11.3 Intervention Options and Selection

Principles and Criteria for Intervention Selection

Once the problem type has been identified (whether through crash analysis or other forms of risk assessment), the next step in the process involves the selection of an appropriate intervention. The main aims during this stage are:

• the identification and selection of interventions that are expected to reduce the number and/or severity of crashes of a particular and dominant type;
• to check that the selected interventions do not have undesirable consequences (in terms of safety or otherwise);
• to maximise the benefits that can be achieved with the available funding (ensuring that the selections made are cost-effective over a number of sites);
• to select interventions where the benefits outweigh the costs.

There are a number of issues to consider when selecting interventions. Usually cost and economic efficiency are the first and foremost considerations, but there are also others. It is important to ensure that the intervention is cost-effective and gives a positive benefit-cost ratio, and that it can be implemented within the available budget. Based on issues identified by Ogden (1996) and BITRE (2012), other considerations include:

• Crash reduction – will the intervention maximise crash reduction, and will the treatment deliver the expected crash reduction at this location?
• Effect on traffic – is there likely to be any effect on traffic delay and redistribution of traffic (particularly to lower quality parts of the road network)?
• Health implications – a reduction in death and serious injury is the most obvious issue, but there may be other health implications from the selection of different treatments. For example, certain treatments may lead to safer conditions for walking and cycling, thereby increasing community health outcomes.
• Environmental issues – for example, increased or decreased air and noise pollution, or though removal of roadside vegetation that may have some visual appeal but be a hazard for vehicles.
• Accessibility – there is a need to ensure that access is maintained for key road users, including the elderly and disabled.
• Public acceptance – whether the intervention will be accepted by the community and supported politically; and whether non-compliance will be an issue.
• Public understanding – will road users know how to use the new intervention (e.g. driving behaviour at roundabouts)?
• Feasibility – will it be possible to install the intervention at this location (including issues relating to physical constraints, sight distance, etc.)?
• Legal issues – is there legal provision for the intervention? Will users be able to continue to conform to road laws once it is implemented?
• Compatibility – is the intervention consistent with other strategies and measures that have been previously implemented in similar situations?
• Available staff resources – is there staff available and skilled enough to install and maintain the intervention?

A ‘hierarchy of control’ is often used in risk assessment when selecting and prioritising interventions. As an example, Marsh et al. (2013) suggest that such a hierarchy helps identify a priority order for different types of road safety treatment based on outcomes. They suggested a road engineering hierarchy based on the Safe System approach to help address driver distraction and fatigue. The hierarchy has four levels, and it is suggested that level 1 equates to a level of risk where Safe System outcomes are likely:

• Level 1: Lower the forces transmitted through ‘conflict points’ (e.g. the point of impact between two vehicles) to within the human tolerances (particularly through the management of speed);
• Level 2: Design so the road ‘talks’ to the road user (this includes design to naturally cause safe road user behaviour (e.g. audio-tactile edgelines to reduce lane departures);
• Level 3: Design to provide opportunities for road users to recover from mistakes and non-compliance;
• Level 4: Design to lower the risk of a crash occurring to an ‘acceptable’ level.

Effective Safe System Interventions

There are a large number of safety interventions that can be used to improve road safety. Some have only a small impact on safety, while others can produce substantial reductions in death and serious injury. The concept of ‘high-performing’ infrastructure was introduced in The Role of Safer Infrastructure in Safe System Elements and application, and has been discussed in the context of the Safe System approach in several documents. For example, Turner et al. (2009) provide a distinction between ‘primary’ and ‘supportive’ road safety treatments. Primary treatments are those that make an important step towards a Safe System outcome, i.e. zero fatal and serious injuries. This can be achieved through reducing impact forces to safe levels or by separating different road users. A supportive road safety treatment is one that assists with the delivery of safety improvements, but only in an incremental way. For example, a hazard warning sign may reduce the occurrence of crashes (which can include severe crashes), but will have no impact on the severity of a crash, should one occur. It is strongly recommended that primary treatments are employed where possible, in an effort to reach the Safe System objectives. Examples of some key treatments are provided in Figure 11.3.

Issues specific to LMICs regarding the use of such treatments are discussed further in Intervention Effectiveness in LMICs in Intervention Option and Selection.

NZTA (2011) refers to Safe System Transformation treatments for rural roads. These are defined as treatments that are likely to address high percentages of the fatal and serious injury crashes of the three key main crash types for rural roads (run-off-road, head-on and intersection crashes). It is recognised that these treatments are typically higher cost, and that they need to be implemented over a longer time period. Examples of such treatments include expressways (4-laning and 2+1 treatments), median and roadside barriers, wide medians, clear zones (areas free from hazards on the roadside), grade separation (overpasses and interchanges), roundabouts, and effective speed management.

Crash modification factors

One of the most important considerations in selecting interventions is knowledge of the safety benefit of that treatment. This benefit is often described as a crash reduction factor (the expected percentage reduction in crashes), or crash modification factor (CMF, which is the multiplier by which the crashes before treatment are adjusted; e.g. a CMF of 0.8 indicates that there will be an expected reduction of 20% in crashes). A number of sources exist that provide information on this issue (also see Box 11.1):

• The Handbook of Road Safety Measures (Elvik et al., 2009): this book provides extensive information on 128 road safety measures, including those relating to road design, maintenance, traffic control, vehicles, driver training, education, enforcement and post-crash care. Some of the interventions included under the road design and traffic control headings include roundabouts, roadside safety treatments, road lighting, speed limits, pedestrian treatments and road markings. A rigorous approach is taken to the selection of studies and calculation of benefits. Studies from around the world are included, with a heavy emphasis on European research.
• Examples of assessed road safety measures - A short handbook (EU, 2006): this handbook is the main output from the EU-funded Rosebud project. The handbook (available from http://partnet.vtt.fi/rosebud/products/deliverable/Handbook_July2006.pdf) includes information on economic evaluation for various road safety measures, including infrastructure, road user and vehicle measures. Based on the benefit-cost ratio, each measure is ranked as ‘poor’, ‘acceptable’ or ‘excellent’. The content is international, but largely reflects European research.
• Effectiveness of Road Safety Engineering Treatments (Austroads, 2012): this document (available here: https://www.onlinepublications.austroads.com.au/items/AP-R422-12) provides information on crash effectiveness for 57 infrastructure safety treatment types, and 126 crash effectiveness values have been derived for them. The research basis is international, but there is an emphasis on research from Australia and New Zealand.
• Highway Safety Manual Volume 3 (AASHTO, 2010): Part D provides Crash Modification Factors (CMFs), allowing the quantification of expected changes in crash numbers resulting from changes to geometric or operational elements. This document draws on international research, but has a strong focus on research from North America.

Box 11.1: CMF Clearinghouse

Given the objective of the Safe System approach is to eliminate death and serious injury, it is important to understand the effect that different interventions have on fatal and serious outcomes. However, much of the research on intervention effectiveness provides information on casualty reduction (i.e. reduction in deaths, serious injury and minor injury combined) or on change in all crashes (including non-injury). This is an important distinction, and it is unfortunate that information on fatal and serious outcomes is so scarce. Although it is desirable to minimise all crash types, including crashes that do not result in injury, an overall reduction in fatal and serious injury is paramount. Safety professionals should not be put off using interventions that have a neutral effect on minor and non-injury crashes, and there may actually be situations where such crashes will increase (typically through a reduction in severity of the crashes that do continue to occur at a treated location).

In the absence of information on the effect of interventions on fatal and serious crash outcomes, information on casualty reduction should be used, although an element of engineering judgement may also be required when using this information. The expected reduction in fatal and serious crash outcomes is often higher than the reduction in all casualties. As an example, BITRE (2012) found the impact on crashes from installation of roundabouts to be greater for higher severity outcomes:

• effect on property damage only: 52% reduction
• effect on casualties: 71% reduction
• effect on fatalities: 79% reduction.

Similar trends were seen in a European study by Jensen (2013). Therefore, using the casualty reduction will often lead to a conservative value for the expected reduction in fatal and serious injury.

The following matrix (Table 11.1) provides a basic summary of road safety treatment options and their effectiveness on some of the key crash types that result in fatal and serious injury. Detailed information on each of these treatments can be found in the documents referenced in Selecting Interventions in Intervention Option and Selection. Broad indicative costs are also provided.

Table 11.1: Road safety countermeasures matrix
1. head on
2. Junction
3. rear-end
4. run-off-road
5.Motorcycle
6.pedestrians

Note: $ = low cost; $$ = medium cost; $$$=high cost.

The following four case studies, from Puerto Rico, Portugal and Slovakia all show example of the effective use of interventions to improve safety.

As can be seen from Table 11.1 speed management is a treatment that is able to address most key crash types. Where speed has been identified as an issue, speeds can be lowered using effective speed management. This will result in fewer fatalities and serious injuries, provided compliance is high or additional enforcement measures are utilised.

Selecting Interventions

Interventions should be selected to suit a particular site, route or area, and to address the crash type occurring at that site, route or area. Crash types can be identified through reactive (crash based) or proactive identification methods (see Assessing Potential Risks And Identifying Issues).

Single interventions can be used, or more commonly combinations of interventions can be selected to combat a particular crash type or issue. The final intervention selection requires expert judgement about the factors that have contributed or may contribute to the occurrence of crashes.

A number of guides provide advice on appropriate interventions to address specific crash problems. PIARC (2009) provides a detailed set of options in the Catalogue of design safety problems and potential countermeasures. Advice is provided for road function, cross-section, alignment, intersections, public and private services, vulnerable road users, traffic Signing and marking and roadside features. PIARC published The role of Road Engineering in Combatting Driver Distraction and Fatigue Road Safety Risks to highlight Driver Distraction and Fatigue from the view of the safe systems approach. Figure 11.4 shows an example of Hierarchy Level 1 Treatments. The document outlines a hierarchy of treatement approaches and provide engineering solutions to address the problem of driving distraction and fatigue road safety risks. To highlight the the issue potential countermeasures for vulnerable road users the Vulnerable road users: Diagnosis of design and operational safety problems and potential countermeasures and appendix were published.

A number of interventions are presented that address safety issues relating to each of these topics. In each case, information is provided on the road safety problem for each issue. Treatment types are then presented, along with photos of the treatment, basic descriptions, indicative costs, crash types addressed, and the affected road user groups. The example below (Figure 11.4) shows potential solutions for issues related to unforgiving roadsides (categorised under Roadside Features).

Several online tools exist that provide similar, and in some cases more detailed information. Some cover a wide variety of treatment types, while others concentrate on particular crash types. Both Austroads (engtoolkit.com.au) and iRAP (toolkit.irap.org) have developed online toolkits that provide guidance on treatment options to address different road safety issues. They are both regularly updated and revised, capturing the most recent findings in road safety. Each includes detailed information on crash problem types and treatments, including indicative costs, safety and other benefits, implementation issues, and references.

The Austroads toolkit is designed to address safety issues identified through crash investigation and road safety audit (safety deficiencies). Detailed information is provided on solutions, including links to relevant design documents. The Road Safety Toolkit is targeted to those working in LMICs, and has been translated into a number of languages, including French, Arabic, Spanish and Mandarin. An image from this toolkit is shown in Figure 11.5.

Although the focus of this manual is on infrastructure interventions, it is important to ensure that multi-sector approaches (e.g. those involving education and enforcement) are considered, as these will often have a greater impact on safety than infrastructure measures alone. This is especially the case in LMICs where there may be lower levels of compliance, or less understanding by the general public regarding the intent of safety interventions. Issues specific to LMICs and intervention effectiveness are discussed in Intervention Effectiveness in LMICs in Intervention Option and Selection while the following case study provides an example of a combined infrastructure and education programme to address pedestrian safety in South Africa.

As another example, the SURE handbooks (particularly the handbook ‘Plan d’actions et realisation des actions’) provide guidance on the French methods for selecting safety interventions on the network (see Box 9.8).

Intervention Effectiveness in LMICs

Most of the information on intervention effectiveness is based on research conducted in HICs. Aside from research on behavioural interventions in LMICs, there are very few studies on the effectiveness in these countries, especially the effectiveness of road safety infrastructure treatments. This issue is significant, as it cannot be expected that interventions used in HICs will have the same outcomes when used in LMICs.

This issue has been highlighted in several studies. OECD/ITF (2012) suggests that there are many context/environment-related elements that influence actual crash reduction, and that this is a more critical consideration in LMICs. For example, provision of a hard shoulder might improve safety in a HIC. However, in a lesser developed country it might encourage improper use, such as the installation of stalls for selling items to travellers, which could decrease the overall safety of the roadway. An understanding of these issues of context is obviously critical to the successful implementation of safety treatments.

There also appear to be several barriers to the successful implementation of road safety infrastructure treatments. Turner & Smith (2013) conducted a workshop to identify issues around the implementation of infrastructure treatments in LMICs. Several effective infrastructure treatments were discussed, and barriers for each explored. Although a number of issues were identified for each treatment type, many of these fell into similar categories. These included cost, issues with compliance, design and implementation issues, public acceptance and maintenance.

Cost was raised as an issue for many of the treatments, although interestingly not for all. For some of the highly effective treatments, cost does not appear to be a significant issue. Of greater interest is that issues relating to compliance and design/implementation were raised for more of the treatments than cost. Compliance of treatments by road users is a significant issue in LMICs, and it is very likely that the treatment effectiveness will be lower as a result. Issues that suffered from this compliance issue were:

• roundabouts (resulting from failure to give way, and even vehicles circulating in the wrong direction);
• pedestrian crossings (failure of traffic to give way);
• pedestrian footpaths (obstruction and misuse by vehicles (including motorcycles));
• signalised intersections (failing to stop);
• shoulder sealing (misuse as an extra lane, parked vehicles, roadside trading);
• off-road cycle/motorcycle paths (obstructions on paths, misuse by inappropriate vehicles).

This issue of compliance indicates the need for a multi-sector response to the road safety issue. The use of safer infrastructure needs to be supported by appropriate education and enforcement, as discussed earlier in this manual.

Design and implementation issues were also thought to have an impact on the effectiveness of treatments. If a treatment is not well-designed and the installation is not of a high standard, the crash reduction potential will not be met. This was considered an issue for all of the treatments discussed. This situation can only be improved through improving the skills and capacity of those working in LMICs, including the sharing of knowledge regarding good practice.

Lastly, maintenance is also an issue that will impact on the crash reduction effectiveness of treatments. It is common for treatments to deteriorate to levels where they become less safe (or possibly to a point where they are higher risk than if the treatment was not present). Appropriate funding is required to ensure that treatments are maintained. Training may also be required regarding the issue of maintenance.

Although all of these issues are likely to be concerns for HICs, they are possibly more pronounced in LMICs, and will certainly have an impact on treatment effectiveness. The extent of this effect is not known, but it can be expected that because of these issues, treatment effectiveness is likely to be different (typically less) than when the same treatments are used in HICs.

Given the absence of good information on treatment effectiveness in LMICs, it is recommended that crash reduction values from HICs be used as a starting point when selecting treatments. However, the issues discussed above need to be carefully considered, and appropriate revisions made to expected benefits. In the longer term it is hoped that the knowledge base regarding treatment effectiveness in LMICs will improve, but this will only occur if appropriate monitoring, analysis and evaluation occurs in these countries (see Monitoring, Analysis and Evaluation of Road Safety). The case study from Puerto Rico shows the effective use of rumble strips to lower roadway departure crashes.

Innovative Safety Treatments

They will be required in order to help eliminate death and serious injury. Many safety treatments currently result in some residual serious crash outcomes, and so improvements to current treatments will be required. As treatments are applied to new situations (including in LMICs) there will be a need to adapt them for better outcomes. There are also a number of highly effective interventions that exist and are used in some countries but not at all in others. There is a need for road agencies and supporting organisations to be innovative and adopt new approaches, provided these are implemented based on an evidence-led approach. It is recommended that road agencies investigate new interventions and learn from overseas experiences.

Reasons that some highly effective treatments are not used in some countries include:

• lack of knowledge regarding the treatment and its effectiveness;
• lack of experience of how to install and maintain a treatment;
• issues regarding transferability and differences in local conditions (see Innovative Safety Treatments in Intervention Options and Selection);
• concern about legal liability if something goes wrong;
• concern about public understanding or acceptability.

Road agencies should be careful in their selection of new treatments, ensuring that they have been rigorously tested and have demonstrated safety benefits. Demonstration projects can be an effective way to assess promising treatments, and prepare for a wider roll-out (see Strengthening Capacity to Set and Deliver Targets.)

It is suggested that a methodological approach be taken towards innovation, outlined in the following steps:

• Know your problem. Identify the target crash type, the road user type and the locations that need to be targeted. See Assessing Potential Risks And Identifying Issues for more details on problem identification.
• Identify possible solutions. This can include solutions that are used overseas, or it can be an adaptation of an existing treatment. See earlier in this chapter for more guidance on intervention selection.
• Assess the solutions. It is important to research treatments thoroughly to ensure they are likely to be beneficial for safety outcomes as well as other policy objectives. This assessment can be based on documented experience from other road agencies. For newer treatments, driver simulators are sometimes used to determine the likely effects. In some instances, treatments can be installed in a controlled environment (e.g. off-road, or in a low speed area) to determine the likely effects.
• Trial the selected solution. A demonstration project can be an effective way to test the treatment within a specific context and in a controlled environment. This can also help prepare for a wider roll-out.
• Monitor, analyse and evaluate the trial. Ensure that the outcomes are as expected, and that there are no adverse effects to any road user’s safety. This evaluation should include an assessment of the cost effectiveness of the new treatments, especially when compared to existing option. See Intervention Selection And Prioritisation for information on economic appraisal, and Monitoring, Analysis and Evaluation of Road Safety for more information on monitoring, analysis and evaluation.
• Roll out the solution on a wider scale. Continue to monitor and evaluate the treatments, including crash analysis once sufficient data has been collected. Include design and operational information in guidance documents.
• Inform others. If the new treatment is effective, it is important to let others know of this. Information on treatments that have not performed well is also very important for the international road safety community.

A number of the reference documents highlighted earlier in this chapter provide innovative examples of road safety infrastructure treatments. Some agencies actively promote certain treatments (including innovative ones) that they would like to see used more often (e.g. FHWA, 2015, which documents and promotes proven safety countermeasures; while information on innovative intersection design including videos can be found here: http://www.fhwa.dot.gov/everydaycounts/edctwo/2012/geometrics.cfm). Many national and local studies have been undertaken to assess innovative treatments with promise. These studies are usually undertaken by universities and research institutes. Information on these trials can be found in international journals and at conferences, although care should be taken to ensure that such information is robust.

The case study below provides an example of innovative use of intelligent transport systems (ITS) in Thailand and Rapid Way narrow cross section in Hungary.