Article Text

Download PDFPDF

Accuracy of self-report of on-road crashes and traffic offences in a cohort of young drivers: the DRIVE study
  1. Soufiane Boufous1,
  2. Rebecca Ivers1,
  3. Teresa Senserrick1,
  4. Mark Stevenson1,
  5. Robyn Norton1,
  6. Ann Williamson2
  1. 1The George Institute for International Health, Sydney, Australia
  2. 2Faculty of Science, University of New South Wales, Sydney, Australia
  1. Correspondence to Soufiane Boufous, Injury Division, The George Institute for International Health, PO Box M201, Missenden Road, NSW 2050, Australia; sboufous{at}george.org.au

Abstract

In order to determine the accuracy of self-report of on-road crashes and traffic offences among participants in the DRIVE study, 2991 young drivers in New South Wales, Australia who completed the follow-up questionnaire were asked whether they had been involved in an on-road crash or were convicted for a traffic offence while driving during the year prior to the survey. This information was linked to police crash data to determine the level of accuracy of self-report of on-road crashes. There was a high level of accuracy in young drivers' self-report of police recorded crashes (85.1%; 95% CI 78.2% to 92.1%) and of police recorded traffic offences (83.0%; 95% CI 79.4% to 86.6%).

Results suggest that surveys may be useful tools for estimating the incidence of on-road crashes and traffic offences in young drivers. The findings are particularly relevant to jurisdictions where access to administrative data is limited.

  • Traffic crashes
  • self-report
  • young drivers

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Road traffic injuries are a leading cause of death and disability worldwide, particularly among young adults.1 Surveys are widely used to measure the incidence of road related crashes and offences, particularly in low and middle income countries where access to administrative data, such as police or hospital records, is limited. Despite the reliance on self-reports and the importance of good quality data to provide estimates of road related crashes and offences, little is known about the validity of self-reported on-road crashes, particularly among young drivers.

While there have been a couple of studies that assessed the validity of self-report of road crashes and traffic offences,2–4 they all had a relatively long recall period of at least three years. In addition, the only study that examined this issue among young adults had a small sample size which might have affected its findings.4

In New South Wales (NSW), Australia, it is a legal requirement to record most road crashes and traffic offences in the state police database. This information can be used as a reference for assessing self-report of road related crashes and traffic offences. The aim of this study was to determine the accuracy of self-report of road crashes and traffic offences based on data linkage of a survey of a cohort study of young drivers with police crash data in NSW.

Methods

The DRIVE study is a prospective cohort study designed to investigate risk factors for motor vehicle crashes in young drivers. Initially, 20 822 NSW drivers aged 17–24 years and holding their first-stage provisional motor vehicle licence responded to a baseline survey between June 2003 and December 2004. A follow-up self-administered questionnaire was sent to a random sample of 5000 participants two years later. More detailed information regarding the methodology of the study is described elsewhere.5

A total of 2991 participants (60%) completed the follow-up questionnaire, which, in addition to demographic information, contained questions related to driving experiences, driving habits and attitudes over the ensuing period. Additional items queried whether they had been involved in an on-road crash or were convicted for a traffic offence while driving during the year prior to the survey.

All respondents gave written consent for their survey data to be linked (using driver licence numbers) to police crash data held by the NSW Roads and Traffic Authority (RTA). On-road crashes are required to be reported to police and recorded in the RTA database if a person is killed or injured; if drivers involved in the crash do not exchange particulars; and if a vehicle involved in the crash was towed away. Crashes and traffic offences recorded in the police data and that occurred in the year preceding the follow-up survey were selected for the study.

As police data do not include minor road crashes, which might be reported by the study participants, it was not deemed appropriate to estimate the specificity of self-reported events. The analysis focused instead on the sensitivity, also referred to in this paper as the accuracy, of self-reports of road crashes and traffic offences and which was determined by computing the proportion of police recorded events that were reported by young drivers during the year preceding the follow-up survey.

Accuracy of self-report of road crashes and traffic offences was computed across sex, age and socioeconomic status (SES) and 95% CI intervals were calculated for all proportions. SES of participants was computed by linking postcodes of residence to the Australian Socio-Economic Indexes for Areas.6 All analyses were carried out using SAS V.9.1.7

Results

Young drivers who completed the DRIVE study follow-up survey reported more crashes and traffic offences than those recorded in the police database during the one-year study period. Respondents reported five times more road crashes than those captured in the police records (table 1). The difference was not as great for traffic offences, with respondents reporting around 30% more infringements than those recorded in police data (table 1).

Table 1

Police recorded and self-reported road crashes and traffic offences during 1 year prior to resurvey

Overall, 85.1% (95% CI 78.2% to 92.1%) of police recorded crashes were reported by the survey participants. This figure was slightly higher among women (86.3%) compared to men (82.0%), and decreased with increasing age, with 60% of respondents aged ≥22 years accurately reporting crashes compared to 89.2% of those aged 18–19 years. Similarly, accuracy of self-report of crashes in young drivers decreased with decreasing SES level, with 87.5% of those in the highest quartile reporting crashes recorded in police data compared to 78.3% among those in the lowest quartile (table 2). However, none of the differences in the accuracy of self-report of crashes by demographic characteristics examined was statistically significant.

Table 2

Accuracy of self-report of road crashes by demographic characteristics

Similar results were found when examining the accuracy of self-report of traffic offences. Overall, 83.0% (95% CI 79.4% to 86.6%) of young drivers reported traffic offences recorded in police data. The accuracy of self-report was slightly higher in women (83.1%) than in men (82.5%), and decreased with increasing age and decreasing SES (table 3). With regard to crashes, none of the differences in the accuracy of self-report of road traffic offences by demographic characteristics was statistically significant.

Table 3

Accuracy of self-report of traffic offences by demographic characteristics

Discussion

The findings show that a high proportion of road crashes (85%) and road traffic offences (83%) recorded in the police records during the year prior to the survey were reported by young drivers using a self-administered questionnaire. The accuracy of self-report of road crashes was much higher than that reported in two studies that examined this issue in older drivers, aged ≥60 years, which found that between 63% and 69% of crashes recorded in police data were reported by study participants.2 3 Apart from the longer recall period of five years in both studies, recall of recent events in older people is often less accurate than in younger adults.8

The only previous study that is comparable to our study, in terms of the outcome examined as well as the age group, is a cohort study of young adults aged 18–21 years carried out in New Zealand, which reported that half of the crashes recorded in police data had been reported by study participants.4 The study, however, only examined a small number of events and had a much longer recall period of three years, which may have increased recall bias. In addition, the survey information in the New Zealand study was obtained through a face-to-face interview, which might have influenced young participants' reporting of adverse events such as road crashes. Previous research has highlighted the social desirability issues (respondents' desire to present themselves in a favourable light) associated with face-to-face interviews, which leads to increased reporting bias.9 It is also possible that participants' foreknowledge that survey data would be linked to police data, might have resulted in the observed high accuracy rates.

Previous studies that examined self-report of various conditions requiring hospitalisation, including injury, found that the level of accuracy of self-report was inversely proportional to the level of education and SES of participants.10 11 In this study, while the accuracy of self-report of road crashes and traffic offences seems to decrease with decreasing SES, this association was not statistically significant.

Young drivers reported more road crashes and traffic offences than those recorded in the police data. In the case of crashes, this may be due to the fact that many self-reported incidents were minor and therefore not recorded in the police data. Alternatively, some might have occurred outside the state's border and were not captured in the state's police records.

Similarly, the ‘over-reporting’ of traffic offences might be due to some young drivers reporting minor offences, such as parking related offences, which are not recorded in police data. The self-report of such minor offences seems to occur despite the study researchers' efforts to be specific about the type of traffic offences to be reported (those committed while driving).

Paradoxically, young drivers failed to report 15% of crashes and 17% of traffic offences recorded in police data. This might be due to young drivers' unwillingness to divulge information regarding some types of crashes and offences, particularly serious ones, because of perception of consequences such as loss of license or increases in insurance premiums. It is also possible that some young drivers either could not recall the event or incorrectly thought the event occurred more than a year prior to the interview.

Of the randomly selected study participants, 60% completed the follow-up survey. While response bias might have influenced the findings, we have compared socio-demographic characteristics recorded in the baseline questionnaire between those who participated and those who did not, but found no significant differences.

Conclusions

The results show high levels of accuracy in young drivers' self-report of police-recorded road crashes and traffic offences. Surveys may therefore be very useful tools for estimating the incidence of road crashes and traffic offences in young drivers, particularly when short recall periods (less than one year) are utilised. The findings are particularly relevant to jurisdictions where there is a high dependence on population surveys in the absence of reliable administrative crash data to inform the development and implementation of preventive strategies.

What is already known on the subject

  • Surveys are widely used to estimate on-road crashes and traffic offences.

  • Previous studies which used surveys with long recall periods (over three years), found that participants failed to report between one third to half of crashes recorded in police records.

What this study adds

  • Young drivers' self-report of police-recorded road crashes and traffic offences, that had occurred one year prior to the survey, were highly accurate (well above 80%).

  • Surveys may be very useful tools for estimating the incidence of road crashes and traffic offences in young drivers, particularly when short recall periods are utilised.

  • The findings are relevant to jurisdictions where there is a lack of access to reliable administrative data on road crashes and traffic offences.

References

Footnotes

  • Funding The DRIVE Study was funded by the National Health and Medical Research Council of Australia, Roads and Traffic Authority of New South Wales, NRMA Motoring and Services, NRMA-ACT Road Safety Trust, NSW Health, and the Motor Accidents Authority of New South Wales. SB, RI, TS, MS and AW receive salary funding from the National Health and Medical Research Council of Australia.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval The study was approved by The University of Sydney Human Research Ethics Committee and the New South Wales Health Ethics, Sydney, Australia.

  • Provenance and peer review Not commissioned; externally peer reviewed.