89 found that despite elective case load increase by 23% after implementation of an ACS model of care, the work relative value units (wRVUs) decreased by 8% among those now in a primarily elective practice. Meanwhile, wRVUs increased by 140% among acute care surgeons.

For these reasons, ranging from fragmented care to the loss of general surgery as a profession, regionalization should be approached cautiously, if at all.