Extended Research Article

i

• Toolkit

• • Download report PDF
  • Download report documents
  • Download report documents
  • Disclosure of interest
  • Download report XML
  • Citation Tools
  • Print
  • Responses to this report (3)
  • Permissions information

View Award

This study found that speed limit interventions that use signs and lines plus education and promotion can reduce casualties, and have significant public support and compliance once implemented.

{{author}}{{author}}{{($index > metadata.AuthorsAndEtalArray.length-1) ? ',' : '.'}}

Ruth Jepson, Graham Baker, Claire Cleland, Andy Cope, Neil Craig, Charlie Foster, Ruth Hunter, Frank Kee, Michael P Kelly, Paul Kelly, Karen Milton, Glenna Nightingale, Kieran Turner, Andrew James Williams & James Woodcock.

• Detailed Author information
• Detailed Author information

Ruth Jepson ^1,*, Graham Baker ², Claire Cleland ³, Andy Cope ⁴, Neil Craig ⁵, Charlie Foster ⁶, Ruth Hunter ³, Frank Kee ³, Michael P Kelly ⁷, Paul Kelly ², Karen Milton ⁸, Glenna Nightingale ¹, Kieran Turner ^1,2, Andrew James Williams ⁹, James Woodcock ¹⁰

¹ Scottish Collaboration for Public Health Research and Policy, University of Edinburgh, Edinburgh, UK
² Physical Activity for Health Research Centre, University of Edinburgh, Edinburgh, UK
³ School of Medicine, Dentistry and Biomedical Sciences, Queen’s University Belfast, Belfast, UK
⁴ Sustrans, Bristol, UK
⁵ Public Health Scotland, Edinburgh, UK
⁶ Centre for Exercise, Nutrition and Health Sciences, University of Bristol, Bristol, UK
⁷ Department of Public Health and Primary Care, University of Cambridge, Cambridge, UK
⁸ Norwich Medical School, University of East Anglia, Norwich, UK
⁹ School of Medicine, University of St Andrews, St Andrews, UK
¹⁰ Centre for Diet and Activity Research, University of Cambridge, Cambridge, UK
^* Corresponding author Email: ruth.jepson@ed.ac.uk

Declared competing interests of authors: Ruth Hunter is a member of the Public Health Research (PHR) Funding Board. Frank Kee is a co-investigator on the Game of Stones trial (PHR 14/185/09 and NIHR129703), the Supporting MumS trial (NIHR131509), the Global Health LINKS Research Group (NIHR 16/137/85) and Improving the Oral Health of Older People in Care Homes: a Feasibility Study (TOPIC) (NIHR 17/03/11). He is a principal investigator of the MECHANISMS study (MR/RO11176/1) and the Healthy Urban Living and Ageing in Place (HULAP) study (GCRF-GIAA18-19). Furthermore, he is a member of the following panels: Medical Research Council (MRC) Public Health Intervention Development (PHIND) Funding Panel (2013–18); MRC Better Methods, Better Research Panel (2020–present); MRC Non-clinical Fellowship Panel (2020–present); UK Research and Innovation Future Leaders Fellowship Panel (2020–present); Agile COVID Panel (2020–21); Policy Research Unit Commissioning Panel (2016 and 2018); Long COVID Panel (2021); ADD (‘Our Future Health’ study) Advisory Board (2020–present); School of Public Health Advisory Board (2018–present); MRC Longitudinal Studies Funding Panel; and Methods Advisory Group. He was also a member of the PHR Funding Board (2009–13; chairperson 2014–19). Michael P Kelly received grants from the Wellcome Trust, The Dunhill Medical Trust, the National Institute for Health and Care Research (NIHR), the Arts and Humanities Research Council (AHRC) and Marie Curie, and received NIHR and AHRC consultancy fees. He is a member of the Scientific Advisory Board Systems Science In Public Health Economic Research (SIPHER), University of Sheffield. Andrew James Williams received a grant from Sustrans/Transport for Scotland for £15,255 to conduct a systematic review into the association between modes of travel and loneliness/social isolation (McHale C, Williams A, Cormie V. Systematic review of research investigating the relationship between social disconnection and transportation activities. PROSPERO 2021 CRD42021232445 URL: www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42021232445) (5 months from November 2020).

Funding: {{metadata.Funding}}

{{metadata.Journal}} Volume: {{metadata.Volume}}, Issue: {{metadata.Issue}}, Published in {{metadata.PublicationDate | date:'MMMM yyyy'}}

https://doi.org/{{metadata.DOI}}

Citation:{{author}}{{ (($index < metadata.AuthorsArray.length-1) && ($index <=6)) ? ', ' : '' }}{{(metadata.AuthorsArray.length <= 6) ? '.' : '' }} {{(metadata.AuthorsArray.length > 6) ? 'et al. ' : ''}}. {{metadata.JournalShortName}} {{metadata.PublicationDate | date:'yyyy'}};{{metadata.Volume}}({{metadata.Issue}})

Report Content

Responses to this report

Response by Dr Andrew Williams (SCPHRP), Dr Glenna Nightingale (SCPHRP), Mr Neil Craig (Yunus Centre at Glasgow Caledonian University) on 22 July 2024 at 4:16 PM
Author Repsonse to comment
Thank you for your interest in and comments on our research. In the process of applying for the funding for this project, we established that the likelihood of being able to detect a statistically significant change in collisions or causalities was very low, and this was made clear to the funder. The absolute numbers represented by the percentages in the abstract are found in Tables 18 (Edinburgh) and 19 (Belfast) and the related text on pages 69 and 71). We took the decision theoretic approach (DTA) described in the report (Chapter 6) to the economic evaluation of the 20 mph speed limits, having recognised that it would be very unlikely that statistically significant changes in collisions and casualties would be observed. We did not carry out a full cost effectiveness or cost benefit analysis in the study. Using the DTA approach we took the measured costs of the intervention and the uncertainty around effectiveness and overall benefits to make an assessment around how likely is it that the benefit would exceed the cost. We concluded it was likely but we acknowledged the uncertainty. Transport networks and more widely cities are systems with many interdependencies which makes the application of traditional statistical methods particularly challenging. Many of the statistical methods used to assess effectiveness and cost-effectiveness make assumptions around each individual involved being independent, when on the road our behaviours are dependent on the others around us (drivers, pedestrians, etc.). Decision and policy makers routinely face the challenge of making decisions when there is a lot of uncertainly as most social situations are systems, while statistical methods were developed for use in experimental trials. Research, including this study is increasingly seeking to identify methods which address these limitations of traditional methods. Finally, regarding the concern about the third paragraph on page 88, the key consideration here is that 20 mph speed limits are an example of a preventive intervention. The main cost of preventive interventions is often in the initial implementation, installing the signage and the awareness raising campaign in the case of 20 mph speed limits. After implementation the costs reduce to maintenance and enforcement, however the benefits continue to accumulate. Economic evaluation requires that the profile of current and future costs is compared to the profile of current and future benefits. Therefore, it was appropriate that we took into consideration the potential reduction in collisions and casualties over a longer timeframe, in this case the follow up period of the study. The DTA approach identified that for the costs of the 20 mph limits in Edinburgh preventing at least 1.35 fatal casualties over any time period would make the investment cost effective. You are correct that the reduction of 11 casualties is comparing two different time periods, so not appropriate. However, if we translated the 14 casualties in 36 months (3 years) down to a 22 month period to match the follow-up period that would be somewhere between 8-9 casualties prevented. Subsequently comparing to the follow-up period (3 fatal casualties in 22 months) the reduction in fatal casualties would be 4-5, which is more than the 1.35 required to make the intervention cost effective. Inclusion of the monetary value of non-fatal casualties prevented would reinforce this conclusion. In practice, to the extent that 20mph limits continue in the future to reduce speeds and reduce collisions, fatalities and casualties, then the benefits may continue to accumulate over time. Therefore, the cumulative figure in the study is potentially an underestimate of the total benefit derived from the initial investment in 20mph limits. Annual rates (converted into fatalities/casualties prevented per year) would give a partial picture of total benefit over time arising from the initial investment in 20mph limits. Overall, 20 mph speed limits in urban and residential areas are relatively cheap interventions which have the potential to benefit the health of the public in multiple ways, but as we acknowledged, assessing the scale of these benefits is not easy or simple.

Response by Simon Newton on 2 April 2024 at 3:13 PM
Response to correction

The posting of the above “Corrections” response for this article helps a lot in its understanding, and the authors should be commended for taking the time to do it. However, as a “member of the public”, I remain concerned about a small aspect of its presentation that I believe has large consequences, and that is with respect to the results for the Edinburgh 20 mph speed limit implementation. The abstract indicates a percentage reduction for fatalities of 23% but does not give the raw data behind the figure (a reduction from 5 to 4 per annum as rounded-up figures; or from 4.67 to 3.61 using the raw data). It is not hard to find on the internet quotations by various parties based on this research paper. I found one in a local Edinburgh paper, and one by a local Edinburgh councillor. Both of them reported the reassuringly high reduction in fatalities of 23%... but neither pointed out that this was actually a reduction of 1 per annum, from a very low level of 5 per annum. Any statistician would tell you that this cannot possibly be a statistically significant reduction, because the sample size is too small. Furthermore, there is unlikely ever to be a statistically significant reduction in the fatality rate owing to the “floor effect” where reduction of an already low number is very hard to achieve. As a consequence of the abstract’s presentation (lacking raw data, and without statistical significance indicated) misleading claims have been made by people of influence to the public for the 20-mph speed reduction scheme regarding fatalities. Apart from misleading the public, the non-significance of the fatality rate reduction means that the cost-effectiveness calculations of the scheme are flawed, and it is probable that the scheme is in fact not cost effective according to the calculation methods used. This has policy and political consequences. I would also like to point out a sentence (on page 88, third para.) which I think is incorrect, and is related: “The reduction of 11 casualties observed in the study (see Table 18) is well in excess of the 1.35 required for the monetary value of the fatalities prevented to exceed the costs of designing and implementing the 20-mph limits.” In fact, the figure “11” comes from the raw data, 14 minus 3, with 14 being the number in the 3-year pre-implementation period, and the 3 being the number in the 22-month post-implementation period. When annual rates are calculated, these convert to the figures 5 and 4 mentioned above, and used the for the final statement for fatality reduction (23%). It is simply wrong to quote the cumulative figure of “11” (especially without explicitly mentioning the time period); if anything, the annual rate should be quoted. 

Response by Dr Andrew Williams on 14 March 2024 at 9:01 AM
Corrections to published report

Corrections to the report content since publication. Please see Response to Report document in the Download report documents tab.