Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Parikh PP, Parikh P, Mamer L, McCarthy MC, Sakran JV. Association of System-Level Factors With Secondary Overtriage in Trauma Patients. JAMA Surg. 2019;154(1):19–25. doi:10.1001/jamasurg.2018.3209
Is regionalization of trauma care, including the distribution of trauma centers, associated with secondary overtriage?
This cohort study of statewide data shows a 12.2% rate of secondary overtriage. System-level factors, such as trauma center distribution and choice of destination by emergency medical services professionals, significantly increase the rate of secondary overtriage.
These results provide the building blocks for designing an optimal network of trauma centers and developing educational and outreach training programs for clinicians at level III and nontrauma centers regarding criteria for the transfer of an injured patient, the application of telemedicine, and the potential policy implications for transfers.
Studies show that secondary overtriage (SO) contributes significantly to the economic burden of injured patients; thus, the association of SO with use of the trauma system has been examined. However, the association of the underlying trauma system design with such overtriage has yet to be evaluated.
To evaluate whether the distribution of trauma centers in a statewide trauma system is associated with SO and to identify clinical and demographic factors that may lead to SO.
Design, Setting, and Participants
A retrospective cohort study was performed using 2008-2012 data from the Ohio Trauma and Emergency Medical Services registries. All patients taken to level III or nontrauma centers from the scene of the injury with an Injury Severity Score less than 15 and discharged alive were included. Among these patients, those with SO were identified as those who were subsequently transferred to a level I or II trauma center, had no surgical intervention, and were discharged alive within 48 hours of admission. The SO group was analyzed descriptively. Multiple logistic regression was used to identify system-level factors associated with SO. Statistical analysis was performed from August 1, 2017, to January 31, 2018.
Main Outcomes and Measures
The primary outcome was the occurrence of SO.
Of 34 494 trauma patients able to be matched in the 2 registries, 7881 (22.9%) met the inclusion criteria, of whom 965 (12.2%) had SO. The median age in the SO group was 40 years (interquartile range, 26-55 years), with 299 women and 666 men. After adjusting for age, sex, comorbidities, injury type, and insurance status, the study found that system-level factors (number of level I or II trauma centers in the region [>1]) were significantly associated with SO (adjusted odds ratio, 1.98; 95% CI, 1.64-2.38; P < .001; area under the curve, 0.89). The reasons for choice of destination by emergency medical services (specifically, choosing the closest facility: adjusted odds ratio, 1.65; 95% CI, 1.37-1.98; P < .001) and use of a field trauma triage protocol (adjusted odds ratio, 2.21; 95% CI, 1.70-2.87; P < .001), significantly increased the likelihood of SO.
Conclusions and Relevance
This study’s findings suggest that the distribution of major trauma centers in the region is significantly associated with SO. Subsequent investigation to identify the optimal number and distribution of trauma centers may therefore be critical. Specific outreach and collaboration of level III trauma centers and nontrauma centers with level I and II trauma centers, along with the use of telemedicine, may provide further guidance to level III trauma centers and nontrauma centers on when to transfer injured patients.
The Emergency Medical Treatment and Active Labor Act, passed in 1986, mandates that emergency departments evaluate and stabilize any injured patient. Once they are stabilized, patients may be transferred to a higher-level facility if the initial receiving hospital lacks the resources to provide definitive care.1-3Quiz Ref ID Although there might be medical reasons necessitating transfer, it has been shown that patients are often transferred to level I trauma centers for nonmedical reasons.4 Consequently, a proportion of patients are discharged home within 48 hours after transfer to a level I trauma center from another facility, which is referred to as secondary overtriage (SO).5,6 Secondary overtriage is an unnecessary interfacility transfer of minimally injured patients and presents a resource-sensitive challenge to a state’s trauma system and trauma centers. It delays definitive care and imposes financial burdens, with a mean cost of $5917 to $12 549 per patient.7,8 Moreover, SO shifts resources away from severely injured patients who truly require care at a trauma center.5 For viability of trauma centers, judicious allocation of trauma resources has become a priority; therefore, it is essential to reduce triage errors, including SO. Developing a strategy to minimize triage errors will require an evidence-based approach to evaluate factors associated with SO.
Previous studies suggest that clinical factors such as injury type and location, as well as patient-level factors, including sex, race/ethnicity, and insurance status, could lead to SO.4,9 However, to our knowledge, the association of system-level factors within a trauma system with such errors has not yet been evaluated. Because SO occurs in a hospital setting and not at the scene of the injury, the purpose of this study was to understand and assess the association of trauma system design, including location of trauma centers, with such unnecessary transfers of patients, leading to SO.
Quiz Ref IDIn the state of Ohio, trauma centers are designated as level I, II, or III per American College of Surgeons verification standards.10 Furthermore, Ohio is geographically divided into 8 prehospital emergency medical services regions (also known as homeland security regions) to oversee the delivery of adult and pediatric prehospital emergency medical services.11 Therefore, in our study we analyzed data based on these regions. This study was approved by the Wright State University Institutional Review Board, who waived patient consent, as only deidentified data were obtained.
The Ohio Department of Public Safety provided us with deidentified patient records probabilistically linked from the Emergency Medical Services Incident Reporting System and Trauma Registry for 2008-2012. The inclusion criteria for both the Emergency Medical Services Incident Reporting System and Ohio Trauma Registry are available on the Ohio emergency medical services website.12,13 We derived 34 494 such records with complete data for consideration in this study. These data included information on the level of hospital where the patient was transferred from the scene of the injury and the second hospital (if any). All trauma patients who were transported to a nontertiary trauma center (ie, level III trauma center) or nontrauma center (NTC) from the scene, had an Injury Severity Score less than 15, and were discharged alive were included in the study. A subgroup of these patients who were subsequently transferred to a tertiary trauma center (level I or II trauma center), did not require a surgical procedure at a level I or II trauma center, and were discharged alive within 48 hours of admission were included in the SO group, as described in the literature.5,6 The group of patients who were not transferred to a level I or II trauma center and had no surgical intervention at a level III trauma center or NTC were included in the nontransferred group.
Patient-specific factors comprised age (adults, 18-65 years; and elderly individuals, >65 years), sex, race/ethnicity, and insurance status. Clinical factors obtained from the admitting hospital comprised Glasgow Coma Scale score, systolic blood pressure, heart rate, respiratory rate, Injury Severity Score, type of injury (penetrating vs blunt), mechanism of injury, primary clinical diagnosis, preexisting conditions, and discharge disposition. Clinical diagnoses were grouped based on primary hospital diagnosis groups for adult trauma patients as defined by Ohio Emergency Medical Services.13 System-level factors comprised the homeland security regions (ie, regions 1-8) within the state, number of trauma centers in the region, and reason for choosing the destination facility.
Statistical analysis was performed from August 1, 2017, to January 31, 2018. Incidences of SO, along with the underlying primary clinical diagnosis, were descriptively analyzed for each of the 8 homeland security regions in Ohio. The demographics in both the SO and nontransferred groups were analyzed to identify factors that uniquely differentiated the groups. Continuous data were presented as median (interquartile range) and proportions were presented as percentages. We compared the SO and nontransferred groups across various factors using the Mann-Whitney test (for continuous data) and Fisher exact test (for proportions).
Multiple logistic regression was then developed to determine if system-level factors contributed to the occurrence of a SO after adjusting for demographic and clinical factors. A stepwise regression was used to identify key factors, which were then included in a multiple logistic regression. All P values were from 2-sided tests and results were deemed statistically significant at P < .05. We verified the model’s quality based on the area under the curve and derived odds ratios of system-level factors.
Quiz Ref IDOf the 34 494 total patients able to be matched in the 2 registries, 7881 met the inclusion criteria; 965 of these patients (12.2%) were transferred to level I or II trauma centers, leading to SO, and 6916 (87.8%) were discharged directly from level III trauma centers or NTCs. Mean SO by homeland security regions in the state of Ohio is presented in Table 1. As shown in the Figure, the primary clinical diagnoses of SO patients (at the first hospital) were other unspecified injuries (to head and body), concussion, and fracture of vertebral column (without spinal cord injuries) or face bones.
Table 2 shows the comparison of patient demographics and clinical factors of the SO and nontransferred groups.13 Overall, there was a higher proportion of men in the SO group than the nontransferred group (666 [69.0%] vs 2292 [33.1%]; P < .001) and patients in the SO group were younger than those in the nontransferred group (median age [interquartile range], 40 [26-55] vs 78 [62-86] years; P < .001). The median values (interquartile ranges) Injury Severity Score of the SO and nontransferred groups were comparable (4 [2-6] vs 4 [4-6]). Quiz Ref IDComparing system-level factors, the SO group (n = 965) and the nontransferred group (n = 6916) differed significantly in the reason for choice of destination hospital by emergency medical services (closest facility, 505 [52.3%] vs 2603 [37.6%]; P < .001; use of a field trauma triage protocol, 156 [16.2%] vs 619 [9.0%]; P < .001) and number of level I or II trauma centers in the region (699 of 965 [72.4%] vs 4565 of 6916 [66.0%]; P < .001) (Table 3).
Table 4 discusses the results of the multiple logistic regression analysis to identify patient-level, clinical, and system-level factors that were significantly associated with SO after adjusting for other factors. Quiz Ref IDPatient-level and clinical factors, such as young age, male sex, insurance type, primary clinical diagnosis, absence of preexisting conditions or comorbidities, and injury type were associated with high SO. The system factors, such as number of level I or II trauma centers in the region (>1) were significantly associated with SO (adjusted odds ratio, 1.98; 95% CI, 1.64-2.38; P < .001). The reason for choice of a destination facility by emergency medical services, specifically, choosing the closest facility (adjusted odds ratio, 1.65; 95% CI, 1.37-1.98; P < .001) and use of a field trauma triage protocol (adjusted odds ratio, 2.21; 95% CI, 1.70-2.87; P < .001), significantly increased the likelihood of SO. The area under the curve of the model was 0.89.
To our knowledge, this study is the first to assess if trauma system design, including distribution of trauma centers, is associated with SO. Secondary overtriage presents a resource-sensitive challenge to tertiary trauma centers and the overall trauma system in general. To improve use of resources, it is essential to reduce such errors; thus, many authors have studied SO and its association with the trauma system.4-8 Our work, however, suggests that trauma system design itself could contribute to such errors.
We observed that 12.2% (965 of 7881) of the patients available to us for 2008-2012 experienced SO in Ohio. Higher rates of SO are not uncommon and have been observed in previous studies, where a 26% to 38% rate of SO was reported for a single institution.5,6 Our study also confirms previous findings on patient-level factors and clinical factors affecting SO, including age, sex, insurance type, and injury type.4,9,14 We also observed that the most common diagnoses for SO were concussion and fracture of the vertebral column or facial bones, which agrees with previous findings.7,14,15 This finding might be owing to the fact that in Ohio, many level III trauma centers and NTCs lack specialties such as neurosurgery or plastic surgery. As interfacility transfers in Ohio are primarily driven by patients’ needs,16 the lack of appropriate specialties in level III trauma centers and NTCs could lead to the transfer of minimally injured patients to major trauma centers. Taking an institutional approach to establish clear guidelines for care of such patients, along with appropriate collaboration with the physicians at level III trauma centers and NTCs could minimize unnecessary transfers of these patients. For example, as evidenced by the data, patients with head injuries as the clinical diagnosis at the first hospital led to SO; patients with “other unspecified injuries” at the first hospital who experienced SO also had concussion, other head injuries, or superficial unspecified injuries or contusion as primary clinical diagnosis at the second hospital. We have observed in practice that patients with head injuries who are receiving anticoagulation therapy are often transferred to level I or II trauma centers even when they have no signs of intracranial hemorrhage. These low-risk patients are usually monitored in the level I or II trauma center overnight and discharged the following day. In such cases, clear guidelines for treating and monitoring these minimally injured patients by neurologists (when a neurosurgeon is not available) could be developed and care could be provided at the level III trauma center or NTC. We established such guidelines at our institution (a level I trauma center) that were implemented at our affiliated level III trauma center (with education of their clinicians). We observed a reduction in the transfer of these low-risk patients and higher satisfaction among patients and clinicians. Such an approach could be generalized to improve collaboration among level I and II trauma centers and level III trauma centers and NTCs in Ohio to reduce SO. The use of telemedicine could help alleviate this problem further and, consequently, reduce the burden of SO on the Ohio trauma system. The application of telemedicine, especially telepresence of a trauma surgeon, has been used previously in the management of trauma patients and those undergoing emergency surgery.17,18 Telemedicine has been associated with improved outcomes, reduced costs of trauma care, and improved access to care in rural areas.17-19 We, therefore, believe that collaboration of level III trauma centers and NTCs, especially in rural areas, with urban level I and II trauma centers in Ohio could lead to a better decision-making process regarding the transfer of minimally injured patients.
Beyond clinical and patient-level factors, we identified system-level factors that seem to be associated with SO. Emergency medical services’ choice for a destination hospital during field triage was a factor contributing to SO, especially when patients were triaged based on the closest facility and use of a field trauma triage protocol. This finding may indicate that emergency medical services personnel are following the triage-based protocol or guidelines available to them. We observed regional variations in the rates of SO owing to variations in the number of level I and II trauma centers, which was also confirmed in a multiple logistic regression model. Regions with a higher number of level I and II trauma centers experienced more than a 96% increase in rates of SO compared with the nontransferred group (corresponding to the adjusted odds ratio of 1.96). Although this finding may not be clinically intuitive, a variety of factors could explain it. As SO occurs at the hospital level and not at the scene, the lack of clear guidelines, training, and/or experience of triage physicians or providers at the level III trauma centers and NTCs could play a role. Appropriate triage and transfer of trauma patients by emergency department physicians has been shown to correlate with the volume of moderately to severely injured patients they treat.20 The fact that most of the SO cases (831 of 965 [86.1%]) in our study were transferred from NTC facilities supports this reasoning. When the transfer decision is made by the initial institution, similar to what happens in Ohio, other social, financial, and/or convenience factors could play a role in patient transfers as well. Medicolegal consequences of not transferring these patients by emergency department physicians at level III trauma centers and NTCs may also encourage such transfers. Thus, the development of specific guidelines for managing low-risk, minimally injured patients; establishing standard transfer policies; and providing appropriate training resources to the clinicians at level III trauma centers and NTCs are essential to minimizing unnecessary transfers of these patients.
Although the socioeconomic and legal factors could lead to SO, the design of a trauma system includes the accuracy of its triage decisions. It is important to identify the optimal number of level I and II trauma centers in the region to ensure accessibility for the regional population, but not contribute to SO. Clustering of trauma centers has been reported in many states,21 which may be owing to regional or state trauma regulations (such as in Ohio22) and other financial motives of the hospital systems. Although some clustering of level I and II trauma centers is inevitable in Ohio owing to rural regions, it appears to increase unnecessary transfers and consequently affects resource use of the system, providing an opportunity to further improve the distribution of trauma centers and, potentially, trauma regulations in the state.22
The study was limited by its retrospective design and a possibility that the data were unique for 1 state in the United States. The study did not consider the knowledge and experience of either emergency medical services personnel or emergency medicine physicians in triaging patients, and was limited to what was documented in the database and limited by the available population for study. However, since the study uses data from 8 different regions of the state, it may mitigate inherent biases.
Secondary overtriage, the transfer of minimally injured patients to level I or II trauma centers, poses a significant burden on a trauma system. Our findings suggest that interfacility transfers are affected by system-level factors, in particular, the presence of 2 or more trauma centers, in a geographical region. Optimizing the number and location of trauma centers based on incidence rates of trauma is important and must be explored further. Developing better guidelines for patient transfer, and better training and education outreach to the emergency department physicians, in conjunction with the use of telemedicine, may further reduce SO.
Accepted for Publication: June 4, 2018.
Corresponding Author: Priti P. Parikh, PhD, Department of Surgery, Wright State University, 128 E Apple St, Ste 7000, Dayton, OH 45409 (email@example.com).
Published Online: September 19, 2018. doi:10.1001/jamasurg.2018.3209
Author Contributions: Dr P. P. Parikh had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: P. P. Parikh, P. Parikh, McCarthy, Sakran.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: P. P. Parikh, P. Parikh, Mamer.
Critical revision of the manuscript for important intellectual content: P. P. Parikh, P. Parikh, McCarthy, Sakran.
Statistical analysis: P. Parikh, Mamer.
Obtained funding: P. P. Parikh, P. Parikh.
Administrative, technical, or material support: McCarthy.
Supervision: P. P. Parikh, McCarthy, Sakran.
Conflict of Interest Disclosures: None reported.
Funding/Support: This work was funded by research grant 669853 from the Ohio Department of Public Safety.
Role of the Funder/Sponsor: The funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Meeting Presentation: This article was presented at the 13th Annual Meeting of the Academic Surgical Congress; January 30, 2018; Jacksonville, Florida.
Additional Contributions: Jose Pestana, Department of Biomedical, Industrial and Human Factors Engineering, Wright State University, assisted with preliminary data extraction and analysis. He was not compensated for his contribution. We also thank the Ohio Department of Public Safety Division of Emergency Medical Services (EMS) personnel and the Ohio Trauma Committee for sharing with us the data and insights into the Ohio trauma system, the Ohio Trauma Registry, and the Ohio EMS Incidence Reporting System.
Create a personal account or sign in to: