Surgery

Factors associated with Interhospital transfers of emergency general surgery patients from emergency departments

a b s t r a c t

Background: Emergency general surgery (EGS) conditions account for over 3 million or 7.1% of hospitalizations per year in the US. Patients are increasingly transferred from community emergency departments (EDs) to larger centers for care, and a growing demand for treating EGS conditions mandates a better understanding of how ED clinicians transfer patients. We identify patient, clinical, and organizational characteristics associated with inter- hospital transfers of EGS patients originating from EDs in the United States.

Method: We analyze data from the Agency for Healthcare Research and Quality Nationwide Emergency Depart- ment Sample (NEDS) for the years 2010-2014. Patient-level sociodemographic characteristics, clinical factors, and hospital-level factors were examined as predictors of transfer from the ED to another acute care hospital. Multivariable logistic regression analysis includes patient and hospital characteristics as predictors of transfer from an ED to another acute care hospital.

Results: Of 47,442,892 ED encounters (weighted) between 2008 and 2014, 1.9% resulted in a transfer. Multivar- iable analysis indicates that men (Odds ratio (OR) 1.18 95% Confidence Interval (95% CI) 1.16-1.21) and older pa- tients (OR 1.02 (95% CI 1.02-1.02)) were more likely to be transferred. Relative to patients with private health insurance, patients covered by Medicare (OR 1.09 (95% CI 1.03-1.15) or other insurance (OR 1.34 (95% CI 1.07-1.66)) had a higher odds of transfer. Odds of transfer increased with a greater number of comorbid condi- tions compared to patients with an EGS diagnosis alone. EGS diagnoses predicting transfer included resuscitation (OR 36.72 (95% CI 30.48-44.22)), cardiothoracic conditions (OR 8.47 (95% CI 7.44-9.63)), intestinal obstruction

(OR 4.49 (95% CI 4.00-5.04)), and conditions of the upper gastrointestinal tract (OR 2.82 (95% CI 2.53-3.15)). Relative to Level I or II trauma centers, hospitals with a trauma designation III or IV had a 1.81 greater odds of transfer. Transfers were most likely to originate at Rural hospitals (OR 1.69 (95% CI 1.43-2.00)) relative to urban non-teaching hospitals.

Conclusion: 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.

(C) 2020

  1. Introduction

Emergency general surgery (EGS) involves treating common condi- tions such as complicated diverticulitis, intestinal obstruction, and ap- pendicitis [1]. These conditions account for over 3 million (or 7.1%) hospitalizations per year and have increased 150% over the last 10 years [2,3]. As the population ages and access to emergency surgical care declines, patients are increasingly transferred to larger centers for care [1,2,4,5]. Most of these transfers originate from smaller community

* Corresponding author at: 600 Highland Ave. CSC K6/140, Madison, WI 53792-7375, United States of America.

E-mail address: [email protected] (S. Fernandes-Taylor).

or freestanding emergency departments (EDs) that are challenged by limitED capacity and resources to care for these patients [6]. The declin- ing availability of on call specialists as well as the consolidation of healthcare services will further lead to transfers. Consequently, the growing demand for treating these conditions and disparities in access to these services mandates a better understanding of how ED clinicians transfer patients.

Despite the need for community hospitals to transfer patients, there are no clear guidelines to inform either patient selection or the informa- tion that should be included in hospital communications to ensure high quality handoffs. This gap in guidance has led to inconsistent transfer decision making and worse outcomes for EGS patients who are trans- ferred, creating an imperative for further study [2,4,7]. Prior research

https://doi.org/10.1016/j.ajem.2020.12.012

0735-6757/(C) 2020

suggests that 20% of surgical transfers are deemed potentially unneces- sary either because patients did not need specialty surgical intervention or because patients were too sick to benefit from a higher level of care [8]. At the same time, delays in transfers have been shown to lead to worse outcomes for patients. This, coupled with recent evidence that suggests that on average EGS patients admitted to centers with high quality trauma programs or to hospitals with high volume EGS admis- sions have lower mortality rates, suggests a need for a better approach to identify patients in need of transfer and that they end up in an appro- priate center capable of treating these patients succesfully [9,10]. EGS patients have the potential to benefit from transfer protocols designed to facilitate timely transfer and streamline interfacility communication to confer better outcomes, much in the way that trauma and myocardial infarction patients have [11-13].

We identify clinical and organizational characteristics associated with Interhospital transfers of EGS patients originating from the ED. We utilize National Emergency Department Sample (NEDS) data, which, to the best of our knowledge, have not been used previously to characterize patient transfers from a clinical decision-making perspec- tive [14]. We anticipate that our results will help to inform the develop- ment of protocols that facilitate the early identification of EGS patients requiring transfer.

  1. Methods
    1. Study design and setting

We analyzed data from the Agency for Healthcare Research and Quality Nationwide Emergency Department Sample (NEDS) for the years 2010-2014. We chose NEDS to answer our question of interest be- cause it is the largest all-payer emergency department (ED) database in the United States and is designed to produce national estimates of hospital-based ED visits. The NEDS is a Healthcare Cost and Utilization Project database and is created by sampling the State Inpatient Data- bases and the State Emergency Department Databases. The NEDS in- cludes information on clinical and resource relevant variables and contains approximately 31 million ED visits per year (unweighted). We identified adult patients (>=18 years old) with EGS conditions using American Association for the Surgery of Trauma (AAST) ICD-9-CM diag- nosis codes (n = 47,442,892) [1,15]. We used these data to describe re- ferring hospital and patient characteristics, employing weights to provide national estimates. The discharge weights are provided with the data and were calculated by stratifying NEDS hospitals on the vari- ables used to create the sample: geographic region, trauma center des- ignation, urban/rural location, teaching status, and ownership [14]. A weight was calculated for each stratum by dividing the number of na- tional ED visits in that stratum (from American Hospital Association data) by the number of NEDS visits in the stratum. The weighted esti- mates thus reflect the number of ED visits nationwide.

    1. Measurements

We characterized patients based on EGS diagnosis indicator var- iables defined by International Classification of Diseases, Ninth Revi- sion, Clinical Modification (ICD-9-CM) codes categorized into the following diagnosis or systems-based groups: resuscitation, abdom- inal conditions, upper gastrointestinal tract, intestinal obstruction, hepatic-pancreatic-biliary, colorectal, hernias, soft tissue, vascular, cardiothoracic, and other. These categories are described in detail elsewhere [16]. Patients’ sociodemographic characteristics included sex, age, Insurance type (Medicare, private insurance, Medicaid, self pay, no charge, other), and the quartile of the ZIP code-level median income for the patient’s residence. Patient-level clinical characteris- tics included Charlson Comorbidity Index, any surgical procedure performed before transfer (defined using ICD-9-CM procedure codes), and whether the admission occurred on a weekend [3,15,17,18].

Hospital-level characteristics included total number of ED visits (irrespec- tive of diagnosis), trauma center status, rural or urban location, teaching status, region of the US (northeast, south, midwest, west), and hospital ownership (government, non-profit, private/investor-owned).

    1. Outcomes

Our primary outcome variable was any transfer to an acute care hos- pital. Patient-level socioDemographic and clinical factors as well as hospital-level factors were examined as predictors of transfer from the ED to another acute care hospital. Missing values were excluded listwise.

    1. Analysis

We report summary statistics for patient and hospital characteristics as percentages for categorical variables and as means/standard errors or medians/interquartile ranges for continuous variables. All estimates are weighted to be nationally representative [14]. We describe bivariate dif- ferences between patients who were transferred and patients who were not transferred using Chi-squared analyses, t-tests, and Wilcoxon- Mann-Whitney tests as appropriate. Multivariable analysis includes pa- tient and hospital characteristics as predictors of transfer from an ED to another acute care hospital. We use a multilevel, logistic regression model to assess predictors of transfer. Our approach uses a competing risk analysis to account for death and transfer to non-Acute care facilities [19]. In analyses that assess EGS diagnosis groups that predict transfer, absence of the diagnosis is used as the reference. Data analysis was per- formed using SAS 9.2 software. This study was deemed exempt by the University of Wisconsin-Madison Institutional Review Board.

  1. Results

Of 47,442,892 ED encounters (representing approximately 5% of ED visits, weighted) for EGS conditions between 2008 and 2014, 1.9% re- sulted in a transfer. Bivariate analysis is summarized in Table 1. On aver- age, Transferred patients were older than patients who were not transferred (56 years old vs. 42 years old; p < 0.0001) and were more likely to be covered by Medicare. Transferred patients also had a greater Comorbidity burden than patients who were not transferred. The most common EGS diagnoses in the overall ED patient population were gen- eral abdominal conditions, soft tissue conditions, and upper gastrointes- tinal tract conditions. Upper GI tract and hepatobiliary conditions were more prevalent in the transferred patient population (upper GI 21% vs. 3%; hepatobiliary 14% vs. 4% p < 0.0001).

Hospital characteristics are summarized in Table 2. Transferred pa- tients were less likely to originate at a Level I or II trauma center (p < 0.001). Lower volume EDs were more likely to transfer patients (p < 0.0001). Transfers were most common in the midwestern US, and rural hospitals were most likely to transfer patients followed by urban non-teaching hospitals (p < 0.0001). Hospital ownership was not significantly associated with transfer.

The results of our multivariable analysis are summarized in

Table 3. Patients’ sociodemographic characteristics were significantly associated with odds of transfer. Men were more likely to be trans- ferred (Odds ratio (OR) 1.18 95% Confidence Interval (95% CI) 1.16-1.21), as were older patients (OR 1.02 (95% CI 1.02-1.02)). Rel- ative to patients with private health insurance, patients covered by Medicare (OR 1.09 (95% CI 1.03-1.15) or other insurance (OR 1.34 (95% CI 1.07-1.66)) had higher odds of transfer. Relative to patients in the lowest income zip codes, patients residing in zip codes in the two highest Household income groups were less likely to be trans- ferred. Odds of transfer increased with a greater number of comorbid conditions, and and patients with three or more comorbidities were significantly more likely to be transferred (OR 2.30 (95% CI 2.18-2.43)) compared to patients with an EGS diagnosis alone; the

Table 1 Characteristics of emergency general surgery patients transferred to acute care facilities in the nationwide emergency department sample 2010-2014

transfer most frequently included resuscitation (OR 36.72 (95% CI 30.48-44.22)), cardiothoracic conditions (OR 8.47 (95% CI

7.44-9.63)), intestinal obstruction (OR 4.49 (95% CI 4.00-5.04)),

Not Transferred (n = 46,534,407)

Transferred

(n = 908,485)

P value

and conditions of the upper gastrointestinal tract (OR 2.82 (95% CI 2.53-3.15)). Weekend admissions were slightly more likely to result

Age (years, mean (standard error))

Gender, % (n)

42.3 (0.05) 57.0 (0.42) <0.0001

in transfer (OR 1.03 (95% CI 1.01-1.04).

Relative to Level I or II trauma centers, hospitals with a trauma des- ignation III or IV had a 1.81 greater odds of transfer. The odds of transfer

Female 58.4 (27166343) 52.6 (477748) <0.0001

day of the week of admission, % (n)

were more than twice as high in the Midwest (OR 2.06 (95% CI

Monday-Friday

72.6 (33791032)

71.1 (645704)

<0.0001

1.56-2.72)) relative to the Northeast, and the odds of transfer were

Saturday-Sunday

27.4 (12743375)

28.9 (262781)

slightly higher in the western (OR 1.27 (95% CI 0.97-1.66)) and south-

ern US (OR 1.08 (95% CI 0.79-1.46)) relative to the northeast. Relative to government or privately owned hospitals, transfers were less likely

26th-50th

27.7 (12896753)

34.4 (312876)

51st-75th

22.4 (10420030)

19.5 (177572)

We analyzed data from the National Emergency Department Sample

76th-100th

15.7 (7286530)

10.4 (94029)

to inform efforts to identify priority emergency general surgery (EGS)

Charlson Comorbidity Index, % (n) populations and to develop protocols for streamlining interfacility

0 84.4 (39108948) 68.2 (619674) <0.0001

Primary payer, % (n)

Medicare

18.7 (8723865)

45.2 (410385)

<0.0001

Medicaid

23.5 (10937725)

14.2 (129346)

Private insurance

30.3 (14091917)

26.5 (240524)

at private, investor owned hospitals (OR 0.84 (95% CI 0.65-1.10)).

Self-pay

22.3 (10372674)

9.0 (81375)

Transfers were most likely to originate at rural hospitals (OR 1.69

No charge

1.0 (481699)

1.0 (7556)

(95% CI 1.43-2.00)) relative to urban non-teaching hospitals.

Other

4.1 (1926527)

4.3 (39298)

Median household income national quartile, % (n) 4. Discussion

0-25th 34.2 (15931095) 35.7 (324007) <0.0001

1

11.8 (5493910)

17.9 (163039)

transfers from emergency departments (EDs). To our knowledge, this

2

2.7 (1234985)

7.6 (69390)

is the first study to characterize interhospital transfers for EGS patients

3

1.5 (696564)

6.2 (56381)

using the NEDS database, a nationally representative database of ED

Diagnosis group description, % (n) visits. Our results indicate that older patients are more likely to be trans-

Hepatic-pancreatic-biliary 4.3 (2021676) 13.6 (123419) <0.0001 ferred from EDs, perhaps because they have care already established

Upper gastrointestinal tract 3.3 (1538094) 21.0 (190672)

Soft tissue

32.0 (14902138)

14.8 (134285)

elsewhere or their family has a preference regarding where they receive

Colorectal

7.0 (3241168)

6.0 (54403)

care. Transferred patients are also more likely to have multiple comor-

Intestinal obstruction

0.6 (294591)

9.7 (88349)

bid conditions and certain EGS diagnoses, including resuscitation, car-

General abdominal conditions 50.0 (23253145) 28.6 (259964) diothoracic conditions, intestinal obstruction, and upper GI tract

Vascular 0.6 (285903) 2.4 (21762)

Cardiothoracic

0.1 (34192)

1.2 (10691)

conditions. This is consistent with previous research showing that

Hernias

1.9 (875857)

1.9 (17713)

transferred patients from inpatient settings tend to be older, male, and

Other Resuscitation

0.2 (84989)

0.0 (4337)

0.3 (2890)

0.3 (2655)

have more chronic conditions [6,16,20-23]. In addition to their comor-

bidity burden, this population experiences greater delays in transfer,

odds of transfer for patients with 2 (OR 1.76 (95% CI 1.68-1.84)) or 1

(OR 1.25 (95% CI 1.20-1.29)) comorbidity(ies) were somewhat at- tenuated. From a clinical standpoint, EGS diagnoses that led to

Table 2

Characteristics of the hospitals that admitted emergency general surgery patients trans- ferred to acute care facilities in the nationwide emergency department sample 2010-2014

Not Transferred

Transferred

P value

(n = 46,534,407)

(n = 908,485)

Trauma Level

<0.0001

I or II

27.8 (12951821)

7.2 (65488)

>= III

72.2 (33582586)

92.8 (842996)

Annual visits, % (n)

<25 K

17.5 (8132878)

59.3 (538967)

<0.0001

25 K-50 K

31.9 (14821622)

23.4 (212193)

50,001-75 K

25.5 (11873659)

10.0 (90475)

>75 K

25.2 (11706247)

7.4 (66850)

Region of hospital, % (n)

Northeast

17.6 (8213303)

9.8 (88890)

<0.0001

Midwest

22.9 (10638692)

41.2 (379566)

South

40.4 (18798632)

30.0 (270695)

West

19.1 (8883780)

18.6 (169333)

Control/ownership of hospital, % (n)

Government, nonfederal

6.7 (3131485)

14.9 (135732)

<0.0001

Government or private, collapsed

64.5 (29996713)

44.9 (408324)

Private, non-profit

16.6 (7738609)

16.8 (153043)

Private, collapsed

3.8 (1789309)

15.7 (143071)

Private, investor-owned

8.3 (3878291)

7.5 (68314)

Location/teaching status of hospital, % (n)

Rural

16.8 (7821583)

50.5 (458419)

<0.0001

Urban nonteaching

40.0 (18590774)

35.0 (317885)

Urban teaching

43.2 (20122050)

14.5 (132180)

making them high value targets for early identification and timely transfer planning protocols [24,25].

We also found that patients with Public insurance and lower income ZIP codes tend to be transferred more frequently. Previous studies dem- onstrate that uninsured and publicly insured patients have 1.5 to 2 times the odds of transfer relative to privately insured patients [26,27]. Although the watershed 1986 legislation EMTALA (Emergency medical treatment and Active Labor Act) prohibits the transfer of pa- tients based on their ability to pay for treatment, the persistent associa- tion between socioeconomic indicators and transfer, even after controlling for measures of underlying patient illness, suggests that so- cioeconomic disadvantage remains a risk factor for interhospital trans- fer. This potentially occurs because the hospitals to which these patients are transferred have greater capacity for un- or under- compensated care even after controlling for hospital facility characteris- tics [28]. However, this phenomenon warrants further study. In contrast to previous work, we found that patients without insurance (self pay) were less likely to be transferred, possibly because these patients are more likely to visit the hospital for less severe conditions that could be managed on an outpatient basis [26]. Establishing protocols for transfer decision-making based on clinical indications will help ensure candi- dates of transfer are identified early and systematically to mitigate biases.

Our results indicate that hospitals in rural areas, those with a Level III or IV trauma designation, and facilities in the Midwest were more likely to transfer patients out of their EDs. This rural effect is robust across health systems, and has increased over time owing to the consolidation of hospitals and the declining availability of on-call specialitsts [2,22]. For EGS patients who are transferred to a higher level of care, fewer than half require surgery [2]. Therefore, potential drivers of transfer for these patients may include rural surgeons’ or anesthesiologists’

Table 3

Multivariable analysis predicting transfer of emergency general surgery patients to acute care hospitals in the nationwide emergency department sample 2010-2014

Parameter

Unadjusted Odds Ratio

Unadjusted 95% CI

Adjusted Odds Ratio

Adjusted 95% CI

Gender Male

1.27

1.23-1.30

1.18

1.16-1.21

Female

1.00

1.00

Age

1.040

1.038-1.042

1.023

1.021-1.024

Expected primary payer

Medicare

2.76

2.48-3.07

1.09

1.03-1.15

Medicaid

0.69

0.63-0.77

0.95

0.87-1.04

Self-pay

0.46

0.40-0.53

0.71

0.62-0.80

No charge

0.92

0.42-2.03

1.72

0.76-1.89

Other

1.20

0.97-1.47

1.34

1.07-1.66

Private insurance

1.00

1.00

Household income of patient’s zip code (median)

26th to 50th percentile

1.19

1.08-1.31

0.96

0.88-1.05

51st to 75th percentile

0.84

0.75-0.93

0.89

0.81-0.97

76th to 100th percentile

0.64

0.53-0.77

0.82

0.69-0.99

0-25th percentile

Charlson Comorbidity Index 1

1.00

1.87

1.78-1.97

1.00

1.25

1.20-1.29

2

3.55

3.31-3.80

1.76

1.68-1.84

3

5.11

4.74-5.51

2.30

2.18-2.43

0

EGS diagnosis group Resuscitation Yes

1.00

83.8

73.3-95.8

1.00

36.72

30.48-44.22

Resuscitation No

1.00

1.00

General abdominal conditions Yes

0.40

0.37-0.44

0.40

0.35-0.46

General abdominal conditions No

1.00

1.00

Intestinal obstruction Yes

16.91

15.88-18.01

4.49

4.00-5.04

Intestinal obstruction No

1.00

1.00

Upper gastrointestinal tract Yes

7.77

7.32-8.25

2.82

2.53-3.15

Upper gastrointestinal tract No

1.00

1.00

Hepatic-pancreatic-biliary Yes

3.46

3.29-3.65

1.92

1.72-2.15

Hepatic-pancreatic-biliary No

1.00

1.00

Colorectal Yes

0.85

0.83-0.88

0.47

0.42-0.53

Colorectal No

1.00

1.00

Hernias Yes

1.04

0.99-1.09

0.56

0.50-0.63

Hernias No

1.00

1.00

Soft tissue Yes

0.37

0.35-0.39

0.31

0.27-0.35

Soft tissue No

1.00

1.00

Vascular Yes

3.97

3.74-4.22

1.77

1.58-1.98

Vascular No

1.00

1.00

Cardiothoracic Yes

16.2

14.9-17.6

8.47

7.44-9.63

Cardiothoracic No

1.00

1.00

Day of Admission

Admitted Saturday-Sunday

1.08

1.07-1.09

1.03

1.01-1.04

Admitted Monday-Friday

1.00

1.00

Total number of visits

1.000

1.000-1.000

1.000

1.000-1.000

Trauma Level

>=III

4.96

4.03-6.11

1.81

1.38-2.37

I or II

1.00

1.00

Hospital Region

Midwest

3.30

2.68-4.06

2.06

1.56-2.72

South

1.33

1.13-1.57

1.08

0.79-1.46

West

1.76

1.47-2.12

1.27

0.97-1.66

Northeast

Hospital Ownership Government, nonfederal

1.00

3.18

2.73-3.72

1.00

1.22

0.99-1.52

Private, invest-own

1.29

1.1-1.52

0.84

0.65-1.10

Private, not-profit

1.45

1.25-1.69

1.00

0.77-1.30

Private, collapsed category

5.87

5.02-6.88

1.09

0.90-1.31

Government/private, collapsed

Hospital Location/Teaching Status Rural

1.00

3.43

2.92-4.02

1.00

1.69

1.43-2.00

Urban nonteaching

1.00

1.00

Urban teaching

0.38

0.31-0.48

0.75

0.56-1.01

discomfort operating on medically complex patients and the absence of critical care resources in small and rural hospitals [2]. Protocols and training tailored to help rural community hospitals care for medically complex patients who do not need specialty surgical intervention have the potential to improve resource use and patient outcomes. In ad- dition there may be a role for telemedicine consultation both in the ED

and in-hospital to prevent an unnecessary transfer and provide addi- tional management guidance.

Despite accepting facilities having more resources, transfers fre- quently confer worse outcomes and higher costs even among propen- sity score-matched cohorts [20,29-31]. Given the complexity of EGS patients who are transferred from EDs, they stand to benefit both

from standardized protocols to streamline transfer communication and procedures, and also from stable transfer relationships with specific ter- tiary care centers. Existing research demonstrates that smaller hospitals typically refer to many different facilities [32]. This is problematic be- cause ED providers waste time calling different hospitals for transfer. Moreover, there is substantial nationwide variability in (1) whether transfers occur for EGS patients, (2) the lengths of stay for patients who are ultimately transferred, and (3) interhospital communication to facilitate transfers [21,33-36]. Gaps in communication are associated with higher mortality [35]. Therefore, a standardized system for assessing patients and initiating interhospital transfers for EGS patients has the potential to reduce unwarranted variability in this process and improve outcomes.

Because EDs rely on timely care processes and are often the first place patients present with EGS conditions, they are a natural fit for de- veloping and testing protocols to improve and standardize the EGS transfer process. Nacht and colleagues comprehensively summarized the need to examine populations that are predisposed to transfer from EDs as essential to characterizing “three critical aspects of emergency care systems: (1) regionalization, including prehospital destination pro- tocols and transferring patients for Specialized care; (2) resource utiliza- tion and distribution, including protocols for diagnostic tests and interventions, workforce planning, and opportunities for telemedicine; and (3) planning for surge capacity.” [37] The authors further note that ED beds are a limited resource, with throughput being of prime im- portance to regulators yet 66% of transferred patients stayed in the ED for more than 3 hours. However, interventions to facilitate interhospital transfers are scant. A single study of a one-page handoff communication tool yielded reductions in length of stay and mortality for all transferred patients, and the majority of efforts to improve handoffs from the ED have focused on communication with outpatient providers and estab- lishing follow-up rather than transfers to other facilities [38,39]. Re- cently, a published opinion on EGS transfers recommended a three part assessment of whether transfer is indicated based on (1) underlying disease (2) the severity of physiologic derangement, and (3) available Hospital resources. Intervention development to improve and standard- ize interhospital transfers for EGS patients has significant potential to reduce mortality, reduce morbidity, and promote the efficient use of re- sources [40]. The resulting intervention at the point of care could in- volve standardized handoff checklists, comprehensive telephone scripts, or a combination thereof to support successful transfers in which patients are transferred swiftly and benefit unequivocally from a higher level of care, including specialty surgical care.

Our study has some important limitations that contextualize our re-

sults. First, NEDS does not allow us to characterize the post-transfer hos- pitalization. As a result, we cannot qualify the outcomes of EGS patients who are transferred from EDs, including morbidity, mortality, and cost. Because NEDS is a population-level data set and is not validated to ad- dress specific clinical problems, our results warrant additional valida- tion in the clinical setting. Moreover, we cannot identify how many different hospitals each facility transfers to nor can we characterize the clinical services available and surgical capacity at transferring facil- ities. Our data are from 2014 and may represent practices that have re- cently shifted. However, because a formalized ICD-10 definition of EGS has not yet been established, 2014 is the most recent year for which we can characterize EGS using a consensus definition. Overall, our anal- ysis provides a comprehensive, representative picture of the factors that predispose patients with an EGS condition to transfer.

Overall, we have shown that the burden of transfer disproportion- ately falls on medically complex and older patients who present at small, rural hospitals. Efforts to improve the interhospital transfer of ED patients with EGS conditions are overdue and would answer a longstanding call to improve handoffs from the ED more generally [41]. Moreover, smaller emergency departments are closing, and more health systems are operating free-standing EDs [42]. Therefore, it is worthwhile for EDs and referring hospitals to prepare joint protocols

for transfers. At a minimum, there is evidence to support pre-arranged transfer relationships between small, community hospitals and larger referral centers, standard communication templates to facilitate inter- hospital communication, and expedited transfer for specific, priority populations. 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.

Funding

Agency for Healthcare Research and Quality K08HS025224-01A1.

Declaration of Competing Interest

AI: Clinical Consultant American College of Surgeons Agency for Healthcare Research and Quality Safety Program for Improving Surgical Care and Recovery.

Angela Ingraham: Clinical Consultant American College of Surgeons Agency for Healthcare Research and Quality Safety Program for Improv- ing Surgical Care and Recovery.

References

  1. Shafi S, Aboutanos MB, Agarwal Jr S, Brown CV, Crandall M, Feliciano DV, et al. Emer- gency general surgery: definition and estimated burden of disease. J Trauma Acute Care Surg. 2013 Apr 1;74(4):1092-7.
  2. Reinke CE, Thomason M, Paton L, Schiffern L, Rozario N, Matthews BD. Emergency general surgery transfers in the United States: a 10-year analysis. J Surg Res. 2017 Nov 1;219:128-35.
  3. Gale SC, Shafi S, Dombrovskiy VY, Arumugam D, Crystal JS. The public health burden of emergency general surgery in the United States: a 10-year analysis of the Nation- wide inpatient sample–2001 to 2010. J Trauma Acute Care Surg. 2014 Aug 1;77(2): 202-8.
  4. Santry H, Kao LS, Shafi S, Lottenberg L, Crandall M. Pro-con debate on regionalization of emergency general surgery: controversy or common sense? Trauma Surg Acute Care Open. 2019 May 1;4(1):e000319.
  5. Voelker R. Experts say projected surgeon shortage a “looming crisis” for patient care.

Jama. 2009 Oct 14;302(14):1520-1.

  1. Philip JL, Saucke MC, Schumacher JR, Fernandes-Taylor S, Havlena J, Greenberg CC, et al. Characteristics and timing of Interhospital transfers of emergency general sur- gery patients. J Surg Res. 2019 Jan 1;233:8-19.
  2. Huntington CR, Cox TC, Blair LJ, Prasad T, Lincourt AE, Matthews BD, et al. Acuity, outcomes, and trends in the transfer of surgical patients: a national study. Surg Endosc. 2016 Apr 1;30(4):1301-9.
  3. Broman KK, Ward MJ, Poulose BK, Schwarze ML. Surgical transfer decision making: how regional resources are allocated in a regional transfer network. Jt Comm J Qual Patient Saf. 2018 Jan 1;44(1):33-42.
  4. Scott JW, Tsai TC, Neiman PU, Jurkovich GJ, Utter GH, Haider AH, et al. Lower emer- gency general surgery (EGS) mortality among hospitals with higher-quality trauma care. J Trauma Acute Care Surg. 2018 Mar 1;84(3):433-40.
  5. Ogola GO, Haider A, Shafi S. Hospitals with higher volumes of emergency general surgery patients achieve lower mortality rates: a case for establishing designated centers for emergency general surgery. J Trauma Acute Care Surg. 2017 Mar 1;82 (3):497-504.
  6. Aguirre FV, Varghese JJ, Kelley MP, Lam W, Lucore CL, Gill JB, et al. Stat heart inves- tigators. Rural interhospital transfer of ST-elevation myocardial infarction patients for percutaneous coronary revascularization: the stat heart program. Circulation. 2008;117:1145-52.
  7. West JG, Trunkey DD, Lim RC. Systems of trauma care: a study of two counties. Arch Surg. 1979 Apr 1;114(4):455-60.
  8. MacKenzie EJ, Rivara FP, Jurkovich GJ, Nathens AB, Frey KP, Egleston BL, et al. A na- tional evaluation of the effect of trauma-center care on mortality. N Engl J Med. 2006 Jan 26;354(4):366-78.
  9. HCUP Nationwide Emergency Department Sample (NEDS). Healthcare Cost and Uti- lization Project (HCUP). 2020. Agency for Healthcare Research and Quality, Rock- ville, MD. www.hcup-us.ahrq.gov/nedsoverview.jsp
  10. Shah AA, Haider AH, Zogg CK, Schwartz DA, Haut ER, Zafar SN, et al. National esti- mates of predictors of outcomes for emergency general surgery. J Trauma Acute Care Surg. 2015 Mar 1;78(3):482-91.
  11. Ingraham A, Wang X, Havlena J, Hanlon B, Saucke M, Schumacher J, et al. Factors as- sociated with the interhospital transfer of emergency general surgery patients. J Surg Res. 2019 Aug 1;240:191-200.
  12. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prog- nostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373-83.
  13. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992;45(6):613-9.
  14. Lim HJ, Zhang X, Dyck R, Osgood N. Methods of competing risks analysis of end- stage renal disease and mortality among people with diabetes. BMC Med Res Methodol. 2010 Dec 1;10(1):97.
  15. Castillo-Angeles M, Uribe-Leitz T, Jarman M, Jin G, Feeney T, Salim A, et al. Trans- ferred emergency general surgery patients are at increased risk of death: a NSQIP propensity score matched analysis. J Am Coll Surg. 2019 Jun 1;228(6):871-7.
  16. Schnipper JL, Mueller SK, Orav EJ. Rates, predictors and variability of interhospital transfers: a national evaluation. J Hosp Med. 2017 Jun;1:12(6).
  17. Mohr NM, Wu C, Ward MJ, McNaughton CD, Richardson K, Kaboli PJ. Potentially avoidable inter-facility transfer from Veterans Health Administration emergency de- partments: a cohort study. BMC Health Serv Res. 2020 Dec 1;20(1):110.
  18. Reimer AP, Schiltz NK, Ho VP, Madigan EA, Koroukian SM. Applying supervised ma- chine learning to identify which patient characteristics identify the highest rates of mortality post-interhospital transfer. Biomed Inform Insights. 2019 Mar;11 1178222619835548.
  19. Broman KK, Hayes RM, Kripalani S, Vasilevskis EE, Phillips SE, Ehrenfeld JM, et al. In- terhospital transfer for acute surgical care: does delay matter? Am J Surg. 2016 Nov 1;212(5):823-30.
  20. Kessler C, Williams MC, Moustoukas JN, Pappas C. transitions of care for the geriatric patient in the emergency department. Clin Geriatr Med. 2013 Feb 1;29(1):49-69.
  21. Kindermann DR, Mutter RL, Cartwright-Smith L, Rosenbaum S, Pines JM. Admit or transfer? The role of insurance in high-transfer-rate medical conditions in the emer- gency department. Ann Emerg Med. 2014 May 1;63(5):561-71.
  22. Huang Y, Natale JE, Kissee JL, Dayal P, Rosenthal JL, Marcin JP. The association be- tween insurance and transfer of noninjured children from emergency departments. Ann Emerg Med. 2017 Jan 1;69(1):108-16.
  23. Hernandez-Boussard T, Davies S, McDonald K, Wang NE. Interhospital facility trans- fers in the United States: a nationwide outcomes study. J Patient Saf. 2017 Dec;13 (4):187.
  24. Yelverton S, Rozario N, Matthews BD, Reinke CE. Interhospital transfer for emer- gency general surgery: an independent predictor of mortality. Am J Surg. 2018 Oct 1;216(4):787-92.
  25. Sokol-Hessner L, White AA, Davis KF, Herzig SJ, Hohmann SF. Interhospital transfer patients discharged by academic hospitalists and general internists: characteristics and outcomes. J Hosp Med. 2016 Apr;11(4):245-50.
  26. Mohr NM, Harland KK, Shane DM, Ahmed A, Fuller BM, Torner JC. Inter-hospital transfer is associated with increased mortality and costs in severe sepsis and septic shock: an instrumental variables approach. J Crit Care. 2016 Dec 1;36:187-94.
  27. Kindermann DR, Mutter RL, Houchens RL, Barrett ML, Pines JM. The transfer instabil- ity index: a novel metric of emergency department transfer relationships. Acad Emerg Med. 2015 Feb;22(2):166-71.
  28. Lauerman MH, Herrera AV, Albrecht JS, Chen HH, Bruns BR, Tesoriero RB, et al. Inter- hospital transfers with wide variability in emergency general surgery. Am Surg. 2019 Jun 1;85(6):595-600.
  29. Herrigel DJ, Carroll M, Fanning C, Steinberg MB, Parikh A, Usher M. Interhospital transfer handoff practices among US Tertiary care centers: a descriptive survey. J Hosp Med. 2016 Jun;11(6):413-7.
  30. Usher M, Sahni N, Herrigel D, Simon G, Melton GB, Joseph A, et al. Diagnostic discor- dance, health information exchange, and inter-hospital transfer outcomes: a popula- tion study. J Gen Intern Med. 2018 Sep 1;33(9):1447-53.
  31. Harl FN, Saucke MC, Greenberg CC, Ingraham AM. Assessing written communication during interhospital transfers of emergency general surgery patients. J Surg Res. 2017 Jun 15;214:86-92.
  32. Nacht J, Macht M, Ginde AA. Interhospital transfers from US emergency depart- ments: implications for resource utilization, patient safety, and regionalization. Acad Emerg Med. 2013 Sep;20(9):888-93.
  33. Theobald CN, Choma NN, Ehrenfeld JM, Russ S, Kripalani S. Effect of a handover tool on efficiency of care and mortality for Interhospital transfers. J Hosp Med. 2017 Jan 1;12(1):23-8.
  34. Katz EB, Carrier ER, Umscheid CA, Pines JM. Comparative effectiveness of care coor- dination interventions in the emergency department: a systematic review. Ann Emerg Med. 2012 Jul 1;60(1):12-23.
  35. McCrum ML, Davis KA, Kaafarani H, Santry H, Shafi S, Crandall M. Current opinion on emergency general surgery transfer and triage criteria. J Trauma Acute Care Surg. 2020 May 26;89(3):e71-7.
  36. Dhingra KR, Elms A, Hobgood C. Reducing error in the emergency department: a call for standardization of the sign-out process. Ann Emerg Med. 2010 Dec 1;56(6): 637-42.
  37. Patidar N, Weech-Maldonado R, O’Connor SJ, Sen B, Camargo Jr CA. Contextual fac- tors associated with hospitals’ decision to operate freestanding emergency depart- ments. Health Care Manag Rev. 2017 Jul 1;42(3):269-79.