National Academies Press: OpenBook

Highway Safety Behavioral Strategies for Rural Areas (2023)

Chapter: Literature Review (Task 3)

« Previous: Data Analysis (Task 2)
Page 48
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 48
Page 49
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 49
Page 50
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 50
Page 51
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 51
Page 52
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 52
Page 53
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 53
Page 54
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 54
Page 55
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 55
Page 56
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 56
Page 57
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 57
Page 58
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 58
Page 59
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 59
Page 60
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 60
Page 61
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 61
Page 62
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 62
Page 63
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 63
Page 64
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 64
Page 65
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 65
Page 66
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 66
Page 67
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 67
Page 68
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 68
Page 69
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 69
Page 70
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 70
Page 71
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 71
Page 72
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 72
Page 73
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 73
Page 74
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 74
Page 75
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 75
Page 76
Suggested Citation:"Literature Review (Task 3)." National Academies of Sciences, Engineering, and Medicine. 2023. Highway Safety Behavioral Strategies for Rural Areas. Washington, DC: The National Academies Press. doi: 10.17226/27196.
×
Page 76

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.

48 Literature Review (Task 3) To further support the creation of a toolkit (i.e., countermeasures and barriers to be included), the research team conducted a literature review that focused on behavioral interventions in a rural environment. The objective of this review was to identify and describe evaluation studies of prevention and intervention programs that have a behavioral approach to reducing rural roadway traffic injuries. Methodology This methodology focused on the primary task of identifying behavioral strategies in rural environments. However, a preliminary literature review was also conducted using Transportation Research International Documentation (TRID) and agency sources (e.g., NHTSA, FHWA, National Safety Council (NSC), etc.) to support the background sections on rural transportation safety, roadway characteristics, risk factors and traffic safety culture. Search Methods The literature review encompassed numerous national and international databases and resources. This review followed guidelines for Cochrane Reviews (https://www.cochranelibrary.com/), which is a structured approach to collecting and synthesizing results of evaluation studies with the intention of informing implementation, dissemination, and translation of evidence-based health interventions. Trial protocols testing behavioral strategies in rural environments were searched through the National Institute of Child Health and Human Development (NICHD) Clinical Trials Registry, but none were found. Requests for reports on rural roadway evaluations were requested through the Safe States Alliance, which is the professional organization for state public health department injury programs. No programs with evaluations were identified. The search for published studies included health-related literature from PubMed and Embase. The search terms included: rural, roadway, safety, injury, crash, and transportation, including singular and plural versions, with evaluation studies as the journal article type. Combinations of these terms were used in inclusive (or) and exclusive (and) versions. The search was not restricted by publication status (e.g., electronic) or language, although non-English articles were only included if an English version of at least the abstract was available. Selection Criteria The search included articles from 2005 through 2020. Studies that measured the effects of an intervention strategy focused on improving roadway safety, with an emphasis on prevention, were included. Studies that had one or more elements of behavioral interventions or measures were also included. Examples of behavioral outcome measures included crashes, driving behaviors (e.g., speeding), surveys of knowledge/attitude/behavior, and observational surveys of behavior (e.g., seatbelt use observations or speed monitoring). Evaluations that collected data measuring the process, impact, or outcome of the intervention were included; studies that described the intervention but did not measure it were excluded. Studies were not selected based on the design of the evaluation, so all study designs were included. The selected papers had

49 to include an intervention that impacted rural roads or drivers on rural roads. Rural was self-identified by the article; the study was selected if it reported focusing on rural factors. (Various definitions of rural were used in the studies, and in some cases rural was not defined.) Interventions that addressed acute care/tertiary prevention were not included. Interventions focused on preventing tractor rollovers were excluded because the majority of rollovers do not occur on the roadway. Data Collection and Analysis Articles were reviewed by the research team to identify eligibility. Members of the research team recorded relevant information in an Excel spreadsheet. Following Cochrane guidelines for summarizing interventions, team members collected specific information for each article: source of article, citation, year published, purpose of study, study population, sample size, rural definition, study groups, study design, recruitment, measures, intervention, results, and limitations. Transportation Safety Overview Recent Changes in Transportation Safety Approaches In 2018, there were more than 36,000 motor vehicle traffic fatalities in the U.S. (NHTSA, 2020) and motor vehicle fatalities were listed within the top three leading causes of death for all age groups (Centers for Disease Control and Prevention, 2021). Because all road users, regardless of their mode or location, should make it to their destination safely, transportation practitioners have created goals and initiatives aimed at reducing fatalities and serious injuries to zero. The goal of zero was first adopted in Sweden as Vision Zero, with the acknowledgement that even one death on the transportation system is unacceptable and preventable. Several complementary U.S. safety initiatives have been created with the same goal. These include Vision Zero, Toward Zero Deaths, and the Road to Zero. The Vision Zero Network is a campaign for “advancing this game-changing shift toward safe, healthy, equitable mobility for all (Fox and Shahum, 2018).” While the Vision Zero Network started as a way to assist early adopter cities, it has expanded its reach to include all local communities. The Vision Zero Network has most notably provided a framework for shifting the thinking around transportation safety from a “traditional approach” to an approach aimed at reaching zero (see Figure 11). Toward Zero Deaths (TZD) is a national strategy led by the American Association of State Highway and Transportation Officials (AASHTO). “The strategy calls for all stakeholders to champion the idea that one death on our nation’s roadways is too many, and we must all work together to bring the annual number of roadway deaths down to zero.” The TZD coalition has created a framework for state and local agencies to implement this strategy into their statewide Strategic Highway Safety Plans (SHSPs).

50 Figure 11: Vision Zero Approach (Fox and Shahum 2018) The Road to Zero (RTZ) Coalition is a coalition of national stakeholders, managed by NSC, which is working towards three main initiatives: “double down on what works through proven, evidence-based strategies, advance life-saving technology in vehicles and infrastructure, and prioritize safety by adopting a Safe System Approach and creating a positive safety culture” (Rand Corporation, 2018). RTZ has also created a new webpage to help stakeholders double down on what works (NSC, 2021). Traditional safety approaches acknowledge that road safety is a complex issue where contributing risk factors affecting crashes are attributed to three components: roadway and roadside features, vehicle factors, and human factors (Treat et al., 1979). The traditional safety approaches also assign 94% of the crash cause to human error (Treat et al., 1979) and aim to change this human behavior. However, along with the common goal of zero fatalities and serious injuries, these three initiatives have shifted toward a new approach called the Safe System, which is “built on the foundation that humans make mistakes, and those mistakes should not lead to death” (FHWA, 2020). As stated in newly released literature from FHWA, “Zero is our goal. A safe system is how we get there” (FHWA, 2020). The Safe System Approach is a holistic approach built on five elements of protection or aspects of safety that include safe road users, safe vehicles, safe speeds, safe roads, and post-crash care. In addition, the Safe System Approach is based on six principles: deaths and injuries are unacceptable, humans make mistakes, humans are vulnerable, responsibility is shared, safety is proactive, and redundancy is crucial. Key to this approach is that, rather than solely attempting to modify human behavior, the design and operation of a roadway should anticipate human mistakes and lessen the impact forces on the body to a level that does not result in a fatality or serious injury should a crash occur (FHWA, 2020). In addition to changing how the transportation sector approaches human behaviors and errors through the Safe System Approach, the zero fatalities initiatives also support growing traffic safety culture. Traffic safety culture is defined as “the shared belief system of a group of people, which influences road user behaviors and stakeholder actions that impact traffic safety” (Ward et al., 2019). A positive traffic safety culture is one that supports prioritizing safety and the incorporation of safety into all aspects of transportation. Traffic safety culture can be categorized into organizational (i.e., looking inward at the safety priorities for improving internal safety within an organization) and programmatic (i.e., looking externally at safety priorities aimed outside an organization). Key to impacting behaviors is identifying and “growing beliefs among road users and stakeholders that increase protective behaviors and decrease risky behaviors” (Otto et al., 2020). Human beliefs are generally

51 shared within our social environment, which includes multiple layers such as: individual, family, friends, workplace, school, community, organizations, and local government (Ward et al., 2019). Therefore, it is necessary to assess these values and beliefs in order to influence them. New research and guidance is emerging on assessing and growing traffic safety culture in the United States. An ongoing project sponsored by FHWA is creating an assessment tool for agencies to measure their level of traffic safety culture, both internally and externally, to identify potential areas of growth. The Traffic Safety Primer provides a seven- step process that can grow positive safety culture as shown in Figure 12 (Ward et al., 2019). Proactive Traffic Safety Communication tools are available through the Traffic Safety Culture Pooled Fund including a primer (referenced above), PPT presentation, conversation guide and poster. Figure 12: Steps to Grow Positive Traffic Safety Culture (Ward et al. 2019) Traffic safety culture awareness has also impacted how practitioners “frame the message” disseminated during public awareness campaigns. Traditional awareness campaigns used to provide statistics around the number of people not participating in the desired behavior (e.g., x percent not wearing their seatbelt); however, newer campaigns focus on proactive traffic safety culture (changing values and beliefs) which “shift our focus to the engagement of the majority of safe road users to influence the behaviors of the smaller group engaging in risky behaviors” (Finley et al. 2019). In other words, this approach highlights the large percentage of people doing the positive or safer behavior (e.g., the Most of Us campaign). Rural Transportation Safety There are more than four million miles of roads in the U.S., and rural roads comprise 70% of them. With only 19% of people living in rural areas and 46% of fatal crashes occurring on rural roads, the risk of fatal or serious injury is twice as high on rural than on urban roads (NHTSA, 2020). To apply the Safe System Approach to rural areas, it is necessary to understand the needs, challenges, and barriers of those areas. When looking at crash numbers, there are some notable statistics that show over-representation of certain crash characteristics on rural roadways. While newer statistics in NHTSA’s Rural-Urban Comparison indicate that the divide between rural and urban areas may no longer exist, the data reveals that there are still disparities when looking at overall proportions of speeding fatalities and seatbelt use. For instance, “the majority of rural fatal crashes (72%) occurred on roads where the posted speed limit was 55 mph or higher” and that “60 percent of rural pickup occupants killed were unrestrained – the highest percentage of any passenger vehicle occupants killed Step 1: Plan & Advocate Step 2: Assess Step 3: Prioritize Step 4: Identify Strategies Step 5: Pilot & Refine Step 6: Implement Step 7: Evaluate

52 among both rural and urban areas” (NHTSA, 2020). In addition, most head-on crashes on rural, non- Interstate roads are caused by a motorist making an unintentional maneuver as a result of driver fatigue, being distracted or driving too fast in a curve (TRIP, 2020). Rural roads are unique for many reasons, some of which include varying terrain and road conditions, wildlife on the roadways, and longer distances. In addition to differences in roadway characteristics, there are also differences in rural road users, culture, and barriers. Research (detailed in the following sections) indicates that there are often more unfamiliar road users due to tourism, differing road user risk perception, and a differing rural culture (e.g., values and beliefs). These pose challenges for rural transportation practitioners who tend to have less resources and funding, but more lane-miles to manage. In striving for a goal of zero deaths and fatalities, improving safety in rural areas will also need to be prioritized. “Based on the crash risk on rural roads, it will be impossible to achieve the goal of zero fatalities without focusing more attention on rural areas. In this case, the consequences of no action in rural areas have a negative impact on everyone, not just rural communities” (Sullivan, 2017). Rural Roadway Characteristics Rural roadways have characteristics that are different from their urban counterparts. Rural roads are more likely to have two-lane routes (86% of rural non-freeway arterial roads have two-lanes [TRIP, 2020]), narrower lanes (“23% of rural collector and arterial roads have lane widths of 10 feet or less” [TRIP 2020)]), longer detours (three times longer in rural areas due to closed or posted bridges ([USDOT 2020b]), more miles of unpaved roadways (roughly 45%of rural road miles are unpaved [Congressional Research Service 2018]), and a larger number of uncontrolled intersections. In addition, there are common road features and design elements that are more likely to be found in rural areas that factor into a reduction in safety including high speeds, no medians, sharp curves, limited shoulders, side ditches, pavement drop-offs, steep slopes, and roadside obstructions/limited clear zones (Congressional Research Service, 2018; TRIP, 2020). As noted in the Safe System Approach, safe roads and safe speeds are a part of the equation to achieve zero deaths. Unlike the traditional approach to safety which aimed to decrease the total number of crashes, a Safe System Approach specifically aims to decrease the number of fatalities and serious injuries. This approach acknowledges that humans make mistakes and therefore shifts the focus to keeping the crash impact on a human body at a tolerable level, rather than solely on eliminating the crash all together. This means that a countermeasure that will decrease the number of fatal and serious injury crashes should be considered even if it may potentially increase the number of total crashes (e.g., a two or more lane roundabout). Additionally, speed has a direct link to the impact on a human body and the ability to survive a crash and can be the difference between fatal and nonfatal crashes. However, the Safe System Approach is not about universally reducing speeds, but rather matching the appropriate speed to the conditions that exist. This includes designing roads and applying countermeasures that meet user expectancies, creating self-explaining roads (e.g., roads which have consistent design, do not have mixed messages, and elicit road user behavior based on design), and creating self-enforcing roads (e.g., provide visual design cues that match the speed limit). Inconsistent design or lack of design, which is found on a great many rural roads, can be deadly. Therefore, design improvements and countermeasures to keep vehicles from leaving their travel lane (e.g., centerline rumble strips) and reducing the effects in the event they do, are an important consideration when looking at behavioral countermeasures or interventions. Rural Road Users and Culture Rural road users include a population that is slightly older, average age of 49 years (TRIP, 2020) and a larger portion of tourists, as 44% of personal VMT are urban residents traveling to a destination outside

53 their home (USDOT, 2020b). Rural road users also include a higher percent of freight (47% of truck VMT and origination of two-thirds of rail freight [USDOT, 2020b]) and slow-moving vehicle travel (i.e., agricultural vehicles and horse and buggies). In addition to a different mix of rural road users, research indicates that the culture of rural road users includes differing values and beliefs than their urban counterparts, which can lead to behavioral actions factoring into crashes. A study by Rakauskas et al. in 2007 found that rural drivers engage in riskier behavior because they have lower perceptions of risk associated with such behaviors. Moreover, rural road users may require behavioral countermeasures guided by local cultural factors in order to be more effective. Safe road users are another part of the equation to achieve zero fatalities, therefore, additional information about rural behaviors and culture is examined in the Rural Roadway Risk and Protective Factors section. Rural Road Barriers Barriers for addressing safety in rural areas include, but are not limited to, lack of hot spot crash locations (e.g., a concentration of crashes in one particular area), funding, staff resources, and technology/communications infrastructure. In addition, longer emergency response times and longer distances to medical services have a detrimental effect on fatality rates in rural areas where medical attention in the golden hour (i.e., the first hour after a crash) is critical. Unlike their urban counterparts, crash locations in rural areas are not typically in hot spot locations. Rural crash locations tend to be random, which makes their identification and elimination more difficult. Instead, identification of aggregated types of crashes are more useful. To better assist rural areas, the systemic safety analysis was created to identify roadway locations with the highest risk for severe crashes in an effort to systemically deploy low-cost countermeasures to locations matching these factors (FHWA, 2021). Although rural areas have a higher percentage of roadways and a larger rate of fatalities, funding from many sources is distributed based on population, of which rural has a lower percentage. “Annually, rural roads received an average of 37.4% of federally obligated highway funds during FY2009-FY2015. This is more than rural roads’ 29% share of VMT on Federal-Aid highways but less than the roughly 65% of Federal-Aid highway centerline miles that are classified as rural” (Congressional Research Service, 2018). While not providing additional funding to rural areas, the Rural Opportunities to Use Transportation for Economic Success (ROUTES) initiative was created to assist in addressing funding disparities in rural areas and provides resources including an applicant toolkit (USDOT, 2020b). Such that need far outweighs funding in rural areas, it is important to ensure that the chosen countermeasures or interventions are either low cost or have the greatest potential to be effective in reducing fatalities and serious injuries. NCHRP Report 622 was conducted in 2008 to determine the effectiveness of the countermeasure listed in the second edition of the Countermeasures That Work: A Highway Safety Countermeasure Guide for State Highway Safety Offices, as well as to create a framework for estimating the costs and benefits for future countermeasures (National Academies of Sciences, Engineering, and Medicine, 2008). Results from that study have since been incorporated into newer editions of the Countermeasures document, now in its ninth edition (Richard et al., 2018). In addition, the NCHRP study was the basis for a research project to create a six step Rural Safety Policy Improvement Index to quantitatively measure the impact of enacting and enforcing rural legislation-based improvement measures on a state basis. While the project results indicated that this framework would provide more beneficial results if completed using specific statewide calculations, it was recommended that a website be created providing cost-benefit information of behavioral safety countermeasures, similar to the infrastructure-related crash modification factors. The research also suggested that while rural application of these measures has challenges, due to over-representation of

54 certain crash characteristics in rural areas, these measures may have more impact than in urban areas if effectively implemented (Knapp, 2011). The lack of funding in rural areas also affects the number of staff resources available, with many staff performing multiple positions, volunteering, or servicing a vast coverage area. This can have an effect on availability of staff to analyze crash data or implement low-cost safety measures, and the ability to enforce behavioral legislation. The lack of staff resources can also affect an agency’s ability to apply for funding. For smaller agencies, the paperwork and documentation requirements for state and federal safety funding can be a large barrier and, in some cases, impact their decision as to whether or not to apply for funding. Technology could be a potential solution in rural areas to assist with the lack of staff resources and vast coverage areas. One particular example is automated technology for speed enforcement in rural areas which have safety benefits and can cover longer distances on the road network. Newer cameras can also detect mobile phone use and seatbelt non-use, though they have yet to be fully evaluated. However, lack of communication infrastructure, the public’s perception of these technologies, and legislation related to their use will also be a barrier for these options. Behavioral interventions for rural areas are not a “one size fits all.” The interventions must fit within the social and cultural differences of the rural community types and differences in road design characteristics as indicated in the rural road safety classification created in Task 1. Therefore, the section on Barriers for the Implementation of Behavioral Rural Traffic Safety Interventions continues this discussion. Rural Safety Coalitions Due to the lack of staffing and resources in rural areas, rural safety coalitions are one way to pool resources for a larger impact. These coalitions have the opportunity to overcome rural barriers, influence the growth of positive safety culture, and carry out the Safe System principle that safety is a shared responsibility. In addition, coalitions can be utilized to choose and implement best safety practices and interventions that fit within a particular community type’s social and cultural construct. Coalitions should be created to include a wide variety of stakeholders. Coalitions composed of “diverse representatives across the social environment” (i.e., among individuals, families, schools, workplaces, organizations, and community-level organizations like local government and law enforcement) tend to be the most effective (Otto et al., 2021). A forthcoming publication entitled Guidance on Growing Traffic Safety Culture: Stories from Rural Communities provides guidance on how to create a coalition which includes establishing a shared purpose, developing roles and expectations, and identifying stakeholder resources. The guide also highlights lessons learned to ensure the coalition remains healthy over time by “building the capacity of coalition members, developing a shared understanding of traffic safety within the community, prioritizing needs and strategies, piloting strategies, supporting broad implementation, and reflecting and evaluating” (Otto et al., 2021).

55 Rural Behavioral Surveillance and Protective Factors Surveillance Health surveillance is the systematic measurement of outcomes and exposures to document the incidence and burden of health outcomes. Surveillance data are an essential component of understanding a health problem, such as contributing causes for motor vehicle crashes, and for use in evaluation studies, such as the impact of policy or behavioral interventions on crash incidence. Considerable data to document the incidence of crashes and crash mechanisms for rural roadways exist. State crash data usually include a rural indicator, although states define rural roadways using different definitions (see Task 2). State data on roadway networks also provide variables to identify rural roads, which have designations such as two-lane highways or farm-to-market status. Census data provide detail about rurality at the state, county, and census tract level, enabling population-based analyses. Crash reports are an excellent source to document crash incidence and mechanisms, but they often have little and potentially inaccurate information about the role of safety behaviors and medical outcomes. Programs such as NHTSA’s FARS and Crash Outcomes Data Evaluation Systems (CODES) projects are surveillance efforts that link data to fill in some of these gaps, and in particular to provide more detail about medical outcomes. These surveillance sources have been instrumental, for example, in identifying vehicle design features as they related to injury severity and diagnoses and have had considerable impact on improving vehicle crash worthiness. Behavioral data are far more difficult to collect. Self-report is often unreliable because drivers have motivations not to report unsafe behaviors, particularly during crash investigations. For example, when describing events of a crash to a police officer, vehicle occupants may differentially report that they were wearing a seatbelt to avoid receiving a citation. Reporting of distractions such as cell phone/texting has similar challenges. Because surveillance data for behaviors are difficult to find and may be unreliable, studies evaluating behavioral strategies have an inherent challenge in availability of outcome measures. Over time, systematic collection of behavioral factors on crash reports has improved, and these advances will facilitate wider evaluations of behavioral strategies. However, the majority of evaluation studies will need to collect their own behavioral data or rely on information from police crash reports. This lack of data will shape the type of evaluation studies that are possible, and this proved to be a limitation in the studies reviewed. Many studies used crash incidence as the primary outcome measure, which in the context of evaluation logic models is a long-term outcome. Without the contextual contributions of the process, impact, and short-term outcomes, interpreting these findings is challenging. Studies may also suffer from being under-powered when the intervention impact is a behavior (and crash occurrence is a relatively rare outcome). Task 2 of this project examined crash data from several states to provide a comprehensive snapshot of rural roadway crash incidence, thus Task 3 did not explore this topic in detail. Since availability of accurate data is such an inherent component of building an intervention evidence base, it is relevant to mention here. As states and communities implement behavioral strategies to improve roadway safety, the need to identify evaluation data sources early will be important for ongoing development of the program and documentation of its impact.

56 Rural Roadway Risk and Protective Factors In the U.S., the overall death rate from traumatic injuries is higher in rural than urban populations, and much of the increase in rural injury death rates is related to motor vehicle crashes. As introduced in the rural transportation safety section, rural fatal crashes account for 46% of all traffic fatalities, but rural road miles only account for 30% of VMT (NHTSA, 2020). Crash-related mortality is inversely associated with population density and per capita income (Baker et al., 1987). Analyses of motor vehicle crashes in several states have shown that fatality rates in the most rural counties are almost double those observed in urban counties (Muelleman and Mueller, 1996; Maio et al., 1992). Many factors may contribute to the increased motor vehicle crash fatality rates observed in rural areas. For example, environmental factors may contribute to increased rates of crash occurrence both among rural residents and among urban residents driving on rural roads. Unlike Interstate roads, rural highways are typically composed of two narrow lanes and do not have crash reduction features such as divided traffic streams, controlled vehicle entrances and exits, graded curves, skid-reducing surfaces, large lane and median widths, and lighted traffic signs (Graham, 1993; Karlaftis and Golias, 2002). In comparison with urban primary roads, rural roads have fewer traffic control devices and speeds are often higher because of the presence of uninterrupted segments of roadway. While many factors contribute, speed is a major factor in serious and fatal injury crashes due to kinetic energy and the fact that humans are vulnerable. Therefore, it is called out separately in the Safe System Approach as one of the five elements (e.g., safe speeds) and should be given consideration especially in rural areas where speeds tend to be higher (FHWA, 2020). Certain types of crashes, such as those involving motor vehicle collisions with farm machinery, are unique to rural environments (Costello et al., 2003). These crashes most frequently involve slow-moving tractors and are exacerbated by the high speeds at which vehicles often travel on rural roadways. In addition, other types of crashes, such as those involving all-terrain vehicles (ATVs) and snowmobiles, are more common in rural areas. Environmental factors may also contribute to increased crash severity. Head-on collisions, which most frequently occur when traffic streams are not separated and which are the most likely of all crash types to cause fatality and severe injury, account for 17% of fatal rural crashes and 9% of fatal urban crashes (Baker et al., 1987). Behavioral factors may have differential contributions in rural and urban environments. While little research has examined differences in behavior and traffic safety culture by rurality, existing studies indicate that rural residents are less likely than urban residents to wear seat belts or to use child safety seats (Zwerling et al., 2001; Baker et al., 2000; Lundell, 2003; Rakauskas, et al., 2009) and they are more likely to have positive blood alcohol tests in crashes (Blatt and Furman, 1998). Rural safety culture may also influence the impact of safety policies. Enforcement of traffic safety laws, such as laws against drunk driving and speeding, may be limited in some rural areas because of the reduced presence of traffic enforcement officials. Rural enforcement may be challenged because rural residents report a higher distrust of the government (Rothe and Elgert, 2003) and a strong sense of autonomy. Rakauskas et al., 2009, compared aspects of safety culture in rural and urban residents and found that rural residents were far less likely to support safety policies or enforcement of these policies. However, residents of rural communities report a stronger sense of connection and shared responsibility for the community, as well as a strong desire to help neighbors in need. Rural safety culture is likely to be highly varied across the country and very specific to local populations. Task 1 of this project delineated the heterogeneity of rural environments, which applies to culture as well as roadways.

57 The Social-Ecological Model is an effective framework to examine behavioral contributions to health outcomes, and to help guide planning of behavioral interventions. The Social-Ecological Model, also called the Social-Ecological Perspective, is a framework which suggests that there are multiple levels of influence on specific health behaviors. While individuals are responsible for instituting and maintaining specific health behaviors necessary to reduce the risk of injury and violence and improve health, individual behavior is determined to a large extent by the social environment (e.g., community norms and values, regulations, and policies). Efforts to prevent injuries and violence are more likely to be successful when stakeholders work within multiple spheres of influence at the same time. The four-level social-ecological model includes the interplay between individual, relationship, community, and societal factors that put individuals at risk of injuries and violence (Centers for Disease Control and Prevention [CDC], 2009). The model can assist in better understanding the determinants of behaviors leading to injuries and violence (Dahlberg, & Krug, 2002). This knowledge is useful when designing, implementing, and evaluating injury and violence prevention interventions. Individual Level. The first level of the Social-Ecological Model identifies biological and personal factors that increase the risk of crashes, such as driving behavior (CDC, 2009). Some of these factors include age, experience, substance use, personality traits such as risk-taking, and the practice of safety behaviors such as safety belt use. Prevention strategies at this level are often designed to promote attitudes, beliefs, and behaviors that will ultimately prevent injury. These individual behaviors are influenced by norms and attitudes at the higher levels. Relationship Level. The second level examines how close relationships with family, friends, partners, and peers increase or decrease the risk of injury. Prevention strategies at this level may include mentoring and peer programs designed to improve safe driving or programs that focus on family-level normative behaviors. Increased parental supervision of novice driving is one example of a relationship-level intervention. Health care provider discussions with their patients on child safety seat use is another example. Community Level. The third level explores the settings, such as schools, workplaces, and neighborhoods, in which social relationships occur and seeks to identify the characteristics of these settings that are associated with traffic injury risk. Prevention strategies at this level are typically designed to impact the climate, processes, and policies in a given system. Social norms and social marketing campaigns are often used at the state level to influence safety behaviors. Other system-based approaches, such as workplaces that require employee seatbelt use, have much potential as well. Societal Level. The fourth level looks at the broad societal factors that help create a climate in which risk behaviors that lead to injuries and violence are discouraged. These factors include social and cultural norms, including the political safety culture. Other large societal factors include the health, economic, educational, and social policies that reduce economic or social inequalities between groups in society (CDC, 2009). Each of the four levels in the social-ecological model can be considered a point of influence. Road safety interventions should include a continuum of activities that address multiple levels of the social-ecological model. These activities should be developmentally appropriate and conducted across the lifespan. This multifaceted approach is much more effective at sustaining prevention efforts over time than any intervention that only focuses on a single sphere of influence.

58 Rural Behavioral Safety Intervention Strategies A total of 1,062 studies were screened, and 42 studies met intervention criteria. Over 95% of the articles that were excluded either did not measure the results of an intervention or did not measure any behavioral components. The 42 selected studies were categorized by type of intervention, type of study design, and behavioral strategy focus. Study Overview by Type of Intervention and Behavioral Strategy Focus Areas Roadway design. The most common type of intervention was a roadway design modification, which accounted for 14 (33.3%) of the 42 studies. Speed was the behavioral focus for 8 (57.1%) of the 14 roadway designs; three of the studies examined both speed and lane control, including one that measured lane placement while passing a bicycle on a two-lane road. One examined speed reduction on rural roads when approaching an intersection to assess readiness to yield to pedestrians. Loss-of-control/roadway departure behaviors were the focus of five (35.7%) of the roadway design studies. Two (4.8%) studies addressed vehicle design, including driving in low visibility conditions and distraction based on GPS screen size. Policies. Policies that impacted rural roadways were the focus of 12 (28.5%) of the 42 studies. The policies examined speed on rural roads (n = 5), motorcycle rider helmet use and general safety (n = 2), and one study each that evaluated graduated driver’s licensure, youth operators of tractors on roadways, ATV safety, Swedish National Vision Zero activities in rural communities, and lighting and marking policies to increase tractor visibility. Education. Eight (19.0%) of the studies evaluated educational interventions, and these focused on ATV safety (n = 2), novice drivers including rural roadway driving (n=3), youth bicycle riding safety, and general traffic safety. The educational program that addressed general traffic safety was a focused strategy for one high-risk rural roadway segment. The educational campaigns focused on individual education strategies, although the community-based educational campaign likely also had an impact at the community level. Campaigns. Six (14.3%) of the studies evaluated a component of a safety campaign. Campaigns addressed child bicycle helmet use, teen risky driving, rural road safety, youth seatbelt use, child occupant safety, and ATV safety. The campaigns focused on a range of audiences, including children (child occupants and bicycle helmet use), youth (novice driving, seatbelt use, ATV safety), rural businesses, and a rural community. The youth campaigns were predominantly set-in schools, although one had a community focus to increase helmet use and one focused on child occupant protection (OP) and was conducted in a clinic. Campaign partners included community organizations, businesses, schools/students, businesses, and local agencies. Study Overview by Intervention Type and Study Design Crash analyses. The most common study design was a crash analysis, used by 20 (47.6%) of the studies. All but two crash analysis studies used police-reported crashes. The availability and generally high quality of police-reported crash data enable efficient analyses. Investment in crash data infrastructures at the national level, as found in most high-income countries, is a critical component of roadway evaluation. Limitations of crash data, however, impact all of these studies. One limitation is the potential for missing records for crashes that are not reported, and these may be more likely in rural areas because individuals in crashes may not want to wait for the police to reach the scene. This limitation is most likely to affect vehicle

59 damage only crashes, and this limitation is not likely to have a large impact on evaluations that used primarily fatal and injury crashes. Another limitation is the potential for data to be miscoded, which has been reported for variables such as party (especially pedestrians and bicyclists), injury severity, and alcohol use. Police crash reports are also challenged in reporting behavioral contributing causes to crashes (this is discussed more in the Study Designs and Quality of Studies section). The most frequent analytic approach using crash data involved examining crash trends before and after a road design change. Most of these studies involved a pre-post analysis of crash trends with intervention and control road segments using Bayesian models to predict impact. The second most common use of crash analyses was to evaluate policies over time. Policy analyses used primarily a pre-post design, and several used comparison groups. Policies evaluated included graduated driver’s licensure, tractor and farm equipment safety, speed control, and motorcycle helmets. In order to collect information about behavioral changes, four of the crash analysis studies also conducted observations to validate actual behavior change. These studies examined helmet use and speeding in response to three policy interventions (a motorcycle helmet law, two studies on speed limit modifications, and one study of speed enforcement). The observational component adds considerably to the quality of the study because the logic model elements of impact and outcome are available. The crash analysis studies that did not rely on police crash reports involved medical records and self- report. Medical records are a good method of collecting detailed data on medical outcomes and some behavioral characteristics, such as alcohol use through blood tests. Some of these data may make it back to police reports. The study that used self-report of collisions was conducted to evaluate a community-level pedestrian prevention program in a low-income country that does not have a police report database (while low-income countries have police reports in response to crashes, many do not have electronic databases of these reports). In high-income countries and areas with lower population density, self-report poses limitations based on accuracy of recall as well as small sample sizes. However, even in this rural community, pedestrian collisions were sufficiently frequent to yield significant results. Simulation. The second most common study design was the use of simulation to measure driver behavior, used by 8 (19.0%) of the studies. Simulation studies are ideal for the objective study of driver response in risky situations. Simulation studies were used most often to evaluate road design interventions, and these included evaluations of roadway speed control (e.g., at curves and high-risk intersections) and lane control. Simulation studies were also used for vehicle design studies to examine driving in low visibility and to study distraction on rural roads based on size of a Global Positioning System (GPS) screen. All simulation studies involved all or partial scenarios on rural roads. Only one study used a simulation approach to evaluate an educational intervention, and this study evaluated a training program teaching drivers how to avoid roadway departure crashes on rural roadway segments. Simulation is a promising but seldom used strategy for educational program evaluation. In this case, the combination of an educational strategy with roadway design intervention provides a more comprehensive strategy to reduce roadway departure crashes. This is the only behavioral strategy for which two types of intervention were examined. Simulation studies, in combination with randomized impact and outcome designs, could provide a more comprehensive picture of intervention effects (this is addressed again in the Study Designs and Quality of Studies section). Simulation designs used standard methods for introducing the simulated drive, ordering drive sequences to eliminate bias, and in measuring standard driving outcomes such as speed differentials and lane placement. Simulation studies, as discussed more in the Study Designs and Quality of Studies section, have

60 a number of limitations when extrapolating intervention impacts to the population level because of limited sample sizes, undefined base populations, and lack of audience segmentation (examination of differences between driver characteristics). Surveys. Surveys were used in 8 (19.0%) evaluations. Five of the studies used surveys in pre-post designs to test educational interventions and one tested a campaign. The educational interventions addressed novice drivers, ATV safety (n = 2), and bicycle safety, and the campaign addressed ATV safety. Three additional surveys also included behavioral observations, and these were all evaluations of campaigns in which surveys provided information about knowledge/attitudes/behaviors and the observations provided objective data about behavioral impacts. Surveys plus observations were used for a child occupant safety campaign in which child safety seat and bicycle helmet use was observed, and for two teen risky driving campaigns in which speeding (n = 1), and seatbelt use (n = 2) were observed. The addition of observational data enhances the quality of the study by contributing both short-term information about knowledge and attitudes and objective impact information about behavior. Observation. A total of ten (23.8%) evaluation studies directly observed behavior through observational studies. Four of these were in combination with crash analyses and were addressed in the crash analysis section (several paragraphs above in this Study Overview by Intervention Type and Study Design section); three of these were in combination with surveys and were addressed in the survey section. Two (4.7%) studies involved observation only. Both examined vulnerable road users. The first examined a road design treatment to improve bicycle passing safety and the second observed vehicles at intersections treated to improve vehicle yield of right-of-way to pedestrians. Studies of vulnerable road users, and especially those involving road design interventions at limited settings, are challenged in using crash data because of small sample sizes. Thus, observational studies are the most efficient in identifying behavior change, while larger outcome analyses are needed for larger, and hopefully randomized, studies. Other study designs. Two studies involved randomized trials, both of which tested educational interventions focused on novice drivers. One study was a purely descriptive design of an educational intervention, and one qualitative study examined a community campaign. Studies by Behavioral Strategy Focus Areas The rural behavioral strategies examined in the 42 evaluation studies included speeding (11, 26.2%), roadway departure crashes (6, 14.3%), novice drivers (6, 14.3%), ATV safety (4, 9.5%), bicycle safety (3, 7.1%), general rural driving safety (3, 7.1%), motorcycle safety (2, 4.7%), farm equipment (2, 4.7%), pedestrian safety (2, 4.7%), child occupant safety (1, 2.4%), distraction (1, 2.4%), and low visibility (1, 2.4%). Speeding. Approaches to control speeding focused on high-risk rural roadways, rural roadways with curves, speed limits, and general speed. The most common approach to control speed was through roadway design. Five studies used simulation to test speed differentials: • Awan et al., 2019 tested speed and lane control for the application of optical circles and herringbone patterns on curves and found that both treatments reduced speed. • Yang et al., 2013 tested speed and lane control for the impact of curbs and found that in clear conditions speed was reduced when curbs were added, but that in foggy conditions, speeds increased without curbs.

61 • Calvi et al., 2019 tested four road design measures: white and red peripheral transverse bars, optical speed bars, and chevrons. All four measures reduced speed with the greatest impact from red peripheral transverse bars. Calvi further interviewed drivers and learned that drivers did not fully comprehend the signage messaging. Adding surveys about perception and understanding was important – it allowed authors to conclude that since participants did not fully understand messaging, it was the optical properties of the signage that led to speed reduction. • Peiris et al., 2018 studied roundabouts, reduced speed limits, and rumble strips in different combinations through a simulation study recreating serious injury crashes. All treatments reduced speed of impact, and the combination of rumble strips and reduced speed limits had the biggest impact on preventing impact. • Meulener et al., 2020, evaluated signage and sign content and found that the Rural Intersection Active Warning System-compliant sign that stated the speed limit at 80 km/h and the non-compliant sign that stated “slow down” showed the biggest reductions in speed. • Jaarsma et al., 2011 evaluated speed limit reduction and low-cost road treatments (e.g., signage) on a rural road segment and found that crashes were reduced 24%. • Jones et al., 2008 tested the impact of speed cameras and found that crashes reduced 19% and serious injury crashes by 44% after installation, significantly higher than non-treated intersections. • Goldenbeld et al., 2005 evaluated a rural speed enforcement program and found that average speeds on roads decreased significantly in intersection areas, with some spillover impact on local roads. • Vadeby et al., 2014 evaluated rural roadway speed limit reductions and found that fatalities decreased but no significant difference in severe injury crashes was found. • Retting et al., 2008 evaluated rural roadway speed limit change and found that mean speeds and the proportion of cars exceeding 80 mph both decreased significantly. • Grabowski et al., 2007 evaluated changes in the maximum rural speed limit and found an increase in fatalities on rural interstates of 17% when the speed limit increased to 65 mph and 36% when the speed limit was increased to 70 mph or greater. Overall, these studies found that speed control through both roadway design and policy approaches significantly reduced average speeds and crashes. The studies generally examined one strategy or a small set of strategies; larger studies that examine the impact of multiple approaches, especially those that combine road design, policy, and campaign approaches, are needed. Roadway departure crashes. Interventions to reduce roadway departure crashes, which tend to be highly fatal, included five roadway design approaches and one educational approach. These studies focused on road segments with high curvature or that were demonstrated to be high risk. • El Esaway et al., 2019 studied the implementation of roadside and median barriers and found they led to reductions in all serious crashes, crashes involving trucks, and off-road collisions.

62 • Khan et al., 2015 evaluated shoulder rumble strips and found a 14% reduction in roadway departure crashes and that they were most effective on roads with moderate curvature and larger paved shoulders. • Pardillo, Mayora and Pina, 2009, evaluated improvements in road friction surfaces and found they reduced wet-pavement crashes by 68%. • Park et al., 2012, evaluated implementation of wider edge lines pavement markings and found reductions in crashes. • Wood and Donnell, 2020, evaluated horizontal curve warnings and found reductions of 20 – 35% in total crashes, fatal and severe crashes, and roadway departure crashes. • Freeman et al., 2015 tested the impact of an educational training video that instructed drivers in how to avoid losing control and running off the road and found through testing in a simulator that the video increased driver response to avoid loss of control. Road design demonstrated effectiveness in reducing loss of control and crashes from roadway departure. The road design features were studied as single interventions, and studies that examine these treatments in combination, especially considering cost-benefit, would be helpful. Only one study examined driver education and found it to be effective. Building the evidence base to consider multiple types of strategies would contribute substantively to current knowledge and practice. Novice Drivers. Novice driver evaluations that included a rural community or rural driving focus included one policy, two campaign, and three educational strategies. • Blackman et al., 2008, assessed the Australian Graduated Driver’s Licensing system for rural communities and found that it was not highly effective in reducing unlicensed driving. • Chande and Roth, 2006, evaluated a teen risky driving program implemented in two rural schools and found that the program increased awareness and that participants reported less likeliness to speed; seatbelt observations found increased use in the rural school. This study did not report numerical findings and did not describe the campaign in detail. • Freund et al., 2019, evaluated a school-based seatbelt campaign program and found that the program increased seatbelt usage, but that schools had many challenges and implementation barriers. • Kumfer et al., 2017, evaluated a computer-based educational program about rural driving, designed for ultimate use in driver’s education. The intervention led to higher scores (improved) on driver safety behavior and rural safety. • Peek-Asa et al., 2014, evaluated a parent-based educational intervention for novice drivers that included rural roadway components and found that self-reported risky driving was reduced, and that the rural roadway component indicated the highest parent communication increase among 26 topics. • Peek-Asa et al., 2019, evaluated a parent-based educational intervention in augmenting the impact of in-vehicle video warning systems among mostly rural drivers and found that the parent-based

63 program led to significant reductions in driving errors, measured through the in-vehicle system, when compared to the in-vehicle system alone or no intervention. Novice driver interventions for rural driving evaluate few overall approaches, but show impact on knowledge, attitudes, behavior, and driving safety. No studies examined crash outcomes. ATV safety. Interventions addressing ATV safety, which are predominantly a rural risk factor, included one policy, one campaign, and three educational programs. • Testerman et al., 2013, evaluated changes in a rural trauma center’s admissions related to ATV injuries before and after a Wisconsin law that required youth ATV training and found that admissions increased following the law. This study had a weak design (before-after case series with no base population, no control, and no components to evaluate compliance). • Jennissen et al., 2017, evaluated a campaign that asked Agri-businesses to post an ATV safety sign. The evaluation included phone follow-up calls with 192 businesses, of which 81% reported displaying the poster. • Novak et al., 2013, evaluated an adolescent ATV safety program implemented in rural middle and high schools and found increases in ATV safety knowledge and use of safety gear. • Jennissen et al., 2015 evaluated a school-based ATV safety program on attitudes towards using the information provided. They found that students who most often rode ATVs and who currently used the fewest safety measures were the least likely to report future use of the program information. ATV safety is a particular focus for rural populations because ATVs are so prevalent for farm use, transportation, and recreation. ATV safety interventions mostly focused on children and had mixed results about effectiveness. Proper messaging to engage ATV users is needed, and the best location for educational programs should be tested. The school-based programs, for example, included many non-ATV riders. Youth agriculture programs such as Future Farmers of America (FFA) and 4H might be better settings and likely have already implemented safety focused programming. However, their programs may not be undergoing systematic evaluation. Bicycle safety. Rural bicycle safety interventions involved one road design intervention, one educational intervention and one campaign. • Kay et al., 2014, examined vehicle speed and lateral placement while passing bicycles before and after placing share-the-road signs, rumble strips, and speed limit signs. Rumble strips shifted vehicles closer to bicycles, road signs shifted vehicles away but did not impact mean buffer distance significantly, and speed limit signs resulted in reduced speed while passing. • Ji et al., 2017, evaluated a middle school bicycle safety program and found reduced reporting of bicycle crashes and improved knowledge, attitudes, and self-reported behaviors. • Swanson, 2010, evaluated a clinic-based program to provide helmets and information about helmets to parents. The evaluation was weak, indicating only that 24 helmets were provided and many in the community were exposed to the program, primarily through a Fourth-of-July event.

64 While the evidence base related to protecting vulnerable road users is increasing, few of these studies address rural populations or roadways. Studies that translate effective programming from urban to rural environments are needed, as are studies that identify ways to reduce risk factors specific to rural environments such as long stretches of narrow two-lane highways and uncontrolled intersections. General rural driving safety interventions included one campaign, one policy, and one educational approach. • Goodrow et al., 2004, evaluated a community campaign to increase safety on a rural road segment in Appalachia. The evaluation reported that there was a measurable reduction in fatalities, but the program activities, the design, and the actual numerical measures were not described. • Zimmerman et al., 2015, evaluated a comprehensive strategy to increase safety in a rural community in Tanzania. The program involved training, safety equipment, and safety promotion materials. Pedestrian (but not motorcycle or motor vehicle occupant) crashes decreased after the intervention, and injury severity overall decreased. • Varnild et al., 2019, evaluated the adoption and impact of the Swedish national road safety policy in a rural region of Sweden. Although safety measures were implemented, changes in crashes were not substantial. Motor vehicle occupant crashes in rural areas were higher for treated than non- treated roads. General rural driving safety evaluations are helpful because they examine multiple components of safety. However, they often address only one community, which provides a limited sample size or population for a rigorous evaluation. Motorcycle safety. Both rural motorcycle interventions focused on policy: one on helmet use only and one on helmet use in combination with alcohol intoxication and mandated operator training. • Passmore et al., 2010, evaluated Vietnam’s mandatory helmet use law in rural areas and found a significant increase in helmet use and a significant reduction in head injuries. • French et al., 200,9 evaluated multiple policy approaches and found that universal helmet laws and mandatory rider education contributed to reduced fatal and nonfatal injuries. Farm equipment. Two studies evaluated farm equipment policies to reduce on-road crashes. • Ramirez et al., 2016, examined state lighting and marking requirements for farm equipment operating on rural roads and found that more stringent requirements were associated with reductions in crashes. • Marlenga et al., 2006, evaluated crashes before and after a Wisconsin law that required training for youth operating tractors on the roadway. This ecologic study found no impact of the law but did not include process or impact measures. Pedestrian safety. Rural pedestrian safety interventions included two road design studies. • Liu et al., 2011, evaluated the implementation of rumble strips on high-speed rural crosswalks and found that they reduced vehicle speeds and could reduce pedestrian collisions by 25%.

65 • Wilson et al., 2015, conducted a feasibility evaluation of a student-led community participatory approach to improve pedestrian safety in a rural community. The only outcome measure was that the intervention idea was received enthusiastically by the community. Other approaches. Evaluations included one child occupant safety campaign in a rural community, one study of visual distraction based on GPS screen size, and one low visibility driving evaluation. • Floerchinger-Franks, et al., 2000, evaluated a state agency’s efforts to engage and support elementary schools in conducting child occupant safety campaigns focused on child seatbelt use, rear-seating, and helmet use. They found that seatbelt use and helmet use increased significantly more in participating than non-participating schools; no difference occurred in rear-seat placement. • Yared et al., 2020, tested visual distraction from a low-size GPS screen when driving on rural road segments and found that the smaller screen led to more navigational errors in the urban but not rural segments, and that speeding was more frequent on rural road segments. • Sharfi and Shinar, 2014, evaluated a Visual Enhancement System (VES) to help drivers in low visibility conditions and found that the VES system led to no difference in speed and that collisions were higher when using the VES. As a body of work, these evaluations indicate that many types of prevention strategies show promise in improving rural roadway safety. The toolkit, entitled Guidance for Improving Roadway Safety in Rural and Tribal Settings, created as an outcome of this project includes resources for how to find these prevention strategies. However, especially when compared to the evidence base for overall road safety and for health outcomes in general, very few evaluations have been conducted. Most standards (e.g., the CDC Community Prevention Taskforce and the U.S. Prevention Services Taskforce) require multiple studies of different designs and population before a prevention approach can be recommended. Study Designs and Quality of Studies The quality of evaluation studies focused on behavioral interventions that reduce crash incidence and injury on rural roads and for rural drivers was weak. Few of the studies used theory-driven evaluation methods, such as a logic model. Because of this gap, it was difficult to understand where or how many of the studies fit within a comprehensive evaluation strategy: specifically, few studies identified if the evaluation contributed to fidelity, process, impact, outcome, dissemination, or translation. In evaluation research focused on behavioral interventions, the general categories of measures include fidelity, process, impact, outcome, dissemination, and translation. Fidelity evaluation measures the program’s plans, procedures, activities, and materials, and, generally, fidelity evaluation should be conducted prior to any large-scale implementation. Process evaluation examines if the program is reaching the intended audience and if the audience understands, accepts, and engages in the approach. Process evaluation is critical for understanding how an intervention works or does not work. Impact evaluation for a behavioral intervention measures the extent to which the intervention altered actual behaviors in the desired manner. Outcome’s evaluation examines whether or not the behavioral changes lead to the intended impact on the health outcome. For example, if a speed intervention shows some impact on overall speed reduction, but does not lead to reductions in driving errors, near-miss crashes, or crashes, then the intervention may not have the correct intended audience (e.g., it addresses low rather than high-risk drivers). For an intervention to be evidence-based, it must have evaluation findings for each of these categories. Dissemination and translation studies examine how evidence-based interventions can be scaled to broader

66 audiences and adapted to new audiences. The majority of studies examined outcomes, usually crash incidence, while a few examined process measures. The focus of this review was on behavioral strategies. However, few studies presented behavioral change theories underlying the intervention approach. The majority of strategies employed individual-level change, and the guiding theory of change was not presented. Combined with the gap in comprehensive steps in evaluation, the synthesis of this literature as a whole did not have sufficient evidence to describe how the intervention works (if results were positive). Overall, the designs used in these studies were weak. Randomization was used in very few studies, and comparison groups were often not used at all. The most common study design was a before-after comparison. In identifying behavioral change, and specifically how behavioral change was achieved, before-after comparisons are far weaker than randomized trials. Quality of study design was weakest for campaigns, which is likely because conducting a comprehensive evaluation of a campaign is costly and challenging, in part because the evaluations require collection of primary data to identify changes in normative attitudes and behaviors. The most common study design was a before-after design, often with a control group, and the most common outcome was police-reported crashes. For example, many roadway design studies, and speed policy studies examined crash rates on road segments before and after the roadway design intervention was implemented. The majority of these studies used road segments without the intervention as a control and estimated the intervention impact using Bayesian models. This is a strong design for outcome evaluation, but only a few of the studies also measured behavior modification. One example is the speed policy study that included speed monitoring before and after road segment intervention; the inclusion of behavior change measures adds considerable strength to the study. Simulation studies are an advantageous design when examining high-risk driving behaviors because they can test behavioral responses without introducing risk to the participant. Simulation studies were used for several roadway design interventions, and in particular those that involve interventions that reduce loss-of- control and roadway departure crashes. Simulation studies were the only design used for vehicle modification interventions. While simulation studies provide strong objective measures of behavior in simulated environments, the studies rarely identified a base population or considered representation of their participants. Since simulation studies tend to involve small sample sizes, the participant pool is unlikely to be representative of drivers overall. Simulation studies often develop sample size estimates based on validity and representation of driving response, so that individual-subject variation is likely to impact the intervention measures. However, the pool of potential simulation study participants is likely to be self- selected to potentially miss the highest risk drivers (which is a concern in many types of safety studies) and also not likely to reflect the overall population-level driving diversity. Simulation studies rarely provided information about recruitment or population representation of their subjects, and also rarely identified a target population. For example, despite the fact that all studies involved rural interventions, no simulation studies mentioned recruiting participants based on rural roadway experience, measured rural roadway experience, or controlled for it. While the overall quality of the studies was relatively weak compared with the general evidence base for health-related studies, a far more important gap is the rarity with which rural, behavioral-focused interventions are evaluated at all. A wide variety of agencies, ranging from schools to youth organizations to state DOTs have implemented education programs for rural road safety, and the majority of these efforts appear to have no associated evaluation activity. Thus, it remains a challenge to identify effective single-

67 component interventions, how these interventions might work together for population-level impact, and how evidence-based strategies can be translated to new populations. Zero Deaths Campaign Overview Vision Zero Website Identification Methods The goal of the review of state Vision Zero programs was to identify content geared specifically to rural populations and to describe the behavioral components of the programs that would impact rural drivers. The project team searched websites in all 50 states and Washington D.C. to identify Vision Zero information. First, the websites were identified through a Google search using the name of the state along with phrases such as "Vision Zero," "TZD," "Vision Zero Fatalities." The words “website” or “Department of Transportation” were used at times to help improve the search when the state website was difficult to locate. For states that did not yield search results with a state or city website, the state’s DOT website was located and searched for the same terms as listed above. Most often this step would prove fruitless, and the most recent SHSP in a PDF format would be sought instead. On the summary spreadsheet, states were color-coded according to their type of website or if there was no website. No color meant there was a state website. Green meant that the state had a city, region, or island (Hawaii) website. Yellow meant there was no state or city website, so the SHSP was used. Some states, such as Colorado and California, had multiple cities with websites. The city used for review was specifically chosen based on a few criteria. If there was only one city website, it was used as it was the only option, such as with New York City, New York. If there were multiple cities with websites in the state, the state’s capital city took precedence if it had a website, such as with Denver, Colorado, over Boulder. If the capital city was not an option, but there were still multiple city websites, the most populated city took precedence, such as looking at Tampa, Florida, over Orlando or West Palm Beach. One state, Minnesota, had a state website, but also had different regional websites with specific behavioral targets in each. While the state website was reviewed, the Metro Minnesota Region was also reviewed as it included the capital of St. Paul and the highly populated city of Minneapolis. Some states or cities had their Vision Zero websites as a separate URL while others simply had it attached to the state or city government website. Twenty-three states and the District of Columbia (making the total 24) were found to have state websites for the Vision Zero initiative, which means the state as a whole adopted Vision Zero as one of its programs. Fourteen states had not adopted Vision Zero, but a city or cities, or regions, within these states had separate Vision Zero websites. In this case, the Vision Zero website that represented the state Capitol was chosen for review (or an island in Hawaii’s case). The remaining 13 states had neither a state nor any city/region websites that identified Vision Zero activities. In these states, the most recent SHSP was reviewed to identify any specific rural content. Vision Zero Website Review Methods For each relevant website or highway plan, the research team searched the content to collect information on any reference to rural roads, populations, or specific strategies with the goal of determining if those areas and roads were the focus of any states’ Vision Zero plans. If the word “rural” was mentioned specifically, the research team identified the context in which the word was used. The research team then examined the website content to answer the following questions: 1) What behavioral or people-based strategies are used

68 within the state/city website? If these were identified, they were summarized; 2) Identify types of messaging used within the website (stories, data, safety tips, videos, resources, news, etc.); 3) Is rural roadway safety mentioned in any way on the website or in the SHSP? If rural was mentioned, the content was summarized; and 4) Does the SHSP mention Zero Deaths/Vision Zero in any way? The home page was first searched for all the information. Some websites did not have any behavioral strategies on the home page, so other pages on the website were searched. For example, some states divided their activities into categories such as “data,” “road design”, or “driver safety,” and the term that most aligned with behavioral strategies was reviewed (in this example, “driver safety”). For states where the SHSP was used instead of a Vision Zero plan, the research team focused on slightly different questions. They examined whether any terms related to “rural” were mentioned in the SHSP, and if so, how many times and in what context. The second question regarded whether the SHSP mentioned Vision Zero, and if yes, how many times and in what context. To find the answer to these questions the word “rural” was searched as well as “Vision Zero,” “Zero Deaths,” or just “Zero.” The two questions on behavioral strategies and types of messaging were not pursued in the SHSP searches. Quality of Vision Zero Websites The appearance and ease of use of websites varied greatly. For example, the state of Iowa’s website was one of the better ones as it had an appealing design that was easy to navigate. It provided bold, clear statistics and safety tips right at the top, making the behavioral focus areas clear. Associated images make the site engaging and easily understandable. The site had a clear, concise section focused on the fact that crashes are avoidable, as most are due to human error. It then provided videos about how to practice traffic safety, as well as real life stories and photos of individuals, which adds an emotional appeal. There is a section called “Message Monday” and the blog that goes with it, which is a campaign for boards on the interstates throughout Iowa that carry safe driving messages. The site also clearly lists the program’s partners, ways to learn more about the program, other websites to look at, the safety plan, other statistics, prevention strategies, and ways to get involved. One website that was less engaging was for Anchorage, Alaska. It included simply one page within the government website that had two images/graphics, a short summary of the program, and a couple of bullet points with recommended safety tips. The site did not provide any data, statistics, or more information about the focus area; however, it did provide links to all of the Vision Zero Annual Reports, the Action Plan, a message from the mayor, presentations, and contact information. While these are good resources, having them only as attached documents and not web content made them difficult to engage with or navigate. It would be beneficial to have more specific, attention-grabbing information about the program and its goals on the website directly, rather than requiring users to search through other PDF documents. Vision Zero Website Content The overall findings were similar within each website group (i.e., state website, city website, or SHSP). Across the board for the state and city website groups, all had variations in their content, but there were also many common types of content, including statistics, data, images/photos/graphics, safety tips, and partner information. For instance, the state of Kansas’ website included safety tips targeted at the priority behavioral areas and used an emotional appeal. The campaign was centered around the slogan, “Who do you make it home for?” Impaired driving was one area of focus and simply included the safety tip, “Don’t drink and drive. Celebrate responsibly so you can be with the ones you love.” Similarly, the safety tip for

69 speeding was centered around being safe for your loved ones and others: “Is getting there faster worth taking a life? If you won’t slow down for yourself, do it for them.” Also, the seatbelt safety tip said, “Always buckle up so you make it home to them every time.” These messages are short and to the point with an emotional appeal to make people think about their actions, and they align with values centered on family and relationships. Many different behavioral strategies were also highlighted when searching through the 38 state or city websites, but the most common were speeding, impaired driving, distracted driving, and seat belts. While all of these behavioral strategies are important for rural populations, no sites segmented the rural audience for their messaging. Audience segmentation may vary if rural safety culture has specific goals and values that differ from urban counterparts. For example, nearly all of the websites addressed speeding, for which general messaging may have some success. However, the specific behaviors and strategies to reduce speeding in rural areas may vary by the types of roadways, driving characteristics, and traffic safety culture. The presence of rural-focused intervention studies in the literature review are one indication that rural speed control may need to be different than urban speed control. Vision Zero sites frequently identified partners, and many of these partners had a rural connection. Some examples of the community partners committed to helping fulfill the goals of Vision Zero listed on the New Hampshire website included AutoFair Automotive Group, The New Hampshire Association of Chiefs of Police, Children’s Hospital at Dartmouth-Hitchcock, the Community Alliance for Teen Safety, the Derry Community Television, and the New Hampshire Automobile Dealers Association. For the states with a statewide website (no color), none mentioned the word “rural.” This included states such as Iowa, Missouri, Montana, Minnesota, North Dakota, and Wisconsin, all of which are known to have many rural areas. For the states that only had city, region, or island websites (green color) only one had mention of the word “rural.” This single mention of “rural” was the city of Tempe, Arizona, and the mention was that a specific “Rural Road” was involved in a current intersection project. Therefore, rural roads and areas did not seem to be a significant priority to these Vision Zero websites even though many of the behavioral strategies emphasized are important in rural areas. The review of the SHSPs for the states with no website (yellow color) looked at different questions but yielded similar results in all states. For 12 of the 13 states’ SHSP, some version of “Vision Zero,” “zero deaths,” “zero fatalities,” or “TZD” was mentioned, with some variation in the number of times mentioned. For instance, the South Dakota SHSP only referenced this term once in a description of a “Drive to Zero Task Force,” but details of this taskforce and its activities were not found on a corresponding website or within the SHSP. The state of West Virginia’s SHSP mentioned moving towards the goal of “zero fatalities” 11 times. Connecticut was the only state that did not have any specific reference to a Vision Zero plan, but it did mention related goals, in that the state wanted bicycle fatalities to be reduced to zero and pedestrian fatalities to be reduced to less than 20 by 2014 (2010 was the latest SHSP available). The word “rural” was mentioned in all 13 SHSPs, with variations in how often it was mentioned and in what context. Idaho’s SHSP only mentioned rural three times, while South Dakota’s SHSP mentioned it 199 times. Throughout all the plans, “rural” was most commonly brought up when discussing HRRR and related programs. One example of such a program was mentioned in both the West Virginia and South Dakota plans. Both defined what a HRRR is, and both stated that there is now an established rule that requires that if the fatality rate on a state’s rural roads increases over a two-year period, the state is required to obligate 200% of its fiscal year HRRR set-aside for projects on the HRRR system. For instance, West Virginia mentioned that if a road experienced an increased fatality rate in this defined time period, it would be required to obligate approximately $1.6 million to HRRR. The word “rural” was also mentioned in the context of defining the word, the discussion of area (i.e., urban versus rural) or road demographics, or the use of it in describing statistics or in graphs/charts. Specific driving risks or behaviors as they apply to rural populations were not mentioned, nor were campaign or messaging focused specifically on rural audiences.

70 In conclusion, states vary in their adoption of the Vision Zero initiative. Some states adopted a statement of implementation, one state had a statewide approach with individual plans by region, and several states had Vision Zero represented at the city level but not statewide. These states are likely to have the least representation of rural populations in their activities. Most states that did not have a Vision Zero website still mentioned it in some way in their SHSP, often as a goal or vision, which reflects their intention or potential for related activities. None of the state or city websites, with the exception of Arizona, mention the word “rural” on their home page, even though many rural areas are included in these 38 states and many of the behavioral strategies under consideration are important for rural areas. The 13 states with SHSPs all mention the word rural in some way even if they are not states typically thought of as having rural areas, such as Rhode Island or Delaware. Therefore, the remaining states could have mention of the word “rural” in their SHSPs even though their state or city Vision Zero websites do not mention it. Overall, rural roads and areas do not seem to be a specific target when discussing Vision Zero in most states, and other behavioral issues seem to take precedence with strategies to address them, such as speeding in urban areas; however, these behavioral issues remain important to address in rural areas. Barriers and Approaches for Overcoming Barriers in the Implementation of Behavioral Rural Traffic Safety Interventions Approximately 14% of the U.S. population lives in rural-designated areas, yet 71% of U.S. roads are rural and rural transportation accounts for 30% of VMT (Congressional Research Service, 2018; Cromartie, 2018). Rural populations thus represent a disproportionate amount of travel, yet research and programming, especially for behavioral strategies, has focused far more on urban environments (one indicator being the number of peer-reviewed publications in the PubMed index focused specifically on urban traffic safety, n = 1,469, and specifically on rural traffic safety, n = 460). Rural populations are highly heterogeneous, and the complexity of rural populations and roadways are an important factor in implementing effective and cost-beneficial prevention and intervention strategies. Behavioral strategies, both educational programs and campaigns, are in particular not a “one size fits all” intervention approach, and these may need regional variation and message segmentation to be effective. A number of trends in rural populations are important when considering how to best develop, implement, evaluate, disseminate, and translate rural-focused traffic safety interventions. After a long decline in rural populations, the U.S. is now experiencing growth in rural areas (Cromartie, 2018; U.S. Census Bureau, 2016). This growth is found throughout the country but is highest in the western U.S. Rural populations tend to have less racial and ethnic diversity than urban areas, with those who are racial/ethnic minorities comprising 22% of non-metro populations and 42% of metro areas. However, two trends are important for future planning: rural diversity is increasing, and this growth is represented in the origins of in-migration (from other U.S. areas and from international origins). The rural population has traditionally had a higher average age for both men and women, but age differences are becoming more heterogeneous throughout rural populations, influenced by two trends. Much of the rural population growth is fueled by in-migration of people of working ages, which contributes to a lower average age; however, in-migration is also high in concentrated areas due to retirement relocation, which contributes to a higher average age. Aging in rural counties largely depends on which of these in- migration trends is more pronounced in each specific county (Currie and Philip, 2019). Older rural populations are more likely to live independently, which will have important implications for the age of rural road users.

71 Rural disparity in education, employment, and income is more diverse (a wider range with higher percentages in high standard-deviation categories) than in urban populations and is changing over time. Rural poverty rates are down from the 2013 peak, yet job growth lags in rural compared to urban counties (Cromartie, 2018; U.S. Census Bureau, 2016). Although median household income is lower in rural populations, the proportion of the population living in poverty, especially among children, is lower in rural populations. A number of barriers and facilitating factors should be considered when addressing rural needs in transportation safety. These are developed below. Geospatial Rural areas, by their definition, are on the lowest end of the spectrum of population density. Rural populations must travel further to reach services, and rural jurisdictions must address safety on a longer road network, often with fewer resources. These geospatial configurations lead to a number of challenges in prevention and intervention programming. Barriers include: • Rural populations are more spread out, thus behavioral message delivery is a challenge. Rural drivers use more roadways (e.g., no interstates) so message delivery on the road (such as through signage) is less efficient. • Rural roads are more heterogeneous, leading to a greater diversity of driving conditions. City drivers mostly use primary roads and interstates, while rural drivers also use roadways such as two- lane highways (with and without shoulders) and unpaved roads. In mountainous rural areas, roadways have much more curvature and sinuosity than city roads. • Rural roads have bigger roadway diversity, which includes high-risk configurations such as uncontrolled intersections and highways with intersection-like crossings. These require specific driving skills, and for intervention, a larger toolkit of behavioral messages. Effective intervention is likely to require road design as well as behavioral strategies. • When conducting behavioral interventions within the diversity of rural populations, it can be challenging to find appropriate control groups that are geographically representative (Robertson and Pashley, 2015). Engaging control groups, which are often further away in rural and urban studies, can also add time and cost to studies. Engagement Particularly for behavioral strategies, engagement of multiple audiences is important. For educational interventions, the participating population must find the program accessible, understandable, informative, and engaging. Regardless of how the program is delivered, the people involved in the development and delivery must be engaged and well trained for thorough in-person or self-led education. In campaigns, broad engagement is even more important and must begin at the inception of the campaign (Batras et al., 2016). The message content, delivery, source, and dissemination should have input from the participating community for optimal effectiveness. Some considerations: • Engagement and participation are essential for effective educational, campaign, and policy approaches. Obtaining this engagement over the larger area of a rural population can be a challenge.

72 For example, a campaign involving local businesses may need to engage several towns to reach adequate sample size. Bringing people together when they live in isolated rural areas can be costly and time-consuming. • Successful behavioral change is based on messaging that resonates with values (Batras et al., 2016). Few studies have identified the types and diversity of values that could serve as the basis for rural road safety messaging, so rural safety programs are starting from further behind than urban areas. Brann, et al, examined elements required to develop a meaningful ATV safety campaign and found rural residents wanted action-oriented information, hands-on interaction, clear and realistic information, and the use of personal stories similar to the target audience (Brann et al., 2011). • In smaller communities, which tend to be closer-knit and with people known to each other, messages need to come from a trusted source. Learning where these pockets of leadership lie within a rural population requires knowledge of that community. In addition, closer-knit rural communities may have inherent distrust in those outside their communities, so partnerships must first build a relationship of trust. • The authorities that oversee rural roadways have more jurisdictional entities than those for urban roads, since rural roadways may go through multiple incorporated towns and unincorporated areas (Golembiewski and Chandler, 2018). For example, one rural roadway may have oversight from multiple city and county agencies. Engaging multiple entities, which may have a good history of working together or not, presents a challenge to rural roadway safety. • Delivering programs in rural environments poses challenges in timing, location, and transportation, as well as engagement across populations and sectors (Novak, et al., 2013). Infrastructure and Resources Rural populations have fewer resources to maintain an infrastructure that spans a larger area. Despite growth in population and socioeconomic status in many U.S. rural counties, infrastructure and resources remain a gap in rural compared with urban communities (U.S. Census Bureau, 2016). These gaps provide several challenges: • Many behavioral strategies engage schools, and rural school systems differ. As rural populations declined over the last decades, school systems have evolved in one of two ways. In one structure, rural towns have maintained their schools, often increasing the number of grades to keep an effective school size. In the other structure, school districts have consolidated so that schools are larger and children from many towns are bused in. These trends have implications for both transportation time (as many more rural children travel further to consolidated schools) and also for the school culture. Schools spanning more grades but representing a smaller community population have a tight-knit culture and the school will be well and highly integrated into the community. In larger consolidated schools, children from multiple communities who do not know each other, or each other’s families will be in classes together. The school will represent multiple communities, and engagement with these communities is likely to be lower. • Within rural small and mid-sized towns, businesses can serve multiple roles to fill gaps. For example, rural schools were among the first to work with healthcare systems to embed mental health services in school settings, in part because these services were not elsewhere available if the

73 town did not have its own hospital or clinics. When designing behavioral interventions, especially campaigns, understanding the diverse roles of local businesses and community leaders is critical. • Rural roads have less intentional design since they have evolved over time, in particular those designed in response to transportation modes no longer prominent (e.g., horse or buggy). Thus, the diversity of addressing challenges along individual road segments is greater (Golembiewski and Chandler, 2018). • Access to definitive trauma care is often delayed due to longer distances that EMS crews, often volunteers, must travel, as well as longer distances to trauma hospitals. Delays in reaching definitive trauma care contribute to higher mortality rates and poorer recovery given the same severity of injury. Thus, prevention is a critical priority in rural areas. Risks Specific for Rural Road Users With significantly more activity in urban than rural road safety research, it might make sense to begin advancing rural roadway safety by translating successful urban programs to rural environments. However, rural populations and environments have a number of unique factors that require approaches to be considered independently from those used in urban areas. Translation of urban programs may miss these factors entirely. • The agricultural, forestry, and fishing industry has the highest occupational fatality rate of all industries in the U.S., and transportation is the second leading cause of occupational death (Census of Fatal Occupational Injuries, 2019). This industry exists almost entirely in rural areas and introduced a number of unique risks that include the presence of occupational equipment on the roadway, operation of vehicles such as ATVs, and a diverse fleet on low-volume streets (occupational vehicles such as logging trucks and combines and passenger vehicles such as pickup trucks and small passenger cars). • Vehicles such as ATVs pose particular challenges for roadway safety. ATVs are used in rural environments for work, transportation, and recreation, yet have a very high crash rate. ATVs are driven both on and off-road and are sufficiently high-powered to operate in rough terrain and on high-speed roads. However, they offer little OP. As evidenced in the review synthesis, a number of programs have addressed ATV safety, yet far more are needed. ATV safety program evaluations are challenging because of the range of populations impacted and the many different types of risk factors. • Rural crash rates are higher, and they have a higher proportion of head-on, roadway departure, and single vehicle configurations. Responding to these more severe crashes is a priority for rural interventions (Golembiewski and Chandler, 2011), yet saturating the roadway with effective approaches is costly. Cultural There is very little data on traffic safety culture segmented by rurality/urbanicity. Some traffic crash and safety behavior data provide clues about rural culture. For example, rural drivers are less likely to use seat belts. However, studies and experiences suggest:

74 • Rural populations often value autonomy and individual responsibility and driving messages must align with these values to be effective. Proscriptive messages, or those that appear to be telling the driver what to do, may not resonate. (Rakauskas et al., 2009). • Rural populations feel a strong sense of community, and smaller cities can feel they do not have as strong a voice as big cities in state politics. Messages that originate from the community may resonate more strongly than messages appearing to come from state agencies. • Since traffic enforcement is less dense in rural enforcements than in urban ones, and carried out by a wider number of jurisdictions, perceived value of following road safety laws may be impacted. One critical tenet of enforcement policy, for example, is the perception of being at risk of being caught violating the behavior. The need for perceived enforcement has been well documented for both speed control and for roadside sobriety testing. Rural populations may have lower levels of perceived likelihood of risk or of being caught than urban populations. Rural Data for Program Evaluation Data, as mentioned in the Rural Behavioral Surveillance and Protective Factors section, are critical components to the evaluation of an intervention program. There are several barriers to rural data collection: • Rural crashes, particularly non-injury crashes, may be less likely to be reported. Because the wait time for police response is longer, there may be a disincentive to report property-damage-only crashes. • Police crash reports are likely to have higher variability since multiple jurisdictions have patrol authority over different roads. While reports from a single jurisdiction, such as an incorporated city, are likely to have more internal consistency, rural roadways have a higher number of jurisdictions covering any roadway, and also have integrated response from police and sheriff’s offices (Golembiewski and Chandler, 2011). • Rural safety strategies must address the risk factors specific to rural transportation, such as farm equipment and ATVs. However, identifying these vehicles on police crash reports is challenging. Each state has its own definition for vehicle type; some have specific categories for ATVs or farm vehicles while others lump these in with motorcycles or semi-trucks, respectively. • Traffic enforcement, given the larger number of roadway miles that traffic enforcement must cover, is not as dense as in urban environments. Thus, studies that use citation, charge, and conviction data as an outcome measure will have a different base population of reports when compared with urban populations, and, furthermore, these citations may differentially represent towns that isolate rural areas. • Technologies such as cameras and traffic volume monitors have provided critical data on transportation exposure as well as driving behavior (e.g., near-miss studies of Interstate crashes in poor weather conditions). These technologies are found primarily on interstates and urban primary roads. Placement on rural roads is challenging because of the lower traffic volume on a higher number of road miles, making these measures difficult to place and costly. Additionally, these technologies are not well accepted in the U.S. and may be even less so in rural areas.

75 Task 3 Key Takeaways Below are some key takeaways from study review: • In general, the studies reviewed indicate that many types of prevention strategies show promise in improving rural roadway safety. • Very few evaluations have been conducted for behavioral-focused interventions, especially for vulnerable road users. • The overall quality of the studies reviewed was relatively weak compared with the general evidence base for health-related studies. • It remains a challenge to identify effective single-component interventions, how these interventions might work together for population-level impact, and how evidence-based strategies can be translated to new populations. • The studies indicated the following generalizations: – Speed control through both roadway design and policy approaches significantly reduce average speeds and crashes. – Road design demonstrated effectiveness in reducing loss of control and crashes from roadway departure. – The few novice driver interventions for rural driving evaluated show impact on knowledge, attitudes, behavior, and driving safety. – ATV safety interventions mostly focused on children and had mixed results about effectiveness. • States vary in their adoption of the Vision Zero initiative (e.g., statewide adoption, state website, city level approach, reflected in the SHSP) and overall, rural roads and areas do not seem to be a specific target when discussing Vision Zero in most states. • Barriers to implementing behavioral rural traffic safety interventions must be considered including geospatial, engagement, infrastructure, resource, rural specific risk, cultural, and rural data barriers. • Future interventions for rural populations must consider growing diversity in race/ethnicity; a wide range and increasing disparity in income and education; and an aging population. Additional research needs that would contribute to current knowledge and practice include: • Larger studies that examine the impact of multiple approaches, especially those that combine road design, policy, and campaign approaches. • Systematic evaluation of youth ATV safety outside of the school setting (e.g., FFA and 4H). • Studies considering cost-benefit of behavioral interventions. Cost-benefit analyses that help prioritize different types of individual and groups of interventions. • Studies that translate effective programming of vulnerable users from urban to rural environments. • Studies that identify ways to reduce risk factors specific to rural environments, such as long stretches of narrow two-lane highways and uncontrolled intersections.

76 • General rural driving safety evaluations that examine multiple components of safety for larger sample sizes (e.g., multiple communities). • Evaluation of the rural road safety education programs from a wide variety of agencies (e.g., schools to youth organizations to state DOT).

Next: Interviews (Task 4) »
Highway Safety Behavioral Strategies for Rural Areas Get This Book
×
 Highway Safety Behavioral Strategies for Rural Areas
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Rural roads have a higher risk of fatality or serious injury than urban roads due to factors such as varying terrain, wildlife, and long distances between services.

BTSCRP Web-Only Document 4: Highway Safety Behavioral Strategies for Rural Areas, from TRB's Behavioral Transportation Safety Cooperative Research Program, documents the overall research effort that produced BTSCRP Research Report 8: Highway Safety Behavioral Strategies for Rural and Tribal Areas: A Guide. Supplemental to the document is a PowerPoint presentation that outlines the project.

READ FREE ONLINE

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!