This appendix lists evidence statements from two reviews, provided by external contractors (see appendix A) and links them to the relevant recommendations. (See appendix B for the key to quality assessments.) The evidence statements are presented here without references – these can be found in the full review (see appendix E for details). It also sets out a brief summary of findings from the economic analysis.
Evidence statement number B1a indicates that the linked statement is numbered 1a in the review 'Barriers to, and facilitators of, the prevention of unintentional injury in children on the road'.
Evidence statement number E1a indicates that the linked statement is numbered 1a in the review 'Systematic reviews of effectiveness and cost-effectiveness of road and street design-based interventions aimed at reducing unintentional injuries in children'. The reviews are available online.
Where a recommendation is not directly taken from the evidence statements, but is inferred from the evidence, this is indicated by IDE (inference derived from the evidence) below.
Recommendation 1: E1a–c, E2a–c, E3a–e
Recommendation 2: B2a–d, E10, E11; IDE
Recommendation 3: E1, E1a–c, E2, E2a–c, E3, E3a–e, E6b, E10, E11
Recommendation 4: E7, E7a, E7b, E9
Please note that the wording of some evidence statements has been altered slightly from those in the review team's report to make them more consistent with each other and NICE's standard house style.
Five UK based studies evaluated area-wide traffic-calming schemes. There were one controlled (+) and three uncontrolled (one [−] and two [+]) before and after studies, and one ecological study (+). Within these studies, casualties, injury collisions and speed outcomes were reported.
There is moderate evidence from two uncontrolled before-and-after studies (both UK) that area-wide traffic-calming may reduce rates of killed or seriously injured children (both [+]). Both studies showed reductions in either killed and seriously injured child casualties or collisions in which a child pedestrian or cyclist is killed or seriously injured, but none of these was statistically significant.
There is moderate evidence from one uncontrolled before-and-after study and one ecological study (both UK) that area-wide traffic calming may reduce child road casualty rates of any severity (both [+]). There is moderate evidence from one controlled and two uncontrolled before-and-after studies (all UK) that area-wide traffic calming may reduce child injury collision rates of any severity (one [−] and two [+]).
Of the two studies that reported child casualty rates, one ecological study showed a statistically significant reduction (rate ratio [RaR] = 0.777 for pedestrians in one of two cities studied, p = 0.002 [+]), while the results in the other city, and the uncontrolled before-and-after study are consistent with a reduction, but do not reach significance (+).
The three studies that reported child injury collision rates (one controlled and two uncontrolled before-and-after studies, all UK) also show reductions, but only one approaches statistical significance when compared with a control group (RaR = 0.524; 95% confidence interval [CI] = 0.258, 1.062 for child cyclists; one [−] and two [+]).
There is weak evidence from two uncontrolled before-and-after studies that area-wide traffic calming may reduce traffic speeds (one [−] and one [+]).
With the possible exception of the much older study (1990), this evidence is judged as directly applicable to similar roads and/or communities in the UK.
Three UK-based studies evaluated single road traffic-calming schemes. These were all uncontrolled before-and-after studies (three [+]). Within these studies, casualties, injury collisions and speed outcomes were reported.
There is weak evidence from two UK-based uncontrolled before-and-after studies to show that single road traffic calming may reduce child road casualty rates. Only one of these studies showed a statistically significant reduction in child casualties from 12 to zero (p < 0.001 [+]). In the other study, numbers of casualties were too small (decreasing from three to zero) to be meaningful (+).
There is weak evidence from one UK-based, uncontrolled before-and-after study that single road traffic calming may reduce child pedestrian injury collision rates (RaR 0.0381, p < 0.001) while child cyclist injury collision rates were also reduced, but non-significantly (RaR = 0.632, p = 0.081 [+]).
There is weak evidence from two uncontrolled before-and-after studies that single road traffic calming may reduce traffic speeds (both [+]). This evidence is judged as directly applicable to similar roads and/or communities in the UK, although the Chorlton evidence is dated.
Four UK-based studies evaluated 20 mph zones (mostly in urban areas). There were one controlled and three uncontrolled (all [+]) before-and-after studies, one of which was adjusted for background trends. There is some overlap between studies. Two of the studies are of 20 mph zones in London; one of which essentially updates the other. There are also small overlaps between these London-based studies and the England-wide study, and potentially between the England-wide study and the study based in Hull. Within these studies, casualties and speed outcomes were reported.
There is moderate evidence from two uncontrolled before-and-after studies (one adjusted for trends on background roads; both UK-based) that 20 mph zones reduce killed or seriously injured child casualty rates (RaR = 0.242, to 0.859 depending on analysis and study, p < 0.05 where recorded [++]). One controlled before-and-after study also showed a reduction in killed or seriously injured child casualty rates in the intervention group when compared with a control group; however, this reduction was not significant (+). It must be noted that this study also evaluated schemes in London and is essentially updated by this uncontrolled before-and-after study.
There is weak evidence from one uncontrolled before-and-after study (London-based), which was adjusted for trends on background roads, that 20 mph zones may reduce child pedestrian killed and seriously injured casualty rates. However, this reduction was not significant once the results had been adjusted for changes in background trends on outside roads (+). One study also showed that 20 mph zones may reduce child pedestrian killed and seriously injured casualty rates (before and after data only reported for this outcome; RaR 0.394, p < 0.001 [+]). As noted above, however, this study is essentially updated by the uncontrolled before-and-after 2008 study. The evidence shouldn't therefore be 'counted' twice.
There is weak evidence from one before-and-after study (controlled data only reported for this outcome) that 20 mph zones may reduce child cyclist killed or seriously injured casualty rates. This reduction approaches statistical significance (RaR = 0.399, p = 0.06 [+]).
There is moderate evidence from three UK-based uncontrolled before-and-after studies (one using adjusted analyses [+]) and one controlled before-and-after study of London schemes (+) that 20 mph zones may reduce child road casualty rates overall, and for child pedestrians and child pedal cyclists when analysed separately (road casualty rates overall RaR = 0.331 to 0.716 depending on analysis and intervention, p < 0.001 where recorded).
There is weak evidence from two studies that 20 mph zones may reduce traffic speeds (both [+]). This evidence is judged as directly applicable to similar roads and/or communities in the UK, although some data is rather dated.
There is weak evidence from one case-control study that living in an area with 0–5 streets with a speed limit of 30 kph may increase a child's risk of injury compared with a child living in an area with 15 or more streets with the same speed limit (OR = 5.3, 95% CI = 1.6, 17.6 [+]).
There is moderate evidence from two controlled before-and-after (injury data time-series) studies (both [+]) in the USA that Safe Routes to School (SRTS) programmes based predominantly on engineering measures may reduce the rates of crash-involved child pedestrians or cyclists, or the rate of child injury road collisions.
In 125 SRTS project areas across California, and after assuming modest (10%) increases in rates of walking and cycling to school due to the programmes (such as increased exposure), a mean reduction of 7% in the all‑injury collision rate with child pedestrians and cyclists was estimated (14% for children aged 5–12) (+). However, the estimated impact on fatal or severe child injuries was less conclusive (ranging from a 52% increase to a 24% reduction, again depending on assumed changes in levels of walking/cycling to school).
The evaluation of 53 projects in three unnamed US States (+) compared linear regression coefficients (giving 'T statistics') between the time-series trends of child injury data for the SRTS sites; these showed significantly greater reductions in crash-involved child pedestrians and cyclists at SRTS sites when compared with at least two of the six 'control' time-series in all three US states (note, all of the 'T' values were negative, indicating that the reductions in crash outcomes in SRTS sites were always lower [if not always statistically significantly lower] than in the comparison time-series.)
This evidence from evaluations of SRTS programmes in the US is judged as partially applicable to similar localities in the UK.
There is weak evidence from one controlled before-and-after study that combined traffic calming, safe routes to schools and education may reduce child road casualty rates when a before-and-after comparison was made (OR 0.722, p = 0.007 [+]); however, compared with the control group the reduction was not significant. This Swedish evidence is judged as partially applicable to similar roads and/or communities in the UK.
There is moderate evidence from three cost–benefit analyses of a variety of area-wide traffic-calming schemes that show that, even in the short term (after 1 year), benefits are likely to exceed costs in most circumstances. However, there was considerable variation in first year rates of return. This evidence was judged to be partly applicable to the UK road setting.
There is moderate evidence from one cost–benefit analysis of advisory 20 mph speed limits that shows that, in the short term, benefits are likely to exceed costs. Similarly, there is moderate evidence from one cost–benefit analysis of mandatory 20 mph zones that shows that, in the medium to long term, benefits are likely to exceed costs. The evidence on 20 mph zones is judged as being directly applicable to other urban roads in England, whereas the applicability of the evidence on advisory speed limits in Scotland may have less applicability in England and Wales because of the different road regulations relating to 20 mph speed limits.
Five studies, four UK and one US-based, discuss risk-taking behaviour among children and young people as a potential cause of collisions (two [+] and three [−]).
Like adults, children and young people often engage in 'common' risk behaviours that are seen as part of everyday life, such as not always using crossings, crossing between parked cars or in traffic.
One UK study reports that teenagers were more likely to take risks on the road than younger children (aged 8+).
Three UK studies suggest that a minority of children and young people engage in 'extreme' risks – such as playing 'chicken' in the road or holding onto the back of buses, and that boys are more likely to do this, and to encourage such behaviour in each other. Such behaviours are regarded in a similar way to thrill-seeking sports.
The economic modelling suggests that setting advisory 20 mph limits is a highly cost-effective way of preventing unintentional injuries on the road (with a net present value [NPV] of £64,209). However, caution is needed in interpreting these results. First, because the studies modelled came from Scotland where the legal definition of what comprises an advisory 20 mph limit is different. Second, the areas where they were introduced were not necessarily comparable for example, in terms of previous collision rates, vehicle speeds and pedestrian usage.
Mandatory 20 mph zones were found to be much more cost effective in areas with previously high casualty rates (1.6 per year per km). The NPV was £90,625 in these areas, compared to a NPV of £25,480 when implemented in low casualty rate areas.
Deterministic sensitivity analyses identified a number of parameters that were important to all interventions: number of casualties in the comparator area, effectiveness of the intervention, the background reduction in casualties and the effective lifetime of the intervention all had an effect.
The modelling did not consider health benefits other than those due to casualties prevented. For instance, it did not consider the health benefits of increased physical activity or the reduction in air pollution due to a change in road use. It also did not consider the impact that road engineering measures to reduce speed can have on other factors such as congestion or noise. These would only be captured by cost-benefit analysis using a full societal perspective.
Fieldwork aimed to test the relevance, usefulness and the feasibility of putting the recommendations into practice. PHIAC considered the findings when developing the final recommendations. For details, go to the fieldwork section in appendix B and online.
Fieldwork participants who work in transport planning were fairly positive about the recommendations and their potential to help prevent unintentional injuries. Many participants stated that although the recommendations were 'nothing new' and were already being applied in general, engaging the health sector in this area was vitally important. They reported that the recommendations could have a positive impact if they helped to engage the health sector in preventing injuries among young people aged under 15.
Some stakeholders said that at partnership meetings (for example, local strategic partnerships) in their local areas the health aspect of the partnership was often missing, that is, the NHS representatives do not attend local injury prevention partnership meetings. Participants particularly noted NHS directors of public health as an important group to engage and ensure that they were active in work in this area. Others noted that stakeholders from education were also unable or unwilling to engage and that it would be an advantage if the recommendations could help engage education services in injury prevention.
The planning stakeholders consistently said in the workshops and focus groups that the recommendations do not add any content for planners. They said that the draft recommendations do not cross-reference to other supporting material, such as that from the Department for Transport. They also thought some of the content of the recommendations is not consistent with Department of Transport guidance and strategies (for example the road safety strategy). Delegates stated that a cohesive approach from government is important.
Overall, delegates said that the recommendations had a role to play in directing the NHS to plays its part in supporting local strategic partnerships and supplying data requirements.
 Net present value (NPV) determines the total monetary benefit of an intervention less its costs (compared with an alternative intervention) when discounted to its present value. A positive NPV occurs when the sum of the discounted benefits exceeds the sum of the discounted costs.