Emergency Medical Services Response to Motor Vehicle Crashes in Rural Areas (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

7 CHAPTER TWO LITERATURE REVIEW (VMT) on public roads occurs in urban areas (1.7 VMT in trillions); only 40% of VMT occurred in rural areas (1.1 VMT in trillions) (FHWA 2002). Although more motor vehicle travel takes place in urban areas, rural areas account for more than half of the nation’s fatal crashes. In 2010, rural areas reported more than 16,000 (54%) fatal crashes that resulted in 18,026 fatalities. Urban areas experienced more than 13,600 fatal crashes (45%) and 14,546 fatalities (NHTSA 2012). The 2010 fatality rate per vehicle miles traveled was two-and-one-half times higher in rural areas (1.83 per 100 million VMT) than in urban areas (0.73 per 100 million VMT) (NHTSA 2012). Although crash rates in general have been declining over the past 10 years, the discrepancy between rural and urban rates remains an issue, as illustrated in Figure 2. FIGURE 2 Fatalities per 100 million VMT by year and location (Source: NHTSA 2012). Several factors appear to contribute to higher fatality rates in rural areas, including differences in travel speeds, use of seat belts, and proximity of emergency care. Rural areas with higher speed limits account for the most fatal crashes (NHTSA 2010). Nearly 70% of fatal crashes in rural areas occur on roads with speed limits greater than 55 miles per hour (mph) whereas most fatal urban crashes, however, occur on roadways with speed limits of 50 mph or less (NHTSA 2010). Rural roadways may also lack the safety features of urban roadways, such as wide shoulders, lighting, and guard- rail/curb systems. Table 1 illustrates the relationship between speed limit and fatal crashes in urban and rural areas. Seat belt use and vehicle type also contribute to the dis- proportionately higher fatality rates in rural areas. In 2009, INTRODUCTION Improvement of EMS performance is a subject of ongoing study across the United States and around the world. The objective of this task is to conduct a national search of avail- able literature and synthesize the information relevant to this topic. This literature review provides background statistics as well as specific practices and metrics used by EMS per- sonnel and health care facilities. Literature was reviewed from various sources, located through online searches, publication databases such as PubMed, existing study bibliographic information, and rec- ommendations by the project panel. In general, documents were limited to the previous 15 years to minimize the influ- ence of out-of-date information. Several categories of docu- ments were obtained directly from state government online sources, including: • Toward Zero Deaths plans/Strategic Highway Safety Plans • Interoperable Communications Plans • State reports on vehicle crashes (Crash Facts and simi- lar documents). The U.S.DOT’s Research and Innovative Technology Administration database of deployment statistics was used as a source of information for deployments of transportation-related emergency response systems. This information was used to guide searches for documents from other sources and to review the relevant activities in the states included in this synthesis. Documents were compiled by SRF Consulting Group staff and categorized for review. Staff members were then assigned several categories for a detailed evaluation and summary. Information in this chapter includes a summary of crash statistics on rural areas, a summary of EMS crash response data, and a synthesis of literature related to EMS crash response in rural areas. RURAL CRASH STATISTICS The majority of motor vehicle travel in the United States takes place in urban areas. Despite a greater number of rural road miles, more than 60% of total vehicle-miles traveled

8 55% of persons killed in rural crashes were unrestrained compared with 50% of those killed in urban crashes. Fur- thermore, nearly 70% of rural pick-up truck occupants killed in crashes were not wearing seat belts, making it the highest percentage of any passenger vehicle occupants killed among both rural and urban areas (NHTSA 2009). RURAL CRASH RESPONSE STATISTICS Rural EMS is provided through a variety of service delivery components and methods across the nation. A network of EMS personnel, including volunteer and career emergency medical technicians (EMTs) and paramedics, use various vehicles, equipment, and facilities to deliver emergency medi- cal care to injured occupants of rural crashes (Knott 2003). After a severe motor vehicle crash, the crash occupant’s survival may ultimately depend on how quickly they receive definitive medical treatment. Dr. R. Adams Cow- ley is credited with coining the term “golden hour” to refer to the 60 minutes immediately following the occurrence of multisystem trauma event. However, a rigidly defined 60-minute interval for survival has since been scrutinized and the relevance of this timeframe is not supported by research. The time-dependency of a successful outcome is dependent on various factors, including the type of injury that has been sustained. Additionally, the literature is not clear on how time sensitive the delivery of definitive medi- cal care is on patient outcomes. For example, research has not established that a clinically significant outcome is cor- related with a given reduction in time in delivering defini- tive medical care. A related issue is the significance of the time required for an EMS unit to arrive at the scene (often measured in minutes) as compared with the significance of the time required for the patient to arrive at a facility that can provide definitive medical care (often measured in tens of minutes or hours). While a “hard” limit of 60 minutes is not regarded as crucial in decreasing patient mortality/morbidity rates, and the time sensitivity that correlates to clinically significant outcomes is not understood, there is general consensus that reducing the time from the occurrence of a motor vehicle crash to the delivery of the patient to definitive medical care has a positive impact on patient outcomes. For example, the Centers for Disease Control and Prevention (CDC) found that severely injured patients who receive care at a Level I trauma center had a 25% reduction in risk of death (NAS- EMSO 2010). The CDC source also emphasized the impor- tance of timely access to trauma facilities, but did not place specific limits on time. What is well documented is that response times are longer in rural areas than in urban areas. Furthermore, more than 36% of rural fatal crashes exceed the “golden hour,” mean- ing it takes more than 1 hour for injured crash occupants to receive hospital care after the crash has occurred. This figure compares with 10% of urban fatal crashes that take an hour or more to receive hospital treatment. Table 2 shows the average EMS response time for crashes in which at least one person died is presented for urban and rural areas. Delays or greater response times in rural areas are often related to increased travel distances. Significant delays may also occur as volunteer EMS personnel may travel first to the EMS station to retrieve the ambulance. In addition, rural areas without well-designed trauma systems may experience further delays in moving severely injured patients from rural hospitals to trauma centers. LITERATURE REVIEW AND SUMMARIZATION Fifty documents were reviewed for their relevance to the topic area. Of these, the major findings for 37 are summarized. Each literature source was categorized into one of the 10 topic Source: NHTSA (2010). TABLE 1 VEHICLES INVOLVED IN FATAL CRASHES BY SPEED LIMIT AND LAND USE

9 areas used here. For larger documents, a summary of contents is included as opposed to providing detailed information. Dispatching The efficacy of using a Global Positioning System (GPS) in EMS vehicles as a means of reducing response times was explored by researchers in “GPS Computer Navigators to Shorten EMS Response and Transport Times” (Floyd et al. 2001). Researchers conducted a two-part test of GPS effec- tiveness. The first part used nonemergency vehicles in pairs, sent from the same origin at the same time to the same desti- nation. One vehicle used a GPS device, the other did not. The second part placed the device in an EMS vehicle with a crew of either one EMT and one paramedic or two paramedics. The first part test runs revealed that there was no sig- nificant difference in distances traveled by the two vehicles. However, there was a significant difference in the meantime to arrival (13.5 minutes versus 14.6 minutes). The authors state that the GPS guidance appeared to have a greater effect at night when roadside signs are more difficult to see and where complex traffic patterns (involving one-way streets) are present. The second part interviews indicated an even split between respondents who found the device useful or somewhat useful and those who rated the device not useful on actual response runs. When asked if the device might be useful in areas where geographic familiarity is poor, all responded that GPS provided benefits. The authors con- cluded that GPS guidance can reduce travel times and that reductions in time will improve as users become more famil- iar with operating the device. A more detailed study of rural EMS response using GPS guidance was conducted in “Improving Rural Emergency Medical Service Response Time with Global Positioning System Navigation” (Gonzalez et al. 2009). Researchers equipped ambulances with GPS devices and recorded trip times for a 1-year period in a rural county in Southwest Ala- bama. The data were collected from both GPS-equipped and nonequipped vehicles and trips were aggregated by total length to the scene to make them comparable for time. Data were all aggregated based on whether the incident involved a motor vehicle crash or not. The authors found that GPS- equipped vehicles had shorter travel times in nearly all cases, with longer trips showing greater advantages. Table 3 summarizes the travel time differences. The authors conclude that rural EMS travel times can be significantly shortened by the use of GPS guidance devices. They also note that limitations of such devices, such as out- of-date mapping data and loss of GPS signals as a result of TABLE 2 FATAL CRASHES BY EMS RESPONSE TIMES WITHIN DESIGNATED MINUTES AND BY LAND USE Source: NHTSA (2010).

10 obstructions or system malfunctions, can affect performance and should be considered when deploying GPS guidance. TABLE 3 OVERALL MEAN EMS RESPONSE TIME GROUPED BY MILES TRAVELED With GPS Without GPS EMS Miles Traveled to Scene No. Calls Mean Response Time (min) No. Calls Mean Responses Time (min) p (t test) 2–5 518 5.1 627 5.3 0.03 6–10 164 8.7 143 11.1 <0.0001 11–15 44 13.3 56 16.3 0.0004 16–20 40 19.1 39 21.3 0.004 >20 25 23.5 28 32.1 <0.0001 Source: Gonzalez et al. (2009), Table 3. Record Linkages/Data Metrics: Real-Time Data Communications and Management Efforts have been made to integrate computer-aided dis- patch with the traffic management systems used in Traffic Management Centers (TMC). Computer-Aided Dispatch— Traffic Management Center Field Operational Test: Wash- ington State Final Report (SAIC 2006) examined the impact of direct data sharing between the Washington State Patrol dispatch and TMC software. The Operation Test had three components: a Primary Alert that transferred data from the computer-aided dis- patch system to the traffic management system for trav- eler information and Washington DOT maintenance/ scene support purposes, a Response Support component that allowed DOT information (construction information, traffic data, or other events) to be automatically trans- ferred to the dispatch system, and Secondary Alerts to provide data to non-State Patrol EMS responder dispatch systems. The Secondary Alert component was not imple- mented in this test because the prospective EMS partner was too small and had too focused a service mission to benefit from the system. The system for passing data between the systems was developed and functioned according to design expectations. However, the authors noted several elements that reduced the expected impact of the system: • The State Patrol and DOT personnel already had approved operational integration plans and had devel- oped methods for communicating data. This made the additional system less significant than if they had not already been working closely together. • The computer-aided dispatch system would not ingest data as originally planned. A separate web-based interface was required for the DOT data to be used. This additional inter- face interrupted the original workflow of the dispatchers. • The operational geographic boundaries differed between the State Patrol and DOT, which resulted in gaps in information being transferred between systems. • The traffic management software had significant laten- cies (up to 4 minutes) when importing data from the dispatch system. • Dispatch system modifications and upgrade schedules limited the speed with which the traffic management system portion could be deployed. • Dispatch and traffic management systems used dif- ferent coordinate systems to define data locations. Translating between these can introduce errors. • Different standards were used to encode the various data elements. For the data sharing to function, transla- tion between them was necessary. • Despite these limitations, it was noted that such inte- gration could be particularly useful in rural areas, where latency tolerances (for dispatch to traffic man- agement system transfers) could be longer and lower staffing levels could benefit from greater automation. A number of states have also begun deploying integrated voice and data radio networks to enable communications across multiple agencies performing a variety of functions. For example, the South Dakota Interoperable Communications Plan (South Dakota Public Safety Communications Council 2007) describes the arrangement, management, and proce- dures for use of a statewide, interoperable digital radio system. Agencies such as ambulance services, air ambulances, and hospital laboratories (for emergency and bioterrorism response) will have preprogrammed radios provided to them by the South Dakota Department of Health. Hospitals are expected to monitor the channel or “talk group” to which they have been assigned. Authorized users of the emergency response talk group include the following: • Law enforcement (federal, state, or tribal) • Fire departments • Any licensed ambulance service (ground or air) • Any hospital recognized by the Department of Health • Any emergency management agency recognized by the Department of Public Safety • Any state or local transportation agency • Transit systems (subject to approval) • National Weather Service offices • Support agencies (such as Red Cross or Salvation Army and service agencies for critical infrastructure). The report states that the high level of integration and interoperability incorporated into the design of the system is critical for states such as South Dakota, which are char- acterized by low population densities and long distances between facilities.

11 Many states have similar documents, although they dif- fer in process, capabilities, and operational rules. The South Dakota study identified Arkansas, Idaho, Iowa, Kentucky, Minnesota, Nebraska, New Hampshire, South Carolina, Vermont, and Wyoming as having communications interop- erability plans or systems in place for emergency respond- ers (South Dakota Public Safety Communication Council 2007). Eight of these 10 states are included in the survey element of this synthesis study. ITS and Transportation Safety: EMS Crash System Data Integration to Improve Traffic Crash Emergency Response and Treatment (Horan et al. 2009) conducted a case study examination of the state of real-time data integration across emergency response and transportation entities. Table 4 (Table 2.1 in the original document) summarizes the infor- mation technologies used at each stage of crash response. These individual information systems exist as “silos” of data, which do not link records, even though they relate to the same crash patient transfer of record of care/discharge. Efforts to improve data sharing and crash response focus on crash identification and improved collaboration between EMS and trauma data systems. These efforts are outlined in the Strategic Highway Safety Plans (SHSP) prepared by individual states. Table 5 (Table 3.1 in the original docu- ment) presents a brief summary. The study also conducted focus groups with represen- tatives from the Minnesota Emergency Medical Services Regulatory Board, the Health Department (State Trauma System), Department of Transportation (ITS Program and Office of Traffic Safety), and Department of Public Safety (Traffic Safety). The group feedback included several impor- tant aspects of EMS data management: • Information collection practices are not defined or enforced at the state level. • Data reporting generally takes place after admission to a trauma center, so it is not available to physicians when the patient arrives. • Communication infrastructure factors may prevent wireless transmission of patient data. • The current consolidation of dispatch centers has cre- ated uncertainty around roles and responsibilities. • Limited availability of staff time hinders data analysis. • Financial concerns limit hospitals’ willingness to share data that could reveal pricing structures. • Data privacy policies hinder sharing of data between agencies. In addition, a case study of the Mayo Clinic in Roches- ter, Minnesota, was completed to assess how data exchange could be enhanced through information technology (Horan et al. 2009). A series of three focus group sessions created a series of findings that were organized into three areas: 1. Operational Linkage Issues: Stored patient data (e.g., medical records) are not available to EMS personnel; data must be entered multiple times into different systems; “siloed” records do not allow for analysis of patient outcomes related to emergency response prac- tices; lack of open standards makes system interoper- ability challenging. TABLE 4 EMERGENCY RESPONSE PROCESS INTERVALS AND SAMPLE TECHNOLOGIES USED Process Intervals Example Information Technologies Used Pre-incident preparation Electronic Personal Health Record (PHR) for emergencies (the AAA card for personal health emergencies) From “crash” to “notification” 911, E-911, AACN technology and integration (e.g., Mayday system) From “notification” to “dispatch” Computer-aided dispatch (CAD), traffic management systems, GPS + [Geographic Infor- mation System (GIS)], mobile data terminals (MDTs), decision support tools, 2-way radios, pagers, cell phones From “dispatch” to “arrival on scene” (in-field care) CAD, patient care record (PCR) systems, traffic management systems, GPS + GIS, MDTs, decision support tools, 2-way radios, pagers, cell phones, navigation systems From “arrival on scene” to “departure to hospital/trauma center” (in-field care and transport) PCR systems, decision support systems, telemedicine applications (remote care), wireless data communications, hospital availability/diversion systems From “departure to hospital/trauma center” to “arrival to hospital/trauma center” [transport and handoff to hospital emergency department (ED)] PCR systems, traffic management systems, GPS + GIS, navigation systems, hospital availability/diversion systems From hospital “admission” to “discharge” Hospital emergency department admissions/registry, trauma registry, electronic medi- cal records, clinical information systems, electronic lab/radiology systems, clinical decision support Post-incident evaluation Crash Outcome Data Evaluation System (CODES), data warehouses, business intelli- gence, crash analysis reporting systems [e.g., Fatality Analysis Reporting System (FARS)], other reporting and analytics Source: Horan et al. (2009), Table 2.1.

12 TABLE 5 COMPARISON OF SHSPS SHSP EMS Related Descriptions SHSP Described Efforts and Demonstrations Minnesota Focus: Creation of a statewide system to reduce crash response times by improving patient to trauma ward routing practices Improvement on ACN and 911 routing communications and development of rural intersection decision support technologies Alabama Focus: Reducing the time from crash to care by ensuring that trauma patients are transported to an appropriate facility with resources to care for patient injuries Provide crash location through advanced GPS tech- nologies; make efforts toward statewide EMS quality and services coordination and increase consumer education on traffic safety. Improve electronic data and voice Maryland Focus: Improving EMS across a range of technology, process, and program improvement Improve electronic data and voice communications for emergency response and improve resource deployment for EMS response California Focus: Reduce crash-related fatalities by at least 5% from 2004 levels through improvements in EMS system communi- cations, response and safety education Advance technologies for locating crash sites, improv- ing EMS access routes, dis- patching, decreasing response times and increasing overall EMS system resources and effectiveness Utah Focus: Review of current systems in order to increase opportunities for crash data use Plans to advance development of technologies to analyze, and distribute crash data in a timely manner across multiple agencies with goals of increasing quality assurance standards Washington Focus: Continued efforts in developing Washington’s EMS and Trauma Care System (EMSTC) Improve communications between response agencies, implementation of dispatch protocols, statewide imple- mentation of GPS technology and continued efforts in part- nerships to improve data Source: Horan et al. (2009), Table 3.1. 2. Organizational Linkage Issues: Need for system- wide, interorganizational approaches to integration; need for individuals to “self-check” performance; greater interaction across agencies would improve trust and cooperation; involvement of stakehold- ers (e.g., legislators, EMS agencies) is needed for improvements. 3. Governance Linkage Issues: Use contracts to enforce performance levels with partner agencies; use infor- mation sharing even when contracts to do so are not in place; seek opportunities with agencies such as the Department of Homeland Security; costs of per- sonnel to implement/manage data systems can be prohibitive. The authors propose that an Integrated Crash Trauma Network is needed to permit access to a broad range of medi- cal and EMS data in a timely, uniform fashion, and recom- mend that their findings be validated and that a prototype software deployment be constructed to verify functionality and provide a base that can be improved through an iterative feedback process. Table 6 summarizes the EMS information contained in SHSPs for the 14 focus states. (Note: The SHSP documents were found from FHWA website links, and may not be the most recent SHSP documents available.) The table summa- rizes the EMS objectives and EMS described strategies, and finds that the emphasis on EMS varies from state to state. In a follow-up study, ITS and Transportation Safety: EMS System Data Integration to Improve Traffic Crash Emer- gency Response and Treatment—Phases IV and V (Schooley et al. 2012), the authors describe the second version of a soft- ware system called CrashHelp that uses a combination of mobile smartphone, multimedia, web server, and location- based technologies to enable information transfer between hospitals and responding paramedics. The study addresses a pilot test of the system in the Boise, Idaho, region. The CrashHelp system has three main components: 1. A smartphone application for paramedics that can be used to communicate voice, video, pictures, and patient condition information. 2. A web application for emergency departments to review multimedia patient condition information, pre- pare for patient arrival, and communicate with medics as needed. 3. A backbone enterprise application server facilitating management and exchange of information between the first two components. The 3-month pilot test included 20 ambulances across two agencies. Each ambulance was provided with a mobile smart- phone, but they were not required to utilize the device at any point during the 3-month trial. Just under half of the para- medics used CrashHelp at least once with positive results. The most frequent features utilized were camera, audio, texting, and notifications. The electronic map and video fea- tures were used less often. It was determined that CrashHelp benefited EMS incidents with higher severity levels and longer transport times as opposed to incidents with short transport times. This finding could be interpreted as the system being of more help to rural incidents where transport times are longer. Potential system improvements resulting from the pilot test included addressing integration with existing EMS and hospi- tal information systems, improving automatic notification of new CrashHelp records, a mobile application to be used on

13 other devices, and the ability to enable hospital referrals. Addi- tional challenges to be addressed include EMS picture-taking protocols, flexibility for new features, and paramedic versus emergency department expectations. A second phase of the pilot began in summer 2012 to address these challenges and further explore the impact on patient care within rural settings. Record Linkages/Data Metrics: Retrospective Data Communications and Management National Emergency Medical Services Information System (NEMSIS) is a collection of software tools and data reposi- tories intended to facilitate the collection, aggregation, and dissemination of information related to emergency medical service response and outcomes. In “National Emergency Medi- cal Services Information System (NEMSIS)” (Dawson 2006), the author outlines the five recommendations made by NHTSA in 1996 that formed the foundation of NEMSIS, including: • Adoption of a uniform set of EMS data elements. • Development of reliable, accurate mechanisms for col- lecting and transmitting data. • Creation of comprehensive information systems to bet- ter assess patient outcomes and cost-effectiveness. • Collaboration of EMS and other health care providers to develop integrated information systems. • Development of a system to provide feedback to those who generate and input data. NEMSIS funding began in 2001. The data set has been housed at NSHTA’s National Center for Statistics and Analy- sis since 2005. The most recent version of the NEMSIS Data Dictionary is 3.2.6. Plans to create integrated records systems continue to be studied. The authors of “Developing a Statewide Emergency Medical Services Database Linked to Hospi- tal Outcomes: A Feasibility Study” (Newgard et al. 2011) developed a system of probabilistic record matching using LinkSolv, a statistical software package that can match records with incomplete data. Using patient care reports (run sheets) from EMS agencies, 60 NEMSIS data fields, and 23 additional fields needed to complete patient care record matching, the authors attempted to track care TABLE 6 COMPARISON OF SHSPS FOR 14 FOCUS STATES SHSP State EMS Focus SHSP EMS Objectives SHSP Described Strategies Arkansas Yes Improve access to crash sites and prehospital data col- lection systems, and develop statewide trauma system. Install median gaps and acceleration/deceleration lanes, establish statewide trauma system, enhance data collection for traumatic injuries, and update systems to meet NEMSIS data elements. Idaho Yes Quick and effective response to address the care of injured crash occupants and improve emergency communication. Improve emergency scene management through training and funding to reduce rural transport times for patients. Iowa No Kansas No Kentucky Yes Improve accessibility of EMS data through a statewide reporting system for EMS ambulance runs. Use Section 408 funds to implement Kentucky Emergency Medical Services Information System. Mississippi Yes Improve EMS response times in rural areas, and data linking between EMS, enforcement agencies, ambu- lance services, emergency departments and hospitals. Increase funding of statewide trauma system, and improve access and integration of data. Montana Yes Develop an effective and integrated EMS delivery system. Provide comprehensive data collection and information system, and incorporate AACN. Nebraska Yes Provide support for 44 designated trauma centers through registry and data information. Implementation of statewide trauma registry, regulations requiring use of the NEMSIS patient reporting data set. New Hampshire No South Carolina Yes Expand EMS in rural areas where response time is greater than 10 minutes; expand communications; improve location coding for rural areas. Implement NEMSIS data collection related to rural crash types; implement electronic data capture. South Dakota Yes Web-based data collection software to increase contri- butions of response times and information. Mandate all hospitals to collect and report data on trauma patients within a trauma registry. Working with [National EMSC Data Analysis Resource Cen- ter (NEDARC)] to examine data collection, with the potential to integrate crash, hospital, and ambulance service data. Work with Governor’s EMS Advisory Committee on trauma regis- try legislation. Vermont No West Virginia Yes Reduce delays in discovery and rapid response, trans- port times, and limited resources. Establish data collec- tion system with injury surveillance and EMS run data. Ensure EMS access and coverage statewide, improve commu- nication, develop electronic patient care record, and imple- ment electronic EMS run form and rural inclusive trauma sys- tem/trauma registry. Wyoming No

14 records from the initial crash reporting though patient dis- charge following care. The report The REACT Project: Rural Enhancement on Access and Care for Trauma (Garrison et al. 2002) summa- rizes the results from the Rural Enhancement of Access and Care for Trauma (REACT) project performed by East Caro- lina University and the Eastern Carolina Injury Prevention Program. The REACT project itself was a follow-up to the 1992 NHTSA-sponsored Rural Preventable Mortality Study (RPMS). According to the RPMS, eastern North Carolina had an overall preventable mortality rate of 29%. The main objective of the REACT project was to decrease this rate of preventable deaths from injury in rural settings. Intervention took place in the same region where the 1992 RPMS was conducted. This region, served by the trauma service of Pitt County Memorial Hospital, is composed of 29 counties in rural eastern North Carolina. The intervention phase of the REACT project had three components: • Partnership with the Eastern Regional Trauma Coalition to develop trauma care guidelines for the treatment of trauma patients, which addressed the defi- ciencies identified in the 1992 RPMS study. • Guideline-focused, in-depth training for emergency medical personnel in the region. • Feedback to emergency medical personnel on their conformance to the guidelines. The primary means of incorporating these three compo- nents into the existing trauma service was use of the Stan- dards of care, Training, And Feedback (STAF) model. This model was directed toward prehospital and hospital emer- gency providers in the rural areas served. Based on the intervention, there were two subsequent evaluation components: • Assessment of the compliance with trauma care guide- lines during the intervention phase. • Determination of the preventable mortality rate for the region during the intervention year to determine if intervention had an impact. As a result of this intervention, improvements were seen in both compliance and the preventable mortality rate. Com- pliance was measured over each quarter during the year and showed improvements from the first through fourth quar- ters. A comparison of the 1992 RPMS study and the 1998 REACT project show that preventable death rates were cut nearly in half. • The REACT Project—1997/1998 – Overall preventable death rate of 15% – 31% of cases had some aspect of inappropriate care. • The RPMS—1992 – Overall preventable death rate of 29% – 68% of cases had some aspect of inappropriate care. Overall, implementation of the STAF model appeared to reduce the rate of preventable trauma deaths in rural areas, when the REACT Project is compared with the RPMS. This finding was also supported by other mortality studies con- ducted previously in both Michigan and Montana. An overview report by NHTSA, The Crash Outcome Data Evaluation System (CODES) and Applications to Improve Traffic Safety Decision-Making (NHTSA 2010), provides insight into CODES, which is a part of its State Data Program. The basic concept of CODES is to link crash records to injury outcome records. These injury outcome records are obtained on scene or en route by EMS, by hospi- tal personnel or at the time of death. Analysis includes both injured and noninjured people to reduce bias that may result by not including data from unexpected outcomes. Application of CODES and other related programs result in four main objectives: • Objective 1: Identify Traffic Safety Problems—Because CODES data are population-based, they can be used to help identify safety issues, including potentially sig- nificant crash outcomes. Some safety issues identified relate to teen driver crashes and passenger injuries, hospital charges and durations for motorcycle-related injuries, seat belt usage, and child injuries in passenger motor vehicles, among others. • Objective 2: Support Traffic Safety Decision-Makers— Through CODES data, individuals in charge of mak- ing state and local traffic safety decisions can be better informed and educated. This helps prioritize traffic safety issues with other public health issues. • Objective 3: Support Traffic Safety Legislation—Many traffic safety decision-makers are working toward leg- islation that can result in meaningful safety impacts. With CODES data, legislators are more aware of traffic safety issues in their state. • Objective 4: Educate the Public—Educating the gen- eral public is critical because they are the motorists that make up the statistics in the CODES data. Properly educating the general public on traffic safety informa- tion and providing it through a convenient means may help to reduce crashes. The CODES report provides a case example of a study completed in Alabama. The study “Does Increased EMS Pre-Hospital Time Affect Patient Mortality in Rural Motor Vehicle Crashes? A Statewide Analysis” was per- formed by the Center for the Study of Rural Vehicular Trauma. The report identifies a distinction between rural and urban fatalities, which was found by measuring EMS

15 Combix Corp. created a system that used GPS positioning and vehicle sensors to detect and relay crash data. Intrado functioned as the ACN service provider and routed calls as needed. The project demonstrated that all of the components for ACN could be successfully integrated and used as part of an existing E911 system. In Germany, the Emergency Call Center (ECC) receives data from the vehicle following an automatic crash detection or manual activation. The ECC then identifies the appropri- ate responder based on crash location and characteristics and notifies the appropriate responder. The use of software to interpret AACN data and pro- vide responders with information about potential injuries is detailed in “Reducing Highway Deaths and Disabilities with Automatic Wireless Transmission of Serious Injury Probability Ratings from Crash Recorders to Emergency Medical Service Providers” (Champion and Augenstein 2003). The report outlines the time-related issues in rural crash responses and contrasts them to urban statistics. Table 7 (Table 2 in the original document) illustrates the time dif- ferences for each phase of crash response. TABLE 7 AVERAGE ELAPSED TIMES IN FATAL CRASHES IN 1998 (MINUTES) Time Intervals Urban % Unknown Rural % Unknown 1. Crash to EMS Notification 3.6 46 6.8 37 2. EMS Notification to Scene Arrival 6.3 47 11.4 35 3. Scene Arrival to Hospital Arrival 26.6 72 36.3 67 4. Crash to Hospital Arrival 35.5 71 51.8 68 5. Recommended Time for ED Resuscitation (No Data in FARS) 15 15 Average Totals 51 67 Source: Champion and Augenstein (2003), Table 2. The study table noted the following points: • These are U.S. average elapsed times that consist of shorter and longer times and vary greatly by state. • Time intervals 2 & 3 do not include the elapsed time from crash to EMS Notification. • Bolded times indicate the average elapsed times that exceed benchmarks of 1 minute for EMS notification, 10 minutes for EMS scene arrival, and 45 minutes for hospital arrival in fatal crashes. The authors noted that increased times between a crash and notification of EMS personnel are associated with higher percentages of crash occupants who die at the scene of injury. They concluded that AACN notifications within 1 response, scene, and transport times. Average EMS pre- hospital times resulting in fatalities were 42.0 minutes for rural incidents and 24.8 minutes for urban incidents (Gon- zalez et al. 2009). Crash Detection, Locating, and Reporting The crash-to-responder notification interval is an area that has been studied for its potential to reduce EMS response times. AACN systems use vehicle sensors (or manual triggering by occupants) to contact an external entity (generally a service provider) who can then assist the traveler or connect them with a PSAP. AACN, the successor to the Automatic Collision Notification (ACN) systems, incorporates vehicle sensor data and implements the Vehicular Emergency Data Set. These fea- tures are expected to enhance the usefulness of AACN over earlier ACN deployments, some of which are described here. Crash Location Systems (CTC & Associates 2003), a synthesis report prepared in 2003 for the Wisconsin DOT, documented ACN system development and availability to that point. Historical highlights included the formation of the Multi-Jurisdictional Mayday (MJM) Project in 1995, which facilitated operational ACN tests in Colorado, Min- nesota, New York, and Washington State. The MJM’s active phase ended in 1998 with the completion of operational tests and publication of evaluation reports. In October 2000, the National Mayday Readiness Initiative completed a set of rec- ommendations, which included: • Updating training standards for call takers to properly receiving ACN information and defining a process for accreditation of ACN service providers • Creating a directory of all public safety agencies in the United States • Improving data sharing procedures for emergency response agencies • Developing uniform and acceptable business practices for ACN service providers • Continuing focus on developing/deploying CAN. The Minnesota Mayday Field Operational Test had two goals: (1) develop solutions for routing incoming AACN notifications to appropriate responders, and (2) obtain and interpret vehicle data to determine characteristics of the crash (e.g., rollover, final vehicle position). It used General Motors’ OnStar technology as its foundation and sought to minimize startup costs through use of existing infra- structure. Cases where crashes were so severe that the ACN hardware was disabled were identified as a place for additional research. The first end-to-end trial of an AACN system was in Har- ris County, Texas. A partnership among the Greater Harris County 911 Network, Intrado Communications, Southwest- ern Bell, Veridiant Engineering, Plant Equipment, Inc., and

16 minute of a crash are technologically possible and economi- cally feasible, potentially reducing response times by up to an average of 5.8 minutes. In March 1997, NHTSA funded development of the URGENCY 1.0 software package. URGENCY used data from vehicle sensors to triage crashes and assign a severity indicator ranging from 0% to 100%. The authors state that the combination of AACN and URGENCY software will improve response times and occupant survival rates. Reduc- tions in fatalities as high as 20% resulting from the use of AACN systems are believed to be possible. Other studies cited by the authors indicate reductions of up to 6%. The authors conclude that it is both technologically pos- sible and economically feasible to have EMS crash notifica- tion within 1 minute, EMS scene arrival within 10 minutes, and trauma center arrival within 45 minutes for many crashes. However, they do not make specific statements about rural crashes. Telemedicine The Telemedicine Journal and e-Health study “Tele- medicine Reduces Discrepancies in Rural Trauma Care” (Amour et al. 2003) is one of the first attempts to put a quantitative metric on the benefits of telemedicine for rural trauma patients. The goal of this project was to measure the effectiveness of allowing specialized surgeons to consult with local physicians on how to treat rural trauma patients. It involved Fletcher Allen Health Care (the level 1 trauma center in Burlington, Vermont) and four rural hospitals. Surgeons from the trauma center were equipped with video and audio transmitters in their homes and at work so they could consult physicians at the rural hospitals. Some key aspects to the study: • Clinical outcome measures were developed before the implementation. • Evaluation questionnaires were designed for the patients and users of the telemedicine system. • Multiple telemedicine sites were set up at the trauma center as well as a telemedicine site in each of the sur- geons’ homes. A telemedicine consult was instituted when the patient had a Glasgow Coma Score of less than or equal to 13, penetrating truncal trauma, respiratory distress, or amputation, or when the physician decided it was needed. The outcome of the teletrauma consults was evaluated by two means: (1) comparing the patients of telemedicine and nontelemedicine with a standardized scoring system, namely the Injury Severity Score (ISS); and (2) interviews and questionnaires given to referring and consulting physi- cians on the effectiveness of the telemedicine treatment. Forty-one trauma consultations were performed, with 49% consisting of motor vehicle crashes. Three of these con- sultations were deemed to be lifesaving in the post-surgery interview. However, this experiment did not find telemedi- cine to be statistically beneficial compared with patients who did not receive telemedicine. Teletrauma patients had average adjusted ISS scores of 25.3 compared with the non-teletrauma patient score of 18. Teletrauma did not statistically decrease total time from injury to arrival at trauma center; however, the overall mean travel time was 34.8 minutes less (p = 0.26). Also, the length of stay of teletrauma patients was not significantly longer. The lack of conclusive results may stem from the smaller number of telemedicine consults and the fact that patients who received telemedicine treatments had much more severe injuries. However, the surgeons who completed the ques- tionnaires had a positive view of telemedicine. In 61% of the cases, they believed that telemedicine had improved patient care, and in 67% of the cases they believed that the recom- mendation could not have been made by telephone. The Journal of TRAUMA® Injury, Infection, and Critical Care published a study on the effect of implementing tele- medicine stations in Mississippi. The “Impact of Telemedi- cine upon Rural Trauma Care” (Duchesne et al. 2008) study equipped seven local hospitals in the state with remote con- trollable cameras providing access to an experienced surgeon at the University of Mississippi Medical Center. This trauma center receives an average 3,500 trauma patients a year, of which 60% are transfer patients. The data for the study were collected over a 5-year period, which included the 2.5 years before and after the telemedicine equipment was set up in the local hospitals. Data collected for this study included the mode of transportation; length of local hospital stay; and transfer time, or was the time between the initial report at the local hospital and the arrival at the trauma center. The telemedicine system allowed the physician at the local hospital to request a consult or a “stat” from a trauma center surgeon. Requesting a consult would put the local physician and patient into a queue. However, if the patient required immediate assistance, a “stat” would be requested, which put the conference first in the queue. A total of 351 trauma patients were presented to the local hospitals during the first half of the study and 463 during the second half. Of the 351 original patients, 100% were transferred to the trauma center, and of the 463 telemedi- cine patients, 11% were transferred to the trauma center and 1.1% were transferred to another local hospital. There was no significant difference in the mode of transporta- tion used for transfer. The length of stay pre-telemedicine was 47 hours compared with 1.5 hours post-telemedicine. The transfer time (the time between the initial report at

17 mortality during the same time in the control region, and 7% before the discontinuation. • The mortality rate did not change in the control region between the two time periods. Before the discontinuation, the two regions had trans- ferred presenting patients 24% of the time for the test region and 25% for the control region. After, the test region trans- ferred 15% and the control region transferred 51%. Before the discontinuation, the median transport times for the test and control regions were 2:07 (hr:min) and 2:15, respec- tively. After, transport times were 3:10 and 2:10, respec- tively. This study concluded that without the availability of air medical transport, the odds of death increased among severely injured patients in rural areas. Also, the lack of air medical transport increases transfer time between rural hos- pitals and trauma centers. The “Helicopter Use in Rural Trauma” (Shepherd et al. 2008) study, published in Emergency Medicine Australasia, focuses on providing statistics for helicopter versus vehicle transport for rural trauma patients in the Australian outback. The study also provides insight to the effectiveness of an on- flight physician. The study data come from the documented activity of the Helicopter EMS (HEMS) in rural northwest- ern New South Wales. Helicopter trauma incidents were found by reviewing the helicopter operator’s activity log from January 2004 to November 2006. Response and scene times were estimated from engine hours in the helicopter’s maintenance log. Ambulance travel times were estimated from GPS mapping by finding the most direct route from the Tamworth Rural Referral Hospital to the accident scene. Two hundred and twenty-two trauma missions were identified with the activity log, of which 171 had records with complete data for analy- sis. Eighty-seven of the 171 of the trauma injuries were vehi- cle-related, and 129 were taken to Tamworth Rural Referral Hospital. Table 8 (Table 3 in the original document) shows the local hospital and the arrival at the trauma center) for pre-telemedicine was 13 hours compared with 1.7 hours post-telemedicine. The implementation of telemedicine significantly reduced the total transfer time and length of stay for trauma patients presented to local community hospitals in Mississippi. The decrease in these times is thought to come from the better understanding of the initial care, and through improved communication with the trauma center. Air Medical Transport This section focuses on the use of air medical transport (heli- copter and fixed-wing aircraft) centered on performance (transport time) and cost-effectiveness measures. Academic Emergency Medicine published a study on the effect of the lack of helicopter transport on the mortality rate of severely injured patients. “Injury Mortality Following the Loss of Air Medical Support for Rural Interhospital Trans- port” (Arthur et al. 2002) compared two regions, each consist- ing of four rural hospitals and one tertiary trauma center. The study compared the mortality rate of the two regions before and after the “test” region had discontinued the use of heli- copter transport. The comparison data were collected from the 3 years before helicopter transport was discontinued in the test region and the 3 years after it was discontinued. The patient data consisted of the ISS; level of neurologic function on the Alert, Voice, Pain, Unresponsive scale; and mortality up to 30 days after discharge from the tertiary trauma center. • Over the 6-year test period, 38% of presenting patients from the control region and 20% of the presenting patients in the test region received inter-hospital transfer. • The mortality rate of transferred patients was 26% after the loss of helicopter transport compared with 9% TABLE 8 COMPARISON OF TIMES FOR TWO-WAY HELICOPTER JOURNEY AND ONE-WAY ROAD RETRIEVAL Source: Shepherd et al. (2008), Table 3.

18 them with 7,854 ambulance records to determine response times categorized by distance of transport. The authors con- clude that for distances of greater than 10 miles from scene to care facility, simultaneous dispatched helicopters result in lower overall transport times. For nonsimultaneous dis- patched helicopters, ground ambulances result in shorter average transport times for distances of less than 45 miles. A number of notes are given to qualify the conclusions, including the following: • It has not been definitively established that patients derive a benefit from air medical transport. Although some studies cited in the report indicated decreased patient mortality, this is only for the most severely injured patients. • Air medical transport is substantially more expensive (by a factor of 5 to 10) than ground transport. • Geography, roadway characteristics, availability of landing zones, and other factors unique to the study area may limit the ability to generalize findings to other areas. “Cost Effectiveness Analysis of Helicopter EMS for Trauma Patients” (Gearhart et al. 1997) addressed the cost- effectiveness of air medical transport. The basis for com- parison chosen by the authors was cost per life saved and cost per year of life, which was discounted to current dollars for the analysis. Determination of additional lives saved used a proba- bilistic model based on Trauma Score—Injury Severity Score calculations. To ensure that calculated values agreed the mean transit time for helicopter and vehicle transport for different distances. There was no significant difference between outcome of patients treated by the ambulance office and the outcome when a physician was on site. Because of this, the author suggests caution in mandating that a physician be on all helicopter missions, especially in areas with limited physi- cian availability. “When Is the Helicopter Faster? A Comparison of Heli- copter and Ground Ambulance Transport Times” (Diaz et al. 2005) is a retrospective analysis and comparison of transport times for ground ambulances and helicopters. The authors differentiate between simultaneous and non- simultaneous dispatch, which is when an ambulance and helicopter are directed to the scene immediately on receipt of a call versus when a responder arrives on scene by means of ground transport and then assesses whether to request helicopter response. The analysis used data from Fresno and King counties in California over a 4-year period (1996–2000) for ground ambulances and a 3-year period (1997–2000) for helicop- ters. Distance for ground vehicles was determined from odometer readings recorded by ambulance crews. The response distances for helicopters were taken from on-board GPS receivers. To correct for roadway versus “straight line” miles flown by helicopters, a factor of 1.3 was applied to the odometer readings. The analysis made use of 715 simultaneous and 360 non- simultaneous helicopter dispatch records and compared TABLE 9 EFFECTIVENESS OF HELICOPTER EMS Author (Year) No. of Patients No. of Expected Deaths No. of Observed Deaths Expected Mortality (e/n) Standard Mortality Ratio (o/e) Additional Lives per 100 Flights Baxt (1987)* 574 36.4 30 0.063 0.82 1.1 Cameron (1993) 242 41.8 34 0.17 0.81 3.2 Schmidt (1992)* 407 57.0 42 0.14 0.74 3.7 Baxt (1985) 1,273 240.7 191 0.19 0.79 3.9 Hamman (1991) 259 32.0 20 0.12 0.63 4.6 Rhodes (1986) 130 28.6 22 0.22 0.77 5.1 Our study (1997) 604 90.3 50 0.15 0.55 6.7 Schwartz (1989) 168 36.7 25 0.22 0.68 7.0 Baxt (1983) 150 20.6 10 0.14 0.49 7.1 Campbell (1989) 168 50.0 31 0.3 0.62 11.3 Boyd (1989) 103 45.5 33 0.44 0.72 12.1 Source: Gearhart et al. (1997), Table 1. *These studies report two separate cohorts, which are combined in this table. Notes: • Expected deaths (e) are calculated with TRISS analysis. Only the air cohort is used from studies including both air and ground cohorts. • Additional lives per 100 flights are calculated as: (e/n)[1 - (o/e)]lO0.

19 with the range previously established in other studies, the authors reviewed research to compare their values with other estimates for additional lives saved per 100 flights. Table 9 (Table 1 in the original document) presents this comparison. The authors completed a cost assessment based on data from a university-based hospital with a helicopter service. Costs included capital, operations, personnel, and overhead (e.g., insurance, administrative) elements. Using these data, the cost-effective measure (cost per life saved) for helicop- ter transport was estimated at $60,163 with a discounted cost (at 3%) of $2,454 per additional year of life. This fig- ure measures favorably when compared with other medical intervention, such as the cost-effectiveness of prehospital paramedic system, which was $8,886 per additional year of life. The authors conclude that air medical transport of trauma patients has cost-effectiveness similar to other life- saving measures. Apart from comparisons of transport methods, the effi- cacy of stabilization of rural trauma patients at Level III emergency departments before transport to Level I trauma centers was examined in “Stabilization of Rural Multiple- Trauma Patients at Level III Emergency Departments before Transfer to a Level I Regional Trauma Center” (Veenema and Rodewald 1994). The authors examined cases in Wayne County, New York. The study included two hospitals with Level III emergency departments. There was no double- physician coverage in either emergency department at any time, although both were staffed 24 hours a day. There was also no organized trauma protocol in use at either emergency department during the study period. A total of 50 patients met the study criteria (Revised Trauma Score > 11, admitted to Level III emergency depart- ment and either died at Level III emergency department or were transferred to Level I emergency department). These cases were divided into three groups based on their time spent at the Level III emergency department before trans- fer to a Level I facility and assessed for unexpected out- comes when compared with model predictions for outcome based on trauma score. Neither the “short time” group (35 to 65 minutes at Level III emergency department) nor the “long time” group (173 to 415 minutes) showed any unexpected outcomes (survival or deaths). However the “middle” group had two unexpected survivors and one unexpected death. The authors conclude that stabilization at Level III emergency departments is viable as an alter- native to long-distance EMS transport to regional trauma centers. However, they caution that small sample sizes and variability in Level III trauma protocol adherence should be taken into account. “Helicopter EMS Transport Outcomes Literature: Anno- tated Review of Articles Published 2007–2011” (Brown et al. 2012) is a review article that presents the most important HEMS outcomes published between 2007 and 2011 as an evidence basis for HEMS use. “Outcomes of Blunt Trauma Victims Transported by HEMS from Rural and Urban Scenes” (McCowan et al. 2007) compared mortality rates of HEMS-transported trauma patients of urban and rural scenes. The study loca- tion was Salt Lake City, Utah, and consisted of a review of records from two HEMS as well as three receiving Level I trauma centers. The review and comparison of urban and rural trauma patient outcomes concluded that there were no significant differences in mortality rates. Therefore, HEMS use for rural trauma patients eliminates the differences in mortality rates that are evident in urban and rural ground- transported trauma patients. “Helicopter Emergency Medical Services (HEMS): Impact on On-scene Times” (Ringburg et al. 2007) com- pared prehospital on-scene times for patients treated by ground EMS with those treated by HEMS. A trauma reg- istry study was performed using data from Rotterdam, the Netherlands. HEMS patients had longer average on-scene times than ground EMS by about 9 minutes, with a logistic regression suggesting a 20% higher chance of dying associ- ated with a 10-minute increase in on-scene time. However, this effect was eliminated for HEMS attended patients. Despite the increase in on-scene time associated with HEMS use, HEMS are able to provide “golden hour” procedures at an earlier time than ground EMS, eliminating the adverse effects of on-scene times. “Helicopter Use in Rural Trauma” (Shepherd et al. 2008) looked to determine whether any time savings were associ- ated with HEMS use. Through a medical records review of multiple hospitals in New South Wales, it was determined that HEMS had a time savings advantage for distances greater than 100 km. Between 50 and 100 km, there were no time dif- ferences between ground EMS and HEMS, with ground EMS being significantly faster under 50 km. This study did not look at patient outcomes, only times-to-trauma centers. “Helicopters and the Civilian Trauma System: National Utilization Patterns Demonstrate Improved Outcomes after Traumatic Injury” (Brown et al. 2010) focused on outcomes of injured patients for helicopter transport and ground transport using the National Trauma Databank. It concluded that helicopter transport increased the chance of survival and discharge to home after treatment. The authors also noted that helicopter transport is frequently used because of distance and geography rather than injury severity alone. The “Helicopter EMS: Research Endpoints and Potential Benefits” (Thomas and Arthur 2012) article reviewed the potential benefits that HEMS provide to patients, EMS sys- tems, and health care regions. A large amount of information

20 in the review highlighted the importance of HEMS. Several key notes regarding HEMS are— • Trauma surgeons support that HEMS response to trauma scenes provides life-saving care during the “golden hour” that is over and above care rendered by an Advanced Life Support (ALS) ground ambulance. • HEMS can cover roughly the geographic area of seven ground ALS ambulances. • It is well known that, particularly for rural locations, prolonged EMS response/transport time results in increased trauma mortality. • HEMS crews’ on-scene times are about 10 minutes longer than those for ground EMS crews. • Similar scene-to-trauma center times for ground and HEMS transports were found in studies conducted in California and the Netherlands. • A 2005 Journal of the American Medical Association study found that HEMS was the only mechanism by which 27% of the U.S. population had timely Level I or Level II trauma center access (within an hour of receipt of emergency call) (Branas 2005). • The authors concluded that new helipad placements and additional HEMS programs “could be an impor- tant, and practical, means of extending trauma center access to populations that currently have none” (Baxt and Moody 1983, p. 2631). • Preliminary analysis has suggested that HEMS is actually no more expensive than the multiple-ground- unit alternative. • Data suggest that the early arrival of those able to provide ALS-level airway and hemodynamic support translates into improved overall outcome and better neurological function. • While there is no “golden hour” for burn patients, epidemiologists and clinicians writing in a Journal of the American Medical Association study (Klein et al. 2009) point out that early care (in the first few hours) at a burn center improves outcome and that HEMS is the sole mechanism by which millions of Americans have access to burn center care within 2 hours of injury. “Helicopter EMS: Outcomes Research, Cost-Effective- ness, & Triage” (Thomas et al. 2013) focuses greatly on HEMS, its potential benefits, and cost-effectiveness. It is difficult in general to study cost-effectiveness, but being that there are limited data regarding ground EMS to begin with, it only complicates things further. The main analysis in the study was of existing literature and study methodologies. According to the study, there are many benefits to HEMS when compared with ground EMS. Some takeaways and important comparisons are: • Fatality rates per mile traveled are 0.4 per million air miles for HEMS compared with 1.7 per million ground miles traveled. • Based on more than two decades of data, U.S. helicop- ter operations report less than one patient death per 100,000 missions. • Early arrival of ALS, especially in rural or isolated areas, results in significant reduction of time to treatment. • HEMS was estimated to be the only means of having timely (less than 1 hour) access to Level I or Level II trauma centers for 81.4 million Americans. From a cost-effectiveness standpoint, a Norwegian study estimated the benefit-to-cost ratio of helicopter EMS to be 5.87:1. In addition comparisons of HEMS to ground EMS on the regional level have indicated that HEMS may be less expensive than deployment of a wide-ranging fleet of ground EMS vehicles. In comparing costs of HEMS to ground EMS: • Per-patient costs from 1991 were $4,475 for HEMS and $2,811 for ground EMS. • Cost per life-year saved has associated costs of $2,454 for HEMS and $8,886 for ground EMS. • The study concluded that with both the benefits and the cost-effectiveness exhibited by HEMS that HEMS, in some form, is a must-have for many U.S. EMS regions. Care Protocols and Procedures The “Rural EMS En Route IV Insertion Improves IV Inser- tion Success Rates and EMS Scene Time” (Cummings et al. 2011) study, published in The American Journal of Surgery, compared the consequences of on-scene intravenous (IV) insertion with en-route IV insertion. IV insertion, which is a component of advanced life support, has been questioned because it can increase the on-scene time and therefore the crash-to-care time. The two metrics that were used to com- pare the two insertion methods were on-scene time and IV insertion success rate. Study data were provided by the EMS provider in DeKalb County in rural Alabama. During 2007, none of the IV inser- tions performed by the EMS provider were done en route. In 2008, all of the IV insertions were attempted en route. The year of on-scene IV insertions produced these results: • Three hundred and six trauma patients received IV insertion attempts on scene • Seventy-six percent were vehicle crash occupants • IV insertion success rate was 79% • Average EMS on-scene time was 19.8 minutes. The year of en route IV insertions produced these results: • Three hundred and forty-one (239 motor vehicle crash) trauma patients received IV insertion attempts en route • Seventy percent were vehicle crash occupants • IV insertion success rate was 93% • Average EMS on-scene time was 13.9 minutes.

21 On-scene time for the en route insertion is statistically less (P < 0.001), and successful insertion percentage for en route insertion is statistically more (P < 0.05). The Rural and Frontier EMS Agenda for the Future: A Service Chief’s Guide to Creating Community Support of Excellence in EMS guidebook (HRSA 2007) was published by the U.S. Department of Health and Human Services, Health Resources and Services Administration, and Office of Rural Health Policy. It was written to give recommendations to rural EMS agencies. It recommends a community assessment and planning effort, in which the EMS agency involves stakehold- ers from the community in the EMS planning process. The planning process should determine the EMS system that the community will support. Committed community stakeholders will thus provide for continued funding of the planned EMS. The EMS should integrate with other “sectors” of the community including public safety, local health care sys- tems, and public health. The integration across the entire range of care may deem the EMS eligible for “pay for perfor- mance” systems set up by federal agencies. The implementa- tion of these “pay for performance” policies will reimburse organizations based on performance and the organization of all “sectors” will increase performance. Another way to win these “pay for performance” reim- bursements is to implement quality improvement systems they may someday be required by regulators. The qual- ity improvement system depends on the agency’s ability to gather and analyze data and to use the data to improve the system performance. Specific tasks that the guidebook recommends to achieve community support and a quality improvement system include the following: • Locating data collection resources that may be avail- able to the EMS. • Building relationships with local schools to develop a data gathering system. • Developing a “quality improvement” team to identify areas of potential improvement. The National Association of State Emergency Medical Services Officials prepared a report, Emergency Medical Ser- vices: Considerations for “Toward Zero Deaths: A National Strategy on Highway Safety” (NASEMSO 2010), highlight- ing many strategies critical to improving EMS response and efficiency. According to the report, the fourth leading cause of nonfatal injuries treated in emergency departments nation- ally is trauma sustained by vehicle occupants during crashes, with more than 2.6 million patients seen in emergency depart- ments each year. Through research, the CDC concluded that severely injured patients receiving care at a Level I trauma center had a 25% reduction in risk of death. The CDC also emphasizes access to trauma facilities as important to out- comes and provides mapping of 1-hour travel times to trauma care for the entire United States. However, there is no specific research cited to tie a 1-hour arrival time to patient outcomes. In response to these statistics, the U.S.DOT has looked to implement a comprehensive EMS system nationwide through the Office of EMS under Traffic Injury Control in NHTSA. However, NHTSA does not directly oversee the approximately 15,000 local EMS agencies and 757,000 EMS personnel supported by states and territories. With the bur- den of reducing losses falling on EMS, public health, and trauma systems, there is no evidence indicating that all means of reducing losses in the post-crash phase have been exhausted. The report looks at several categories, highlight- ing key strategies to improve on those already existing. • Detection and notification systems • 911 access and capabilities • EMS response and capacity • On-scene medical care • Patient transportation paradigms • Definitive care: hospital and specialty care infrastructure • EMS data, registries, and records linkage. According to the report, it is possible to transmit data describing crash severity through existing telematics data definitions and transmission standards, but no standard data dictionary and .xml schema exists for use by telematics device manufacturers. AACN could utilize an urgency algo- rithm to determine the probability of severe injury through vehicle telematics data. AACN could also predict the need for vehicle extrication. Enhanced 911 and Phase II compliance help dispatchers in locating a caller’s address or location within 300 meters, which becomes crucial in motor vehicle crashes. Next-gen- eration 911 would help 911 centers receive and process data other than audio (e.g., text messages, images, video). The report discusses several standards and initiatives related to EMS response and capacity. These include the National EMS Scope of Practice Model and National EMS Education Standards, Vehicle Extrication Education and Competency Standards, Regionalization of Emergency Care, Integrated Ambulance-Based Safety Systems, IntelliDrive for Emergency Response Vehicles, and Evidence-based Emer- gency Vehicle Operations Standards. IntelliDrive for Emer- gency Response Vehicles currently only incorporates flashing lights and sirens as a “vehicle-to-human” means of communi- cating presence. “Vehicle-to-vehicle” interaction would assist drivers of other vehicles in the vicinity of ambulance in mov- ing out of needed lanes. “Vehicle-to-infrastructure” interac- tion is in use, with an example being signal priority. Topics discussed within the On-Scene Medical Care cat- egory include Field Triage Decision Scheme: The National

22 Trauma Triage Protocol and the National Unified Goal for Traffic Incident Management. Patient Transportation Para- digms summarizes Engineering and Design Standards for Ambulances, Helicopter EMS Utilization Criteria, and Ground Ambulance Access to Intelligent Transportation Systems (ITS) Infrastructure. EMS often are taken out of the range of their home 911 system, leaving them vulnerable to the absence of information about road hazards, closures, or changing weather conditions. Ground Ambulance Access to ITS Infrastructure could benefit the crew and patient’s safety, route planning, and resource utilization. The report includes statistics regarding hospital and specialty care infrastructure in relation to trauma systems and prehospital and interfacility telemedicine applications. Based on CDC maps, 8% of land and 57% of the population are within a 1-hour ambulance drive of a Level I or Level II trauma center. In contrast, 29% of land and 83% of the popu- lation are within a 1-hour trip by helicopter. This still leaves thousands of miles of roadways outside the range of Level I or Level II trauma centers regardless of access to helicopters. The report also discusses EMS data, registries, and records linkage. NEMSIS is the nationally recognized EMS data repository that will be used to store EMS data from every state in the nation. Trauma registries may be used as reliable sources for severe injury data, in combination with records linkage, which makes other health care databases available. The Iowa DOT’s “Golden Hour” Project: Recommenda- tions for Reducing the Crash to Care Time (SRF Consulting Group 2010) explored innovative uses of ITS to benefit the medical community in general, and emergency responders specifically. Recommendations for the DOT were developed from two primary sources for information: a literature search of sources relating to trauma care, and a series of interviews with experts working in trauma care in Iowa. Traffic crash fatalities in the United States have histori- cally been a rural phenomenon. Of the more than 40,000 annual traffic crash fatalities, more than 55% result from injuries incurred in rural crashes, even though only 20.8% of the population lives in rural areas (U.S. Census 2000). Of the 450 Iowa traffic fatalities in 2005, approximately 88% resulted from rural crashes (Iowa DOT). The three primary factors for the disproportionate number of rural traffic fatali- ties were roadway facility type, automobile speeds, and dis- tance and time to a trauma care facility. Several themes emerged from the interviews, which pro- vided the basis for recommendations. • Theme 1: Improved Road Condition Information. Nearly all formal and informal interviewees expressed a desire for additional road condition information, including weather, maintenance, and construction details. There was also interest in real-time snow maintenance data to indicate which roadways had been plowed and how much time had elapsed. • Theme 2: Communications Infrastructure. The popu- larity of telemedicine and remote consultations is increasing, and as a result, some hospitals have been purchasing fiber-optic capacity from cable owners to minimize recurring costs and ensure connection qual- ity. The value of remote video surveillance of high- accident rate rural intersections was highlighted by rural PSAP dispatchers, which would require a com- munications system to support the bandwidth needs of digital video. • Theme 3: Availability of Volunteers. Those who expressed this concern saw it as the largest issue faced by rural emergency medical response. • Theme 4: Ambulance/Equipment Management. Several interviewees expressed concerns about the maintenance and condition of vehicles and equipment, particularly in cases where there was no permanent maintenance or storage facility available for ambulances. Several themes also emerged that, while not directly related to the core focus of the study, were seen as important to rural crash response, including the following: • Differences in diagnostic capability, record-keeping conventions, and time pressures deprive the definitive care team of vital information. • At-grade rail crossings can cause delays, and a system that automatically determines when a key intersection is blocked by a train along a projected response route could provide real-time notification to a dispatcher, who could then assess and relay information to a vehi- cle driver as needed. The report recommended next steps in several “Action Areas” for follow-on work by the DOT in cooperation with EMS and trauma care providers: • DOT operations representation on trauma board • Expansion of travel data to local jurisdictions • Improved user interface for emergency responders • Sharing of real-time road maintenance data • Coordination of communications infrastructure • Certification program for DOT personnel as EMTs • Co-location of emergency vehicles at state mainte- nance garages • Real-time rail crossing status. On-Scene and Transport Issues The FHWA Office of Operations published a report titled Best Practices in Traffic Incident Management (Carson 2010), based in part on an international scanning tour spon-

23 sored by FHWA, AASHTO, and NCHRP. The scanning tour focused on Traffic Incident Management (TIM) plan- ning and training, on-scene operations, technology use, and program management and administration. This report encompassed a wide variety of traffic incident management practices and pertains to all agencies that are involved in incident response, including first responders, law enforce- ment, tow and recovery vehicles, highway departments, and EMS. It is organized into four chapters and covers the fol- lowing topics: • Incident detection and verification • Traveler information • Incident response • Scene management and traffic control • Quick clearance and recovery. The report cited the following factors that can lead to the slow or inaccurate detection of traffic incidents: inaccurate report of crash location, dispatch overload, and slow detec- tion (particularly in rural areas). Some tools and strategies that can improve the accuracy and timeliness of incident detection include use of closed-circuit television cameras (CCTV), enhanced roadway reference markers, enhanced 911/automated positioning systems (i.e., next-generation 911), motorist call boxes, and AACN. It also indicated how the degree of emergency response can be either less or more than required, termed “under response” and “over response.” Early and accurate assessment of the scene con- ditions is critical to establishing the appropriate resource needs. The use of CCTV to verify incident severity is cited as a solution. The report identified various communication strategies that can enhance the emergency response process: • Common mutual-aid frequency/channel • Alternative communications devices • Wireless information networks • Mobile unified communications vehicle • Standardized communications terminology/protocol. Several technologies that can assist in timely and accu- rate emergency response were also identified. These include enhanced computer-aided dispatch systems and automatic vehicle location technologies that are used to locate, dispatch, and route emergency vehicles closest to the incident scene to minimize response time. In addition, “dual dispatch” is used in some areas; this involves dispatching emergency response vehicles from two different locations with the first to arrive providing aid and the second unit to return to base. The report identified several areas that contribute to successful scene management. Those relating to this study include the use of emergency scene lighting, which can reduce the time EMS personnel spend on sight, and adher- ence to incident parking plans that provide for a quick in/ out movement of EMS vehicles at the scene of the incident. Underlying all of these issues is the value of engaging stakeholders in incident response planning. Bringing staff together from first responders, law enforcement, tow and recovery vehicles, highway departments, and EMS on a regular basis helps to establish common objectives and a better understanding of other agency needs and objectives. The stakeholder meetings also provide an opportunity for debriefing major incidents, conducting traffic incident man- agement training, and developing traffic incident manage- ment guidelines. Training can be done at several different levels: training specific to traffic incident management within one’s own agency or company, training aimed at increasing awareness of other responders’ roles or exis- tence, and training aimed at improving specific procedural operations. The report lists several training resources and communities of practice that are available. Another practice that fosters interagency cooperation is the creation of a joint emergency/traffic operations center to dispatch and monitor incidents through a common facility. Training, or more specifically continued training, plays a critical role in innovation and improvement. Pro- moted in National Traffic Incident Management Leader- ship & Innovation: Roadmap for Success (Allwell et al. 2012), the second Strategic Highway Research Program (SHRP 2) has developed two products hoping to improve on-scene traffic incident management. These products, a multidisciplinary training course that provides a bet- ter understanding of quick clearance requirements and a 2-day train-the-trainer course facilitating use of the mul- tidisciplinary training, intend to save time, money, and lives on the nation’s highways. The multiagency National Traffic Incident Management Training Program hopes to strengthen traffic incident management programs in responder safety, quick clearance, and communications. Some of the topics covered in these courses include the following: • Statistics, terminology, and standards • Scene and responder safety • Incident notification and response • Arrival at the scene • Initial size-up of the scene • Command responsibility • Traffic management • Situational awareness. EMS Service Challenges A number of challenges in the provision of EMS are sum- marized in Emergency Medical Services in Rural America (Goodwin 2007) a report prepared for the National Confer- ence of State Legislatures. The report organizes issues faced

24 safety, and hospital discharge to help communities develop injury prevention programs. • Institutional-level changes. Develop practice manage- ment guidelines and establish trauma centers staffed with appropriately trained personnel. In response to motor coach incidents resulting in a high number of casualties and injuries, the NTSB made several recommendations including one to the Federal Interagency Committee on Emergency Medical Services in its report Rural Highway Mass Casualty Guidelines (NASEMSO 2011). These recommendations, based on the concept that emergency care systems may not be sufficiently prepared for motor coach crashes in rural areas, would result in the development of the following rural highway mass casualty guidelines: Guideline 1: Evaluation of EMS System Readiness— The EMS Incident Response and Readiness Assessment (EIRRA) is designed to evaluate state, regional, and local EMS agencies’ ability to respond to large emergencies. Highway maintenance and operations, law enforcement, fire/rescue, and emergency management personnel may also be evaluated by the EIRRA along with the EMS agen- cies. The EIRRA specifically evaluates these systems and agencies through benchmarks, indicators, and scoring. Table 10 shows an example of a scoring system for human resources availability: Guideline 2: Prepare to Quantify Resources on a Geo- graphic Basis—Another tool is the Model Inventory of Emergency Care Elements, which provides a visual display of resources and capacity by roadway segment. A future capability of this tool would be to provide highway offi- cials, EMS officials, route planners, and the public with a regularly updated map showing segment capabilities for response. To determine the capability of each segment, six categories would be evaluated: personnel, transportation, communications, equipment/inventory, medical facilities, and other. Guideline 3: Engage and Educate Partners—Recom- mendations for EMS officials to take leadership position in using EIRRA as well as the other tools and resources avail- able on both the state and local levels. By sharing knowl- edge and critical information, improvements can be made to reduce the number of rural highway mass casualty incidents. The 2011 CDC report Guidelines for Field Triage of Injured Patients describes revisions to the guidelines and reasoning for any changes. EMS providers must make deci- sions about the appropriate care and destination for injure patients on a daily basis. These guidelines are intended for use with individual injured patients and provide direction for EMS providers caring and transporting these patients. Published peer-reviewed research was the primary basis for any revisions made to the guidelines. by rural EMS providers into three categories: recruitment/ retention, reimbursement, and restructuring. Recruitment/Retention The current reliance on volunteers for rural EMS (80% of total workers) is cause for concern, given the demographic and economic realities facing rural America. Several states have initiated programs to improve the quality of care in rural areas, including the following: • Financial incentives to attract and retain personnel. • Financial assistance for items such as malpractice insurance. • Use of EMS personnel in expanded health care roles (e.g., immunizations, primary care) • Use technology (such as distance learning) to enhance training opportunities. Reimbursement Fixed costs in rural areas are comparable to urban centers, but patient volumes are much lower, resulting in higher per-patient costs. Potential solutions identified in the report include the following: • Use providers to deliver preventive care, provide public health services, or work in emergency rooms. Currently, no mechanism exists for reimbursement; therefore, no incentive exists to use these providers most effectively and provide reimbursement for their services. • Regionalize or consolidate administrative services to lower per-unit costs. • Enhance community awareness of EMS structures and operations so that they can make informed deci- sions about the type of investment they want to make in their systems. • Enhance access to capital. Although equipment and technology are costly, they are critical to improve quality. EMS leaders need to participate in federal and state dia- logue and planning for health information technology. Restructuring EMS systems have developed independently and, as a result, are organized as “silos” of functionality, rather than an inte- grated system of service providers. The following strategies for establishing a more efficient and cohesive system were summarized in the report: • Investigate cohesive and integrated systems, such as in Hawaii, where the state system is responsible for arranging personnel, facilities, and equipment in the prehospital setting. The system includes injury preven- tion and public education within the EMS system and combines data collection from EMS injuries, highway

25 Statistics regarding injuries and trauma centers across the United States show— • Approximately 30 million injuries were serious enough to warrant visit to a hospital emergency department. • Of these injured patients, 5.4 million (18%) were trans- ferred by EMS personnel. • National Study on the Costs and Outcomes of Trauma identified a 25% reduction in mortality for severely injured patients receiving care at a Level I trauma center. • Only 28% of U.S. residents have access to specialized trauma centers within an hour by helicopter. The report indicated that, ideally, patients with severe or life-threatening injuries would receive care at Level I or Level II trauma centers, with less serious injuries being handled by lower-level trauma centers or emergency depart- ments. However, complexities of patient assessment in the field can affect triage decisions. Existing triage studies use retrospective data, trauma registry samples, single EMS agencies, and single trauma centers. Future research needs to include multiple sites, agencies, and centers to reduce bias and increase generalizability. Field triage in rural settings could be better understood through further research in— • Impact of geography on triage • Issues regarding proximity to trauma centers • Use of air medical services • Using local hospitals for initial stabilization • Secondary triage at nontrauma hospitals. Improved field triage of injured patients can benefit EMS and trauma systems by reducing costs associated with trauma care, and increasing the care provided to the millions injured every year. Planning and Innovation NCHRP has assembled an extensive report to provide guid- ance for implementing AASHTO’s SHSP. Strategies related to EMS improvements are covered in NCHRP Report 500: Guidance for Implementation of AASHTO Strategic High- way Safety Plan, Volume 15: A Guide for Enhancing Rural Emergency Medical Services (Torbic et al. 2005). The document outlines the challenges facing rural EMS crash response, including: • Importance of timely medical care in minimizing fatalities and the long-term effects of injury. • Increased response times in rural areas, owing to delayed incident reporting and longer response times. • Low volume of EMS calls in rural areas makes it dif- ficult to develop adequate financial support, requiring the use of volunteers (65% to 75% of rural EMS per- sonnel are volunteers). • EMS personnel in rural areas often have less experi- ence because of lower call volume. • Growing recruitment and retention challenges of vol- unteer-based EMS agencies. • Lack of comprehensive statewide trauma legislation in several states. The report documented the rates of crashes per 100 million vehicle-miles traveled in rural areas as more than double that in urban areas (2.01 vs. 0.89). The guid- ance document then defines four objectives related to EMS enhancement: TABLE 10 SCORING CRITERIA FOR HUMAN RESOURCE AVAILABILITY Source: NASEMSO (2011).

26 • Integrate services to enhance emergency medical capabilities • Provide or improve management and decision-making tools • Provide better education opportunities for rural EMS • Reduce time from injury to appropriate definitive care. These objectives were then supported by 24 individual strategies, categorized as tried (implemented, but not fully evaluated), experimental (suggested and thought to be fea- sible), and proven (used in one or more locations and eval- uated to be effective). A few of the relevant strategies are listed here. The abbreviations T, E, and P refer to the catego- ries tried, experimental and proven: • 20.1 A1—Establish programs with organizations to utilize nontraditional employees as EMS personnel (T) • 20.1 A4—Integrate information systems and highway safety activities (T) • 20.1 A6—Use mobile data technologies that are interoperable with hospital systems (T) • 20.1 A7—Require all communication systems to be interoperable with surrounding and state jurisdictions (T) • 20.1 C4—Require first care training for all public safety emergency response personnel, including law enforcement officers (T) • 20.1 C6—Provide “bystander care” training pro- grams targeting new drivers, rural residents, truck drivers, interstate commercial bus drivers, and motorcyclists (T) • 20.1 C7—Provide EMS training programs in high schools in rural areas (T) • 20.1 D1—Improve cellular telephone coverage in rural areas (T) • 20.1 D2—Improve compliance of rural 911 centers with FCC wireless “Phase II” automatic location capability (T) • 20.1 D3—Utilize GPS technology to improve response time (T) • 20.1 D4—Integrate AVL and computer-aided naviga- tion technologies into all CAD systems (T) • 20.1 D5—Equip EMS vehicles with multiservice and/ or satellite-capable telephones (T). The report also recommended a number of related actions that, while not directly part of EMS operations, should be con- sidered as part of an overall safety program. These included: • Public information and education programs • Enforcement of traffic laws • Strategies directed at improving the safety manage- ment system • Implementation of other strategies developed for other sections of the overall guidance for implementing SHSPs To facilitate the pursuit of the objectives and assist with the implementation of specific strategies, the guidance document also defined an 11-step plan. This plan covered all aspects of implementation, from the initial definition of problem to performance assessment and transition from an implementation posture to a standard operating procedure.