Interhospital transfer (IHT) in emergency general surgery patients (EGS): A scoping review.

Background/Aims of study: Interhospital transfer of emergency general surgery patients continues to rise, and no system for transfer of emergency general surgery patients exists. This has major implications for cost of care and patient experience. We performed a scoping review to understand outcomes related to transfer and the associated factors and to identify any opportunities for improvement.
Methods: Studies involving emergency general surgery patients with interhospital transfer were identified by searching OVID MEDLINE, EMBASE, Cochrane Library, and Scopus. There were 1,785 records identified. After duplicates were removed, there were 1,303 articles screened in the initial phase. Fifty-eight articles were included in the second phase. Eventually, 21 articles were included in the review. Thirty-seven articles were removed during the full-text screening phase due to the following: wrong publication type (2), wrong population (8), abstract (11), outside the United States (3), and wrong study design (6).
Results: Transferred patients had a higher mortality rate, were older, were more likely to be male and to undergo reoperation, and had higher resource utilization compared to patients who were not transferred. All emergency general surgery patients had a high burden of chronic disease. Unnecessary transfer, typically defined by lack of intervention and discharge within 72 hours, was reported to be 8.8% to 19%.
Conclusion: Emergency general surgery patients have a high rate of comorbidities. Limited physiologic status information prior to patient transfer limits understanding of the necessity for transfer. Areas for improvement include assigning a physiologic status for all patients and utilizing telehealth. More detailed information needs to be captured to determine the appropriateness of transfer.

INTRODUCTION

Emergency general surgery (EGS) is a common health occurrence that impacts an estimated 2 million patients or about 7% of hospital admissions in the United States annually [[1], [2], [3], [4]]. The most common EGS procedures include partial colectomy, small bowel resection, cholecystectomy, operative management of peptic ulcer disease, lysis of peritoneal adhesions, appendectomy, and exploratory laparoscopy/laparotomy [5].

Despite the increased cost of care and complication in EGS patients, the rate of interhospital transfers has increased in recent years [2,6,7]. The increase in transfers has occurred despite previous research showing similar outcomes in procedures that are typically within the skill set of most general surgeons, especially when comparing rural versus urban settings [8,9]. The increase in EGS patient transfers is likely due to a multitude of factors, one of which may include lack of access to a general surgeon. The number of surgeons willing to cover emergency surgery has decreased, whereas the number of surgeons who have specialized has increased [10,11]. This problem is exacerbated by the unequal distribution of surgeons. Roughly 30% of counties in the United States with a total population of 9.8 million people do not have a surgeon, whereas the national average is about 45 surgeons per 100,000 people [[12], [13], [14]].

The American College of Surgeons (ACS) has a well-developed pathway for transfer when it comes to trauma patients, but it does not have any specific guidelines in place for EGS [6,8,15] in part because of the increased complexity associated with EGS. Patients who present with acute surgical conditions can be urgent or emergent, both of which require significant resource allocation. The decision to transfer is difficult and can include many factors including local surgeon skill set and the ability of the hospital to manage acutely ill patients. Other factors including comorbid conditions may increase the need for higher level of care, which again may or may not be available at a local hospital. There are significant risks related to transfer including delay of care, risk of transportation, potential for clinical deterioration, potential for poor handoff communication, and neglected patient preferences [2,[16], [17], [18], [19],20].

The 7 most common EGS procedures, as reported above, do not have a consistent or cohesive transfer protocol for patients who require these procedures or who have the diagnoses that lead to these operations. Given the increasing number of patients who require emergency surgery that could be done by general surgeons, without clear protocols in place, we sought to understand the key characteristics of these patients and their outcomes depending on transfer status available in the current literature. In addition, we sought to identify factors that may have driven the decision for transfer.

METHODS

Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews (PRISMA-ScR) (Fig 1) guidelines were followed to conduct the review and analysis [21]. A protocol was developed and registered with the Open Science Framework at https://osf.io/7g4uf/.

Fig 1

PRISMA 2009 flow diagram.

Search Strategy

A medical librarian (SJB) created the literature search strategy after meeting with the research team to clarify goals and further define selection criteria. The search strategy was built and tested for sensitivity in Ovid MEDLINE using Medical Subject Headings (MeSH) and keywords (see Appendix 3), and the search strategy was translated to 3 other databases: Cochrane, EMBASE, and Scopus. Reference lists of included studies were also reviewed for relevant articles. Searches were run without limits from database inception through June 2, 2021.

Study Selection

Studies were included if they met the following criteria: (1) patients transferred with emergency general surgery; (2) articles with an observational study design; (3) included information about intervention or lack of intervention following transfer; and (4) studies in the United States.

Studies were excluded for the following reasons: (1) patients transferred with traumatic surgical emergencies; 2) patients transferred for surgical procedures that were elective;( 3) studies outside the United States; (4) publications that were systematic reviews or other type of review; (5) studies with EGS patients who were not transferred; (6) randomized control study design; (7) vascular surgery and GI bleed; and (8) studies with only pediatric patients.

Duplicate references were removed, and items were uploaded to Rayyan [22] for blinded and independent assessment of eligibility by 2 researchers (RE and PMT). After the initial title and abstract screening, conflicts were resolved by consensus. When researchers could not reach consensus, titles were included in the full-text screening. PDFs of 49 selected studies were uploaded for full-text screening and independently reviewed by RE and PMT. The consensus model was again used to resolve conflicts in the full-text screening phase. Near the end of the project, research reviewed included references to check for any additional articles that met inclusion criteria. Again, both researchers came to a consensus when determining which articles should be included. This process added another 9 articles (Fig. 1).

Data Extraction and Analysis

Twenty-eight articles were included in the final analysis (Table 1). A data extraction form was created and tested by 2 researchers (RE and PMT). Prior to data extraction, pilot testing of the form with 2 articles was conducted to highlight inconsistencies, and modifications to the form were made by consensus. The data extraction form was built using Qualtrics software version June 2021 (Qualtrics, Provo, UT).

Table 1

Articles included in the final analysis

	Name	Author, year	Journal
1	Unnecessary transfers for acute surgical care: who and why?	Kummerow Broman, K; 2016 [29]	The American Surgeon
2	Factors associated with inter-hospital transfer of emergency general surgery patients	Ingraham, A; 2019 [44]	The Journal of Surgical Research
3	Cost and burden and mortality of rural emergency general surgery transfer patients	Keeven, D; 2019 [8]	The Journal of Surgical Research
4	Transfer of acute care surgery patients in a rural state: a concerning trend	Misercola, B; 2016 [17]	The Journal of Surgical Research
5	Emergency general surgery transfers in the United States: a 10-year analysis	Reinke, C; 2017 [2]	The Journal of Surgical Research
6	Interhospital transfers of acute care surgery patients: should care for nontraumatic surgical emergencies be regionalized?	Santry, H; 2011 [45]	World Journal of Surgery
7	Acuity, outcomes, and trends in the transfer of surgical patients: a national study	Huntington, C; 2015 [7]	Surgical Endoscopy
8	Transferred emergency general surgery patients are at increased risk of death: a NSQIP propensity score matched analysis	Castillo-Angeles, M; 2019 [46]	Journal of the American College of Surgeons
9	High-volume hospitals are associated with lower mortality among high-risk emergency general surgery patients	Ogola, G; 2018 [30]	The Journal of Trauma and Acute Care Surgery
10	Interhospital transfer for acute surgical care: does delay matter?	Kummerow Broman, K; 2016 [28]	American Journal of Surgery
11	Interhospital transfer for emergency general: an independent predictor of mortality	Yelverton, S; 2018 [47]	American Journal of Surgery
12	Triaging to a regional acute care surgery center: distance is critical	Diaz, Jose J; 2011 [48]	The Journal of Trauma
13	Factors associated with interhospital transfers of emergency general surgery patients from emergency departments	Fernandes-Taylor, 2021 [49]	American Journal of Emergency Medicine
14	An evaluation of emergency general surgery transfers and a call for standardization of practices	Bruenderman, 2021 [43]	Surgery
15	Escalation of mortality and resource utilization in emergency general surgery transfer patients	Keeven, D; 2019 [50]	Journal of Trauma Acute Care Surgery
16	Characteristics and timing of interhospital transfers of emergency general surgery patients	Philip, J; 2018 [51]	Journal of Surgical Research
17	Interhospital transfer and adverse outcomes after general surgery: implications for pay for performance	Lucas, D; 2014 [6]	Journal of the American College of Surgeons
18	Transfer status: a significant risk factor for mortality in emergency general surgery patients requiring colon resection	DeWane, M; 2018 [52]	Trauma and Acute Care Surgery
19	Transfer status: a risk factor for mortality in patients with necrotizing fasciitis	Holena, D; 2011 [53]	Surgery
20	Effect of transfer status on outcomes for necrotizing soft tissue infections	Ingraham, A; 2017 [54]	Journal of Surgical Research
21	Effect of transfer status on outcomes of emergency general surgery patients	Philip, J; 2020 [39]	Surgery

Method for Data Analysis

Data were manually extracted and uploaded into an Excel spreadsheet by 2 researchers (RE and PMT), allowing for us to carry out the extraction phase. The same 2 researchers (RE and PMT) individually reviewed data and highlighted common themes in the extracted data. Specifically, we looked at important data markers like age mortality, comorbidities, and insurance status. The data were gathered in an Excel spreadsheet, which allowed the authors to assess the information to perform a narrative synthesis. Both researchers then came to a consensus on the most common and important findings in the data.

RESULTS

All publications reviewed were cohort studies using either a locally created health care system database (n = 8) or one of the publicly available databases (National Inpatient Sample [NIS] = 6; ACS National Surgical Quality Improvement Program [NSQIP] = 6; and National Emergency Department Sample [NEDS] = 1) that have been created to examine patient quality and characteristics at a national level (Appendix 2). The most common diagnoses of patients included the following: appendicitis, gallbladder pathology, small bowel obstruction, diverticular disease, ischemic colitis, and general abdominal conditions. In reviewing the final 21 publications related to emergency general surgery patients, we were able to assess the following elements of the patient cohorts:

30-Day Mortality

Fourteen articles studied mortality. The results showed varying mortality rates as different conditions have higher rates of mortality regardless of transfer such as necrotizing fasciitis. Studies that reported broad EGS conditions had mortality rates that ranged from 2.3% to 12.6% in patients who were transferred from emergency department (ED) or outside hospitals (nursing home transfers were even higher, up to 24.7%). In the same studies, the mortality rate in local admits/patients who were not transferred ranged from 0.4% to 3.1%. Three studies investigating just 1 condition (2 studies on necrotizing fasciitis and 1 study on emergent colorectal surgery) had higher mortality rates ranging from 19.4% (ED) to 25.7% (inpatient transfer [IPT]) to 34.3% nonhospital transfer (NHT) in colorectal surgery group and 8.9% to 15.5% in necrotizing fasciitis patients. In the same studies, the mortality rate was 12.9% in nontransfer of the colorectal study and 8.7%–13.7% in patients with necrotizing fasciitis. Overall, the rate of patient transfer increased from 1.2% to 3% over a 10-year period, whereas mortality decreased by 40%2 (Table 2).

Table 2

The mortality that was reported in patients who were transferred compared to those who were not transferred.

	Mortality
	Transferred (%)	Not transferred (%)
3	MortalityED to ED: 2.3IPT: 7.5	Mortality:LA: 0.4
5	Mortality:Intervention: 4.4No intervention: 4
6	Mortality: 4.9	Mortality: 0.9
7	Mortality: 8.9	Mortality: 1.7
8	Mortality: 10.8	Mortality: 3.1
9	Mortality:High risk: 8.1Low risk: 0.8
11	Mortality: 4.2	Mortality: 1.5
12	Mortality: 2.7
15	Mortality: EDT (7.4), IPT (12.6), NHT (24.7)	Mortality: 3.3
17	Mortality: 10	Mortality: 4
18	Mortality: EDT (19.4), IPT (25.7), NHT (34.2)	Mortality: 12.8
19	Mortality: 15.5	Mortality: 8.7
20	Mortality: 8.9	Mortality: 13.7
21	Mortality: 4.4	Mortality: 1.6

Age and Sex

Age: Nineteen articles reported median age in EGS patients. The median age of EGS patients who were transferred ranged from 47 to 78.1 years. The median age of patients in the local admit group ranged from 44 to 59 years (Appendix 2).

Sex: Seventeen articles included sex in the EGS cohort. The percent of patients who were male in the transferred group ranged from 48% to 57%. The percent of local admissions that were male ranged from 42.3% to 50.9% (Appendix 2).

Socioeconomic Status and Insurance Status

Socioeconomic status: Five articles exploring EGS patients reported socioeconomic status. The makeup of the transfer population included 28.8%–34.6% in the lowest quartile, 25.6%–31.5% in the second quartile, 20.2%–22.5% in the third quartile, and 13.6%–21.2% in the highest quartile. The makeup of the local admission population included 27.4%–28.1% in the lowest quartile, 25.6%–25.8% in the second quartile, 23.9%–24.6% in the third quartile, and 19.9%–22.2% in the highest quartile.

Insurance status: Fourteen articles reported insurance status. The results varied by study design and population. The percent of patients who were transferred and had private insurance was 5%–47.5%; Medicare, 38.8%–82.3%; Medicaid, 5.3%–29; uninsured/self-pay ,1.8%–9%; and other, 2%–10.6%. The percent of patients who were not transferred and had private insurance was 29.3%–45%; Medicare, 34%–47.5%; Medicaid, 8%–28%; uninsured/self-pay, 6%–9.6%; and other, 3.2%–13.3% (Table 3).

Table 3

The insurance status of patients who were transferred and those who were not transferred

	Insurance status
	Transferred (%)	Not transferred (%)
1	Private: 35Medicare: 46Medicaid: 10Uninsured: 8
2	Private: 26.7Medicare: 52.9Medicaid: 10.8Self-pay: 5.4Other: 3.8No charge: 0.4	Private: 30.1Medicare: 45.7Medicaid: 11.4Self-pay: 8.5Other: 3.3No charge: 1.0
3	Private: (ED-ED): 15; (IPT): 5Medicare: (ED-ED): 55; (IPT): 60Medicaid: (ED-ED): 29; (IPT): 27Other/self-pay: (ED-ED): 2; (IPT): 3	Private: 32Medicare: 34Medicaid: 28Other/self-pay: 6
4	Private: 26Medicare: 55Medicaid: 9Uninsured: 6Military/government: 4	Private: 45Medicare: 34Medicaid: 8Uninsured: 9Military/government: 4
5	Private: 29Medicare: 51Medicaid: 10Self-pay: 6Other/no charge: 4
6	Private: 42.7Medicare: 38.8Medicaid: 12.6Uninsured: 3.9Other: 1.9	Private: 39.3Medicare: 37Medicaid: 13.4Uninsured: 7Other: 3.2
9	Private: 8.1Medicare: 82.3Medicaid: 5.3Uninsured: 1.8Other: 2.3	Private: 29.3Medicare: 43.5Medicaid: 12.9Uninsured: 9.6
10	Private: 33Medicare: 47Medicaid/uninsured: 20
11	Private: 29Medicare: 51Medicaid: 10Uninsured: 6Other: 4	Private: 32Medicare: 45Medicaid: 10Uninsured: 8
13	Private: 26.5Medicare: 45.2Medicaid: 14.2Self-pay: 9.0Other: 4.3No charge: 1.0	Private: 30.3Medicare: 18.7Medicaid: 23.5Self-pay: 22.3Other: 4.1No charge: 1.0
14	Private: 25Medicare: 45Medicaid: 21Self-pay: 9
16	Private: 47.6Medicare: 43.7Medicaid: 9Uninsured: 3
19	Private: 40.7Medicare: 31.3Medicaid: 18Uninsured: 5.9Other: 4.1	Private: 31.5Medicare: 29.9Medicaid: 15.9Uninsured: 15.2Other: 7.5
21	Private: 28.9Medicare: 49.8Medicaid: 10.7Other: 10.6	Private: 30.8Medicare: 44.9Medicaid: 11.1Other: 13.3

Unnecessary Transfer

Three studies on EGS patients reported unnecessary transfer, which is typically defined as discharge without any procedure within 72 hours. In the 3 studies that included this outcome, unnecessary transfers were reported at 8.8%, 15.6%, and 19%.

Comorbidities

Thirteen articles reported comorbidities. Comorbidity scores were difficult to calculate or assess based on lack of information and different scoring systems which prevented comparison. Additionally, different studies focused on different comorbidities, making it difficult to report/compare comorbidities between studies. In the 13 studies with available information, the percent of patients who were transferred and had diabetes was 13%–47.7%; hypertension, 42.3%–82%; congestive heart failure, 2.5%–24.3%; renal failure, 3.8%–11%: chronic obstructive pulmonary disease, 9.2%–25.2%; and history of tobacco use, 20%–52%. The percent of patients who were not transferred and had diabetes was 7.7%–50%; hypertension, 26.3%–70%; congestive heart failure, 0.6%–18%; renal failure 0.9%–16%; chronic obstructive pulmonary disease, 3.1%–12%; and history of tobacco use, 10%–36%. We could not report specific comorbidity score because of limited data.

Outcomes: Length of Stay (LOS), Intervention, Reoperation, and Cost

LOS: Fifteen articles reported LOS. Patients who were transferred had a median LOS that ranged from 4 to 14.4 days. Patients who were admitted locally had a median LOS that ranged from 2 to 5.8 days (Appendix 2).

Intervention: Eleven articles reported an intervention. Some studies reported operations only, and others included other procedural interventions such as endoscopic procedures. Patients who required intervention after transfer ranged from 15% to 64% in the EGS population. Some of the most common interventions that were reported included the following standard operations such as laparoscopic appendectomy, cholecystectomy, hernia repair, colectomy, sigmoidectomy, small bowel resection, lysis of adhesions, wound debridement, ruptured abdominal aortic aneurysm repair, and endoscopic interventions.

Reoperation: Six EGS articles reported reoperation. EGS patients who were transferred that required reoperation ranged from 9.1% to 55%. EGS Patients who were not transferred that required reoperation ranged from 3.4% to 5% (Appendix 2).

Cost: Five articles reported cost, and 3 specifically compared the cost of patients transferred to the cost of patients who were not transferred. In all 3 of these articles, the cohort that was transferred had a higher cost than the cohort that was not transferred (Appendix 2).

DISCUSSION

Several key themes emerged from this review. These studies validated the observation that more patients with EGS diagnoses are being transferred to larger centers with the availability of specialized care. The studies noted that patients who are transferred have higher mortality and an increased number of complications. Patients who transferred were also found to have higher resource utilization as noted by LOS and cost. All studies that reported and compared cost (5) and LOS (13) between transferred and direct admit patients showed higher overall cost and longer LOS in the transferred patient cohort. Patients who are transferred with EGS diagnoses were older and transferred for a variety of reasons such as escalating the level of care for the acute problem, associated comorbidities, or a lack of available expertise at the local level. Given the lack of formal regionalization for these types of patients, many of the articles concluded that a standardized approach to transfer is necessary. In addition to trauma, protocol-based transfer in vascular surgery has been studied and demonstrated a safe and efficient transfer system [23]. EGS transfer system can use both the skeleton of the trauma and vascular system to help develop a protocol that appropriately handles the uniqueness of EGS patients. In addition, 30-day mortality, insurance status, comorbidities, and a determination of the need for transfer were able to be evaluated across the studies.

Unsurprisingly, we found that EGS patients who were transferred had a higher rate of mortality when compared to EGS patients who did not transfer. All studies that compared mortality between patients who transferred versus those who did not showed an increase in mortality in transferred patients. Most studies showed a higher mortality rate in EGS transfer patients (ranging from 2.3% to 7.5%) than the general rate of mortality of all EGS surgery including urgent 2.3%, emergent 3.7%, and elective 0.4%, respectively [24].

EGS patients are facing a serious illness regardless of their baseline status, and those who are transferred are among the most acutely ill as demonstrated by the high percentage of patients with serious chronic illnesses. In general, greater than 10% of EGS transfer patients require care in the ICU [25,26] during their hospitalization, something small/rural hospitals may not have. Small and critical access hospitals can provide quality surgical care, but if they do not have the requisite resources in place to adequately manage complicated postoperative patients, such as patients who require invasive monitoring after surgery [27], they will struggle to adequately provide care. Patients who are transferred and end up in the ICU at the referring hospital likely contribute to the observed increased mortality. Patients with a higher chronic disease burden have higher mortality, and high-risk patients are more likely to die regardless of whether they are transferred.

There are many reasons that necessitate transfer including the presence of comorbidities, escalation of care, patient preference, and previous relationship between patient and accepting surgeon. The most common reported reasons for patient transfer included requirement of a higher level of care and continuity of care with a surgeon who had been previously involved in the care, whereas patient/family requests contribute to a small percentage of patients who are transferred [28,29]. Transfer because of a need for a higher level of care is a broad category and does not give us significant information on the patient status. Additionally, it depends on providers giving a reason for transfer. This is important to understand as communication between hospitals plays an important role in patient transfer. There are times in which the initial accepting surgeon is not the same surgeon on service when the patient finally gets transferred to the new facility [7]. In our review, a comprehensive understanding of the specific reasons for transfer was not able to be captured.

Whereas mortality rates remain elevated in patients requiring transfer, overall mortality has decreased over time. Huntington et al compared outcomes following transfers during 2 different time periods (2005–2008 and 2009–2012), and the later time period showed a decrease in mortality and major complications, indicating that outcomes following transfer have improved. The study also reported that transfer patients in the second period of time had a lower Charlson Comorbidity Index score, suggesting a possible lower threshold to transfer [7].

Surgeons working at community or rural hospitals may refer patients because they feel they are not able to adequately care for the patient, either because the issue is outside their scope of practice or because they do not have adequate resources to properly manage the patient (ICU beds, nursing staff, etc). Conversely, accepting surgeons at tertiary facilities typically accept all patient transfers upon request [18]. The current pathway to transfer is not well defined, and as discussed below, some patients are transferred and never undergo an intervention.

When transfers happen, they typically occur for a reason and patients may have better outcomes upon transfer. One study included in the review noted that high-risk patients/procedures, defined as a risk above 4%, had better outcomes when the procedure was done at a high-volume center [30]. This would support high-risk patients transferring to tertiary centers. There is a fine line when trying to determine who should transfer versus who can stay at their presenting hospital. A standardized approach to EGS patients can help decide who would likely benefit to transfer and who can receive care at their original facility.

With the increase in EGS patients who are transferred, there has also been an increase in patients who undergo unnecessary transfer, defined as discharge within 3 days without an intervention [29]. This ranged from about 10% to 20% in the articles included in the study. Even at the low end, there is significant room for improvement. The most common outcome for a transferred patient was no intervention. That is not to say patients who were transferred and did not undergo an operation were inappropriately transferred. Escalation of care may be due to higher level of care outside the operating room or for management of comorbid conditions that are exacerbated by having an EGS diagnoses.

In patients who did require an intervention, most procedures were within the scope of a general surgeon, such as cholecystectomy, wound debridement, lysis of adhesions, small bowel resection, colectomy, and appendectomy. This finding may suggest lack of access to basic surgical care [17]. Current general surgery residency graduation requirements include competence in each of the mentioned procedures, implying that general surgeons should be able to operate on the majority of EGS diagnoses [31]. Previous studies have reported that residents of rural areas, individuals who are historically underrepresented by race, and the socioeconomically disadvantaged are at high risk of living in an EGS desert with little or no access to EGS care [1]. More recently, research has shown that up to 25% of patient transfers were potentially avoidable and that hospital-related factors were more likely to account for potentially avoidable transferred compared to patient-related factors [32,33]. Using this information is a starting point when trying to adapt a better transfer system. Importantly, patients undergoing an unnecessary transfer can negatively impact the availability of beds, leading to a delay in patient care, and may also unnecessarily increase the number of resources necessary to treat the problem.

However, one of the most important limitations in each study and thus in this review is the lack of information regarding what led to the decision for transfer. In other words, was a general surgeon unavailable or did the hospital lack personnel to provide care? Was a surgeon consulted prior to transfer? Who made the final decision to transfer the patient? What was the physiologic status of the patient? Lastly, we were unable to ascertain what percent of patients had transfers due to patient or family request.

Comorbidities in the Transfer Population

Both groups of patients appear to have high levels of comorbidities such as diabetes and cardiopulmonary diagnoses. However, there was limited use of acuity scores, and when used, different scoring systems created a challenge in comparing the groups. In addition, the lack of access to pretransfer records in all studies makes it difficult to determine if comorbidities in individual patients contributed directly to the transfer decision. Transfer of a patient adds to the complexity in caring for the patient.

Although we do not understand the actual reasons for transfer, chronic illness burden underlying acute surgical illness often mandates a multidisciplinary approach that transcends the availability of a qualified surgeon. The high levels of chronically ill patients in both groups was an unsurprising find as the majority of all hospitalized patients have at least 1 comorbidity [28,29].

Insurance Status Among the Cohorts

Insurance status, specifically whether a patient has health insurance, is important as patients who are uninsured or Medicaid beneficiaries are more likely to be transferred than admitted to hospital [34]. Insurance does not appear to play a role in the transfer decision, and only small percent of patients, regardless if transferred or not, did not have insurance. Fifteen articles using the NIS, NSQIP, and data from their own health care system demonstrate no difference in the patient cohorts. Decreased uninsured rates are significant because the reduction of the uninsured rate leads to patients presenting earlier in the course of their surgical emergency [35]. Another interesting finding was that, in transferred patients, the most common insurance was Medicare. Older patients tend to have chronic diseases which can make acute conditions worse. This has implications for rural general surgery care as the rural population tends to be older (more likely to be on Medicare) and less healthy. Rural patients are at an increased likelihood of requiring advanced surgical care (increased age and sicker baseline health), which coupled with the scarcity or rather unequal distribution of general surgeons puts patients with decreased access to appropriate care [36]. This is consistent with previous research as it has shown that EGS patients who live in rural communities are more likely to live without appropriate access to EGS care [1]. It should be noted that in most instances, these studies took place across the timeframe for the implementation of the Accountable Care Act, perhaps limiting the impact of insurance status on the transfer decision.

Lack of Physiologic Status Information Prior to Transfer

The most commonly reported study limitation was lack of access to the physiologic status of patients throughout their illness. Many of the articles used nationwide databases like ACS NSQIP or NIS. Although these resources are helpful to grasp a high-level view of patient cohorts and disease categories, they lack the granularity required to help understand the experience at the local level. These databases contain information such as inpatient utilization, access, charges, quality, and outcomes. Patient outcomes have improved since the introduction and widespread use of NSQIP, although using these databases alone will not achieve improvement in surgical quality [37,38]. Access to physiologic data can give researchers more data to help create a standardized EGS transfer pathway.

The lack of sharing of information between various electronic health record platforms inhibits the ability to capture the status of patients prior to transfer in databases. Assessing the acuity of surgical illness in the setting of chronic disease is likely to have led to many transfer decisions. However, this is just conjecture because none of the studies were able to obtain accurate and complete pretransfer information even when desired.

Collaboration between key stakeholders is needed to obtain physiologic and decision-making information across the transfer space with shared electronic health record information. Ultimately, a standardized pathway can improve patient outcomes as it can help us better allocate resources to care for EGS patients while cutting down on the number of unnecessary transfers. Determining which risk factors are associated with worse outcomes can help inform decision-making ability on which patients will benefit from transfer and how to improve resource allocation and appropriate level of care [39].

To develop a standard pathway, a better understanding of patient status/reason for transfer is needed. Commonly used comorbidity scores such as Charlson Comorbidity Index and Elixhauser scores take disease burden into account. An issue with these scoring systems is they focus on conditions patients have at baseline rather than physiologic status during their acute event [40]. Acute Physiology and Chronic Health Evaluation II is another acuity scoring system but is typically used in ICU care, making it less useful in EGS care [41]. Adding physiologic status to the database, such as vitals, pertinent laboratory test results, imaging findings, and physical examination findings could help develop a scoring system that would give surgeons another tool in determining who should transfer. In clinical practice, more consistent use of the Emergency Surgery Score, a scoring system that has been shown to accurately predict mortality in patients undergoing emergency surgery in multiple surgical specialties, could help predict which patients may need to transfer [42]. Regardless, it would be helpful to have physiologic information in large databases to help further understand trends.

Solutions

A more coordinated regionalized response could better allocate resources and efficiently manage patients as close to home as possible. Another obvious solution is the increased use of telehealth to support care closer to home or assess need for transfer. This has significantly increased in the time of COVID-19 pandemic, and its use will likely stay at a higher rate than prior to the pandemic. Although not the same, a study looking at vascular patients reported that a third of cases in which a transfer request was made did not end up in transfer. This most common reason patients stayed occurred because of reassurance from consulted vascular surgeon. A similar collaboration model could be used with EGS patients. If the goal is to get the right treatment to the right patient at the right time, communication and support between surgeons at referring and accepting centers are vital. Often, many hospital systems do not communicate well with each other. At least 1 group of investigators has recently published an article that was able to capture information such as availability of an on-call surgeon, hospital capacity, patient request, patient status, availability of appropriate diagnostic and treatment testing, and specialist availability as dimensions of the transfer decision.

Limitations Across Studies

We recognize the key limitation of our study, and those that we evaluated question how much of a true difference exists between patients who are transferred and patients who are not transferred. Patients in the different groups were similar based on demographics. Physiologic status was unable to be assessed in any of the studies in our review. We recognize that chronic illness burden underlying acute surgical illness is not just about the surgeon and the operation but also other providers, health care workers, and institutions that provide the highest level of care. What drives the decision to transfer? It remains unclear (Appendix 2).

In conclusion, patients who are transferred for EGS are older and have a high burden of chronic illness. A major limitation to understanding the decision for transfer is the limited ability to obtain pretransfer information about physiologic status and care prior to transfer. With the lack of data sets and no national registry for EGS patients, management is variable, and there is a need for a national EGS data registry that include patients who were operatively or nonoperatively managed and transferred or nontransferred, through which research and national QI initiatives could be developed consistently. This type of information deserves further investigation and understanding of pretransfer environment as well as patient status and hospital access to specialty care. Having accurate patient information that is shared across electronic health platforms and across patient care settings will facilitate better understanding of what leads to the transfer decision. We hope that this will improve patient outcomes and facilitate the development of transfer protocols for the EGS patient [43].

Author Contribution

Ryan D. Emanuelson: conceptualization, data curation, formal analysis, writing of the original draft, and review and editing.

Sarah Jane Brown: methodology (search strategy and protocol), writing of the methods section, and review and editing.

Paula M. Termuhlen: conceptualization, data curation, formal analysis, review and editing, supervision.

Conflict of Interest

None of the authors have any conflict of interest.

Funding Source

This research did not receive any grant from funding agencies in the public, commercial, or not-for-profit sectors.

Ethics Approval

Study exempt from required IRB review due to type of study.

	Limitations	Conclusions
1	Regional study.The definition of clinically unnecessary was based on consensus among authors who were intentionally conservative in the definition.Unable to capture data on patients who were transferred but not admitted	About 20% of EGS are unnecessary. All stakeholders can benefit from transfer guidelines
2	NIS does not include physiologic data.	Hospital-level characteristics better predict need for transfer than patient-related factors.
3	Time spent at the referring hospital or emergency department (ED) was not documented for all patients	EGS transfer patients are an extraordinarily costly population with worse outcomes. Individual referral hospitals and health systems must recognize the significant impact of this population. Resource allocation, care pathways, transfer agreements, and even system regionalization should be considered.
4	Changes in EHR and subsequent lack of systems to collect data on transfer patients.Not able to reliably collect data such as time at outside facility, physiologic status at outside facility, and time of acceptance versus time of arrival.	Although this study confirms transfer patients need the resources for which they were referred to a tertiary center, nearly half of transfer patients undergo basic surgical procedures or do not require intervention. This points to a potential lack of general surgery resources in the community.
5	Absence of physiologic data and dependence on accurate coding.Using NIS, unable to link patients to their previous hospitalization to examine their pretransfer care.	An increasing number of patients with EGS diagnosis are undergoing transfer. Less than half of EGS transfer patients require surgical intervention. Mortality and length of stay are decreasing in the transfer population.
6	Inability to analyze all patients cared for in the referral centers without a transfer, identify the precise details of every transfer, and predict the outcomes if the transfer had occurred earlier.	Transfer patients have worse outcomes compared to patients admitted locally. It is unknown if outcomes are influenced by delays in transfer or inadequate initial care, or simply the result of a more serious disease. Nearly 4 d before transfer to definitive care likely caused harm. Eight diagnoses that, coupled with significant comorbidities and physiologic derangement, may be appropriate as a starting discussion point about regionalization of nontraumatic surgical emergency care.
7	NSQIP does not contain details about the transferring facility.	Nationwide, the rate of patients transferred for surgery is increasing. Although complication rates are markedly higher in transferred patients than in similar nontransferred patients, outcomes have improved for transferred patients over time.
8	Physicians consider several factors when determining whether or not a patient should transfer. This is not reflected in the NSQIP database.	Transferred patients that underwent EGS had a small increased risk of mortality and morbidity compared with patients admitted directly from home. This suggests that it is generally safe and reasonable to transfer EGS patients and regionalization of EGS care is feasible from a patient outcomes standpoint.
9	NIS contains no data on anatomic severity of disease, physiologic status of the patient, and whether diagnosis was present on admission or not, all of which may have a significant impact on patient outcomes.	EGS patients with predicted risk of death of 4% or higher may benefit from transfer to high-volume EGS hospitals (shock, ruptured abdominal aortic aneursym, deep venous thrombosis/pulmonary embolism, liver disease, perforation of intestine, esophagus disease, small intestinal cancer, bowel ischemia, colorectal cancer, nontraumatic bladder rupture, and peritonitis.
10	Data were from a single academic medical center, which limit generalizability to other centers, particularly nonacademic referral facilities. The study population is relatively heterogeneous, encompassing 4 surgical disciplines. This may increase the generalizability of findings but limits the ability to make diagnosis-specific conclusions or recommendations.	Duration of referring facility care before transfer request did not impact posttransfer mortality or discharge to hospice. This suggests effective triage of high-risk patients and permits time to determine transfer appropriateness for lower-risk patients taking into consideration patient needs and available resources.
11	Absence of physiologic status, lack of detail about the hospital stay prior to transfer and referring hospital resources, and inability to track critical care services.	Interhospital transfer in the EGS patient population increases the odds of mortality and is more costly than direct admissions, even after controlling for multiple other contributing factors.
12	Although the use of our severity of illness (SOI) indicators (perforation, SIRS/sepsis/shock, peritonitis, and acute renal failure) strongly predicted poor outcome, there were no standard SOI makers available to stratify the study population.	Age, severity of illness, and distance from a regional referral center explain much of the variation in mortality and can be used for triage to regional EGS centers.
13	Nationwide Emergency Department Sample (NEDS) does not allow for characterization of the posttransfer hospitalization.Unable to qualify the outcomes of EGS patients who are transferred from EDs, including morbidity, mortality, and cost.NEDS is a population-level data set and is not validated to address specific clinical problems.Cannot identify how many different hospitals each facility transfers to nor can we characterize the clinical services available and surgical capacity at transferring facilities	Medically complex and older patients who present at small, rural hospitals are more likely to be transferred. Future research on the unique needs of rural hospitals and timely transfer of EGS patients who require specialty surgical care have the potential to significantly improve outcomes and reduce costs.
14	Data regarding clinical reasoning prompting transfer for the EGS consult cohort were not available, as the consulting surgeon did not speak directly with the transferring facility in those situations.Because specialty services were not included in data collection, overlaps between diagnoses that could be managed by either an EGS or a specialty service could have been missed.	EGS transfers burden patients and increase health care costs. Prior studies have concluded that many EGS transfers are unnecessary, but they have failed to identify the clinical reasoning prompting these transfers. The study indicates that this process is frequently based on subjective reasoning and often without direct communication with the accepting surgeon.
15	Time spent in the referring hospitals is not measured, so it is not possible to know if definitive care was truly delayed.NSQIP only supplies data for patients who underwent operations. There may be a significant number of patients who underwent nonoperative management not included in this study.Cost and charge data are not available in NSQIP, but relative costs of resources such as LOS, OR time, and critical care complications have been well established.	Emergency general surgery patients who are transferred have significantly higher mortality, morbidity, and resource utilization. The type of transferring setting (EDT, IPT, and NHT) also makes a difference. With the progression to a more regionalized health care system, value will be achieved by improving outcomes while lowering costs. Referral centers must be prepared to assume outcome and financial risk as they receive EGS patients.
16	Single academic medical center, potentially limiting the generalizability of our results.Study did not include patients who were initially transferred to medical services who subsequently consulted or ultimately transferred the patient to the EGS service.Some of the results were dependent upon the documentation provided by the referring hospitals.	Study documented the provision of care to patients with a range of EGS diagnoses transferred to a tertiary medical center, including specifics regarding the characteristics of and the care provided at the referring facility as well as details regarding the timing of transfers. A need for general surgery or specialty services and a need for higher level of care were found to be major contributors to interhospital transfers for EGS conditions. At the same time, approximately one third of the patients transferred to the center did not undergo a procedure following transfer.
17	NSQIP membership is overrepresented by tertiary centers. This selection bias likely resulted in an overestimate of the incidence of transfer.Many different operation types were pooled together in the analysis; the risks associated with transfer can vary by operation type.	The incidence of interhospital transfer in surgery is high. Transferred patients have worse outcomes than nonelective direct admissions. However, this difference is largely due to confounding by patient factors, as sophisticated adjustment techniques nearly equalized the risk for adverse outcomes.
18	Patient and perioperative characteristics are not comprehensively available.Data were limited to the variables available in NSQIP. Variables not available in this analysis include readmission as well as minutes/hours from arrival to operation.	The type of transferring institution has a significant impact on postoperative risk adjusted morbidity and mortality after emergent colon surgery, with patients originating at outside hospital wards and/or nursing home/chronic care facilities demonstrating the worst postoperative outcomes.
19	Large sample size based on the NIS derived from claims data. Possible that findings are due to unequally distributed patient-level variables that are not captured by the NIS data set.Elixhauser comorbidity index was used and does not control for acute physiologic derangement.	In patients undergoing surgical intervention for necrotizing fasciitis, interhospital transfer is associated with increased mortality compared to patients undergoing definitive management at the presenting hospital.
20	NSQIP indicates if the patient had a previous operation within 30 d of the index surgery. However, the details of the procedures performed are not available in the data set.NSQIP definition of emergency is somewhat subjective, primarily relying on the surgeon or anesthesiologist to document the case as being an emergency.NSQIP does not capture specific details of the transfer process (such as reason for transfer and potential delay in surgical management that resulted from transfer).	Interhospital transfer status is not an independent risk factor for poor clinical outcomes after surgical management of necrotizing soft tissue infection (NSTI). Although expedient surgical debridement will always remain a basic tenet of NSTI management, the findings of the study provide some reassurance that transfer of NSTI patients before initial surgical management will not significantly jeopardize their outcomes but may increase their length of stay at the accepting hospital should such transfer be deemed necessary
21	NIS is an administrative database of discharge records from US hospitals and lacks information on the physiological status of the patient.Information regarding the transfer process, such as reason for transfer or delays in transfer, was unavailable.Determining the timing of procedures relative to presentation and tracking patients between hospitals in the NIS is not possible.	Transfer status is an independent risk factor for poor outcomes. Additional studies are needed to understand the mechanisms by which patient transfers may lead to poorer outcomes and ameliorate the significant morbidity and mortality experienced by these patients.

	Author, date	Setting	Participants (#)	Length (y)	Ave age (y)T/NT	% MaleT/NT	Database	Cost ($)	LOS (d)T/NT	Intervention (%)	Reoperation (%)
1	Kummerow-Broman 2016	U	21,77	5	57	53	HCS	†
2	Ingraham, 2018	U	17,236,701	6	61.6/58.8	50.1/46	NIS
3	Keeven, 2018	U	663	2	LA: 52; EDT: 58; IPT: 62;	LA: 49; EDT: 50); IPT: 48	HCS	LA: 3,954; EDT: 5,212; IPT: 9,551	LA: 3; EDT: 4; IPT: 7	43	IPT: 55
4	Miscercola, 2016	U	772	1	61.2/54.7		HCS		Median: 4/2	31
5	Reinke, 2017	U	525,913	10	60	49	NIS	Median: 8,213; intervention: 15,425; no intervention: 5,601.	Median: 4.4*	Operation: 33; procedure: 21
6	Santrey, 2011	U	319	3.08	59.2/55	53.4/50.9	HCS		8/
7	Huntington, 2015	CH	1,474,531	8	60.5/57.8	52/43.3	NSQIP		14.4/5.8		10/4.3
8	Castillo-Angeles, 2019	MC	222,519	10	55.5/44	48.1/48.7	NSQIP		Median: 5/2		9.1/3.4
9	Ogola, 2018	MC	3,006,615	1	78.1/59		NIS		High risk: 5; low risk: 3
10	Kummerow Broman, 2016	U	2091	5	57	53	HCS			15
11	Yelverton, 2018	CH	25,021,217	10	60/58	49/46	NIS	7,742/5,820	4.4/3	Broad: 21; narrow: 33; any procedure: 45%
12	Diaz, 2010	U	3,439	4.75	47 (survivors)/64 (not survivors)	Survivors: 47%; not survivors 46%	HCS	Survivors: $25,612; not survivors: $161,653	Overall: 6.4 d; survivors: 4 d; nonsurvivors: 12 d	Survivors: 67.8%; nonsurvivors: 80%
13	Fernades-Taylor, 2021	U	47,442,892	5	57/42.3	47.4/41.6	NEDS
14	Bruenderman, 2020	U	200	0.41	59		HCS		4 d or less (42%); 5–9 d (30%); 10 d or more (28%)
15	Keeven, 2019	U	167,636	3			NSQIP				EDT (4.8%), IPT (7.0), NHT (7.3), DA (2.9)
16	Philip, 2019	U	334	2	60	45.2	HCS			64
17	Lucas, 2013	U	53,464	1	59/54	46/46	NSQIP		7/4		10/5
18	DeWane, 2018	U	12,245	2	EDT (64.3) IPT (63.8), NHT (75), DA (63.9)	EDT (54.8) IPT (54.1), NHT (61.4), DA (53.8)	NSQIP		EDT (10), IPT (14), NHT (13), DA (10)		EDT (14.5), IPT (15.6), NHT (11.7), DA (10.1)
19	Holena, 2011	U	9958	6	52/53	55/57.2	NIS
20	Ingraham, 2017	U	1801	5		59.9/59‡	NSQIP		16/14
21	Philip, 2020	MC	10,730,245	3	60.1/58.7	49.1/45.6	NIS	8,687/6,759	4.3/3.0

#	Search statement
1	General Surgery/
2	exp Surgical Procedures, Operative/
3	(general surgery or surgical procedure* or surgery or surgeries).tw.
4	1 or 2 or 3
5	exp Emergencies/
6	exp Emergency Medicine/
7	exp Emergency Service, Hospital/
8	exp Emergency Medical Services/
9	(emergency or emergencies or emergent or acute or urgent).tw.
10	5 or 6 or 7 or 8 or 9
11	exp Patient Transfer/
12	exp "Transportation of Patients"/
13	("transportation of patients" or interhospital transfer* or inter-hospital transfer* or interfacility transfer* or inter-facility transfer* or transfer patient* or (patient* adj5 transfer*)).tw.
14	11 or 12 or 13
15	exp Hospitals, Rural/
16	exp Rural Health Services/
17	(critical access hospital* or rural or remote or regional*).tw.
18	15 or 16 or 17
19	4 and 10 and 14 and 18

#	Search statement
1	exp surgery/
2	(general surgery or surgical procedure* or surgery or surgeries).tw.
3	1 or 2
4	exp emergency/
5	exp emergency medicine/
6	exp hospital emergency service/
7	exp emergency health service/
8	exp emergency ward/
9	(emergency or emergencies or emergent or acute or urgent).tw.
10	4 or 5 or 6 or 7 or 8 or 9
11	exp patient transport/
12	("transportation of patients" or interhospital transfer* or inter-hospital transfer* or interfacility transfer* or inter-facility transfer* or (patient* adj5 transfer*)).tw.
13	11 or 12
14	exp rural health care/
15	exp community hospital/
16	(critical access hospital* or rural or remote or regional*).tw.
17	14 or 15 or 16
18	3 and 10 and 13 and 17