Introduction
Trauma is a major public health concern and leading cause of death in the United States for those aged one to 44 years old.1 Los Angeles (LA) County (California USA) has not been immune to this problem with an overall trauma center (TC) volume in 2008 of approximately 19,000, which increased to over 23,000 in 2017 across 15 TCs.2 For patients with critical injury, it has been well-established in current literature that treatment at a designated TC improves overall outcomes.Reference Haas, Gomez and Zagorski3-Reference Mullins, Veum-Stone and Hedges6 While Emergency Medical Services (EMS) systems have been designed to identify trauma patients for direct routing to TCs, patients with traumatic injuries may also present to a non-TC via private vehicle or due to EMS under-triage. In these cases, secondary transportation to a TC is required to ensure optimal patient outcomes.
Typically, inter-facility transfers (IFTs) are coordinated between individual hospitals, often involving a transfer center, and require identification of an accepting physician at the receiving center. Private ambulance transport is then arranged by the sending facility. In trauma systems, delays in IFTs often stem from difficulties in identification of a receiving facility, preparation of documentation, successful completion of physician-to-physician communication, and/or arrangement of timely ambulance transportation.Reference Harrington, Connolly and Biffl7-Reference Crandall, Esposito and Reed10 While this traditional process of IFT may be effective for stable patients, delays in IFTs cause harm to patients experiencing time-critical emergencies.Reference Harrington, Connolly and Biffl7 In particular, for critical trauma patients at local community hospitals who require expeditious transport to definitive care for emergent interventions, this process is inappropriately lengthy. Use of an established 911 system offers a rapid alternative for effective IFTs for these trauma patients.Reference Kuncir, Spencer and Feldman8,Reference Eckstein, Schlesinger and Sanko11
In 2015, a new community hospital opened in LA County in an area with a high trauma burden, where frequent transports by private vehicle were anticipated. In response, the LA County EMS system developed a pilot program to activate the established 911 system for rapid transport of these patients to the closest TC. The primary objective of this study was to describe the effectiveness of 911 trauma re-triage with regard to timeliness of IFTs. A secondary objective was to assess protocol adherence by the sending hospital.
Methods
This was a retrospective cohort study utilizing data from the regional trauma registry and individual patient care records. The study was reviewed and approved by the Institutional Review Board at the Lundquist Research Institute (Torrance, California USA; protocol ID # 042614).
The LA County EMS Agency oversees a regional trauma system serving over 10 million residents with 30 fire-based public provider EMS agencies and one law enforcement agency transporting patients to 14 designated TCs and covering LA County’s 4,085 square miles. In 2015, a new 131-bed community hospital was opened in LA serving an urban/suburban region of approximately 1.3 million predominantly black and Hispanic residents. Four years after opening, the emergency department (ED) volume at this community hospital was estimated at approximately 90,000 patients per year. The nearest designated TC was located just under three miles away. Historically, inter-facility trauma transfers in LA County were coordinated by a centralized transfer center and required identification of an accepting physician at the receiving TC. Private ambulance transport was then arranged by the sending facility. Given the high trauma burden in the selected area and anticipation of injured patients arriving by private vehicle, the LA County EMS Agency implemented a 911 trauma re-triage policy to allow for rapid transport of trauma patients from the community hospital to the nearest TC with unconditional acceptance.
The 911 re-triage protocol was developed based on recommendations from a state EMS workgroup along with input from local experts at the LA County EMS Agency and regional TCs.12 Per protocol, IFTs to the TC were conducted via the public fire-based EMS provider who was the sole 911 provider serving this region. Prior notification of the transfer was provided, as able; however, acceptance of the transfer was based on prior agreement between the centers and did not require direct physician-to-physician communication. In advance of implementation, an educational module, along with the criteria and implementation steps, were distributed to ED providers and staff at the community hospital.
Patients included were 15 years or older who sustained a traumatic injury and were transported via 911 re-triage from the community hospital to the nearest designated TC. Patients transported by private ambulance for routine IFTs and those transported to other TCs were excluded. Data were routinely submitted by TCs in LA County to the EMS Agency on all trauma victims with at least one ICD-9 injury diagnostic code within the range of 800-959.9 or ICD-10 S00-S99 or T79.A1-T71.A9 and maintained in LA County’s Trauma and Emergency Medical Information System (TEMIS). Data received from all trauma centers are verified by an epidemiologist based at the LA County EMS Agency for completeness, logical consistency, duplication, and formatting. Data completeness is maintained at >90% for all fields. Identified deficiencies and errors are sent back to the TC for correction, and updated information is automatically uploaded to TEMIS every 24 hours. Quarterly reports are generated and disseminated to the system for use in quality improvement. This study was a retrospective analysis of patient data contained in the TEMIS.
Data were downloaded to an Excel file (Microsoft Corporation; Redmond, Washington USA) from TEMIS from August 1, 2015 through May 31, 2016 by a registered nurse working within her normal job duties, blinded to the study hypothesis and outcomes. Variables abstracted included age, sex, race/ethnicity; mechanism of injury; injury severity score (ISS); EMS times (response time, scene time, and transport time); TC treatments; disposition from the ED (admission, and to what level of care, or discharge); and patient outcome (lived/died) at discharge from the TC. All variables were defined based on the TEMIS data dictionary available to the investigators.13 Time values were recorded by prehospital providers via the Computer-Aided Dispatch system and automatically transmitted to the electronic patient care record for upload into TEMIS. Total transport time was defined as EMS arrival at the transferring hospital to arrival at the TC (ie, inclusive of scene time and transport time). Total transfer time was defined as EMS response to arrival at the TC (ie, inclusive of EMS response time, scene time, and transport time).
Rationale for transport was determined from review of the corresponding prehospital care report for each patient, including the narrative summary of events. Two investigators, board certified in Emergency Medicine and EMS and trained on the 911 re-triage policy, independently reviewed the prehospital patient care report to determine: (1) whether the patient met 911 re-triage criteria by policy (Figure 1); and (2) whether the investigator agreed with the decision for 911 trauma re-triage. Investigators were blinded to the treatment and outcome at the TC. In the case of disagreement, a third senior EMS physician investigator, also trained on the 911 re-triage policy, reviewed the case to determine the final decision.
Figure 1. 911 Trauma Re-Triage.
The primary outcome was median total transport time. Secondary outcomes were the proportion of patients meeting re-triage criteria and whether investigators agreed with the decision for 911 trauma re-triage. Each case was also reviewed to determine whether the patient would have met field trauma triage criteria for direct transport to a TC, based on LA County trauma triage policy.12
Data were maintained in Microsoft Excel and uploaded into SAS 9.4 (SAS Institute; Cary, North Carolina USA) for statistical analysis. Descriptive statistics were calculated with median and inter-quartile range (IQR) or frequencies and proportions, as appropriate.
Results
During the study period, 32 patients with traumatic injuries were transferred via 911 re-triage to the nearest TC. Patient characteristics are shown in Table 1. Twenty-five (78%) were male; 39% black, 45% Hispanic, 13% other or undocumented race/ethnicity, and 3% white. The median age was 31 years old (IQR 24-45 years). Twenty-one (66%) had a penetrating mechanism of injury. Median injury severity score (ISS) was four (IQR 1-10).
Table 1. Patient Characteristics and Transfer Times (n = 32)
Abbreviation: EMS, Emergency Medical Services.
a Total transport time includes scene time and transport time.
b Total transfer time includes response time, scene time, and transport time.
Overall response and transport times were short. The median 911 prehospital provider response time was four minutes (IQR 2-5 minutes), median prehospital provider scene time was 10 minutes (IQR 8-12 minutes), and median transport time was seven minutes (IQR 6-9 minutes). The median total transfer time via 911 was 21 minutes (IQR 18-25 minutes; Table 1).
Seventeen patients (53%) met 911 re-triage criteria as determined by study investigators (Table 2). Agreement between investigators was 97% (31 of 32 cases). Investigators concurred with the decision to utilize 911 re-triage for the same 17 patients; investigator agreement 91% (29 of 32 cases). In cases of disagreement, final classification was determined by a third investigator. The most common criteria met was “penetrating injury to the head, neck, or torso” in 14 cases. Two patients, by investigator judgment, met the criteria “high likelihood of requiring emergent life- or limb-saving intervention within two hours;” both had sustained a gunshot wound to the buttock. “Intubation required” and “extremity injury with neurovascular compromise” were present in one patient each (Table 2 and Table 3). Twenty-seven patients (84%) would have met EMS trauma triage criteria for primary transport to a TC as delineated by LA County trauma triage policy.12
Table 2. Criteria/Rationale for Transfer and Disposition Outcome by Individual Patient
Abbreviations: CT, computed tomography; ED, emergency department; GCS, Glasgow Coma Scale; ICU, intensive care unit; OR, operating room.
Table 3. Frequency of Each Re-Triage Criteria
Note: Patients could meet more than one criterion.
Overall, 22 (69%) patients were admitted to the TC, including seven to the operating room and four to the intensive care unit. Three patients required packed red blood cell transfusions. All patients survived to hospital discharge.
Discussion
This study demonstrated that implementation of a 911 trauma re-triage protocol for patients with critical injuries initially presenting to a non-TC was a feasible approach for timely transfer to definitive care at a designated TC. Furthermore, after a brief initial training, local community hospital providers were able to effectively utilize this 911 re-triage protocol to appropriately transfer trauma patients in more than 50% of the cases. Through rapid transfer to definitive care, this may improve outcomes for select patients needing time-sensitive interventions not available within the community hospital setting.
To date, only two prior studies have assessed 911 re-triage for critically injured trauma patients; both took place within urban/suburban centers in California. Kuncir, et al conducted a retrospective study assessing 911 re-triage in Orange County, California from 2014 through 2015, where a 911 trauma re-triage policy has been in effect to move all trauma patients to the TC since the mid-2000s. They noted a median total transfer time in that system of 122.5 minutes (defined as time of patient arrival at the transferring hospital to time of arrival to the TC).Reference Kuncir, Spencer and Feldman8 This differs from the current study because the time interval was assessed starting from time of patient arrival at the non-TC, which allowed inclusion of decision time prior to initiation of patient transfer, and patients mostly suffered from blunt trauma. Della Valle, et al undertook a retrospective observational study in the San Francisco Bay Area from 2013 through 2015, comparing 911 re-triage to conventional IFT. The authors found a decrease of 81 minutes in median total transfer time (defined as time of patient arrival at the transferring hospital to time of arrival at the TC) between 911 re-triage and conventional IFT (172 minutes versus 258 minutes, respectively).Reference Della Valle, Newton and Kline9 Patients with penetrating trauma had a much shorter median transfer time via 911, 96 minutes, in comparison to blunt trauma via 911 at 198 minutes, suggesting the decision time may be longer for patients with a blunt traumatic mechanism.
In the current study, the median total transfer time from initiation of the 911 response at the transferring hospital to arrival at the TC was 21 minutes. By defining the total transfer time as beginning with EMS activation, this better quantifies the burden on the 911 system and removes the variability of the decision-making process at the transferring center in regards to activating the transfer process. Unfortunately, the differing definitions of total transport/transfer times between the aforementioned studies and this investigation prevent direct comparison of time values. Yet overall, these reports collectively demonstrate the feasibility of 911 re-triage systems as a means to expeditiously move trauma patients to definitive care within a developed regional trauma system. In comparison, transfer times for trauma patients utilizing the standard IFT process range from two hours to greater than four hours in some studies, though varying definitions of transport/transfer time were used.Reference Harrington, Connolly and Biffl7,Reference Della Valle, Newton and Kline9,Reference Crandall, Esposito and Reed10,Reference Porter, Wyrick and Bowman14 It is well-described in current literature that timely treatment at a TC, ideally via direct transport to a TC or secondarily via rapid transfer to a TC from a non-TC, results in reduced morbidity and mortality for trauma patients.Reference Haas, Gomez and Zagorski3,Reference Haas, Stukel and Gomez4,Reference Harrington, Connolly and Biffl7,Reference Sampalis, Denis and Lavoie15,Reference Holst, Perman and Capp16 Though 911 re-triage is not a replacement for direct routing, it can minimize delays to definitive care for critically injured patients who arrive at a non-TC, with a median total transfer time from EMS notification to arrival at the TC of 21 minutes in this cohort. Furthermore, Kuncir, et al demonstrated similar mortality rates for transfer patients utilizing 911 trauma re-triage compared with direct transport of trauma patients within their system.
This study adds to the prior literature demonstrating both the effectiveness (in regards to transfer time) and the appropriateness of 911 use (in regards to protocol adherence) after implementation of a 911 trauma re-triage protocol. Della Valle, et al evaluated a similar set of criteria in the San Francisco Bay area, however, the authors did not evaluate protocol adherence as an outcome.Reference Della Valle, Newton and Kline9 In this study, all injuries identified at the non-TC necessitating 911 trauma re-triage according to protocol were due to penetrating trauma, most often to the head, neck, or torso. Of those transferred that did not satisfy 911 re-triage criteria, the majority suffered from blunt trauma or penetrating trauma isolated to an extremity. All had stable vital signs documented by the 911 provider.
In this study, investigators agreed that 911 transport was not appropriate in all cases that did not meet the 911 re-triage criteria, thus supporting the existing re-triage criteria. Although not all trauma patients will require 911 re-triage, transfer may still be appropriate for some trauma patients via standard transfer procedures as evidenced by the rate of admission in this cohort (69%). Patients may also benefit from subspecialty consultation not available at the non-TC. However, these transfers should be conducted on a non-emergent basis, thus reducing the burden on the 911 system.
Critics of 911 re-triage practices cite concerns over limited prehospital resources and risk of exceeding paramedic scope of practice, as well as the potential for over-use. As evidenced by this study and others, these transfers can be conducted efficiently with time intervals consistent with published average prehospital trauma response times.Reference Eckstein, Schlesinger and Sanko11,Reference Carr, Caplan and Pryor17 In a study by Eckstein, et al, IFTs made up 0.1% of all ambulance transports.Reference Eckstein, Schlesinger and Sanko11 Additionally, in cases where interventions or treatments performed at the referring hospital require continued management that exceeds paramedic scope of practice (eg, infusion of blood products or monitoring of a chest tube), it may be feasible in some instances for hospital personnel to accompany paramedics. Alternatively, where appropriate, the specific interventions may be temporarily discontinued for transport.Reference Eckstein, Schlesinger and Sanko11 In the current study, no patients required any additional personnel from the referring hospital to accompany the paramedics due to interventions exceeding paramedic scope of practice. In the case of the intubated patient, paramedics were able to manually ventilate during the transport.
Finally, concerns of over-use must be weighed against the risk of under-triage. Despite evidence of over-triage in this small pilot study, this degree of over-triage (25%-35%) is acceptable according to the American College of Surgeons Committee on Trauma (Chicago, Illinois USA) in order to prevent dangerous under-triage.Reference Cribari, Rontondo and Smith18 Of note, the vast majority of patients transferred in this study met field trauma triage criteria for direct transport to a TC within the same regional system.12 Additionally, given that this was a relatively novel protocol implemented in a new ED, it is conceivable that with increased protocol familiarity and quality improvement, apparent over-use of the 911 re-triage system would decrease over time. Future analyses are planned to evaluate the level of over-triage subsequent to adapting the 911 trauma re-triage policy system-wide.
Limitations
This study has several limitations. First, the retrospective nature of the study and lack of a comparison group limits what can be inferred from patient outcomes. This protocol was implemented upon the opening of the local community hospital, and as such, there were no prior data on trauma transfers via the standard process against which to compare. Although data from the EMS providers and TCs were available, it was not possible to obtain records from the community hospital; therefore, the rationale for transport is based on the injuries as documented by transporting EMS providers. The time of arrival at the community hospital is also unknown, so the total time to transfer could not be calculated from patient arrival. Although EMS activation times and transport times were relatively short, this study does not attempt to further quantify the burden on the 911 system, including delays in responding to primary 911 calls or need to bring resources in from surrounding areas. Finally, this study took place between one referring and one receiving institution in an urban region within a mature trauma system and significant prehospital resources. As such, study findings may not be generalizable to dissimilar settings including rural areas or regions with developing trauma systems.
Conclusions
This pilot study demonstrates the feasibility and effectiveness of 911 re-triage protocols for rapid transfer of critical trauma patients to a TC in an urban region with a mature trauma system, with a median transport time of 17 minutes. This study serves as a foundation for other regions to develop trauma-centered 911 re-triage protocols, as well as for future studies in the Los Angeles County to assess the on-going county-wide efficacy of this protocol.
Conflicts of interest/funding
No funding was received for this study. There are no conflicts of interest to disclose. These data were presented, in part, at the American College of Emergency Physicians Scientific Assembly 2017, Washington DC, USA.
