Dataset Information

Differential impacts of ridesharing on alcohol-related crashes by socioeconomic municipalities: rate of technology adoption matters

ABSTRACT:

Background

An emergent group of studies have examined the extent under which ridesharing may decrease alcohol-related crashes in countries such as United States, United Kingdom, Brazil, and Chile. Virtually all existent studies have assumed that ridesharing is equally distributed across socioeconomic groups, potentially masking differences across them. We contribute to this literature by studying how socioeconomic status at the municipal level impacts Uber’s effect on alcohol-related crashes.

Methods

We use data provided by Chile’s Road Safety Commission considering all alcohol-related crashes, and fatal and severe alcohol-related injuries that occurred between January 2013 and September 2013 (before Uber) and January and September 2014 (with Uber) in Santiago. We first apply spatial autocorrelation techniques to examine the level of spatial dependence between the location of alcohol-related crashes with and without Uber. We then apply random-effects meta-analysis to obtain risk ratios of alcohol-related crashes by considering socioeconomic municipality differences before and after the introduction of Uber.

Results

In both analyses, we find that the first 9 months of Uber in Santiago is associated with significant rate ratio decreases (RR = 0.71 [95% Confidence Interval (C.I.) 0.56, 0.89]) in high socioeconomic municipalities in all alcohol-related crashes and null (RR = 1.10 [95% C.I. 0.97, 1.23]) increases in low socioeconomic municipalities. No concomitant associations were observed in fatal alcohol-related crashes regardless of the socioeconomic municipality group.

Conclusions

One interpretation for the decline in alcohol-related crashes in high socioeconomic municipalities is that Uber may be a substitute form of transport for those individuals who have access to credit cards, and thus, could afford to pay for this service at the time they have consumed alcohol. Slight increases of alcohol-related crashes in low socioeconomic municipalities should be studied further since this could be related to different phenomena such as increases in alcohol sales and consumption, less access to the provision of public transport services in these jurisdictions, or biases in police reports.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-021-12066-z.

SUBMITTER: Blazquez C

PROVIDER: S-EPMC8569979 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundThere is growing evidence that a range of pre-injury, injury related and post-injury factors influence social and health outcomes across the injury severity spectrum. This paper documents health related outcomes for people with mild, moderate and severe injury after motor vehicle crash (MVC) injuries in New South Wales, Australia.MethodsThis inception cohort study followed 2019 people injured in MVCs, for 6 and 12 months post-injury. We categorised moderate injury as hospital length-of-stay (LOS) of 2-6 days and Injury Severity Score (ISS) of 4-11, while severe injury as LOS ≥7 days or ISS ≥ 12. We examined differences in paid work status, 12-Item Short Form Survey (SF12), EQ-5D and World Health Organisation Disability Assessment Schedule II (WHODAS) outcomes longitudinally from baseline to 12 months between levels of injury severity using linear mixed models for repeated measures. We first considered minimally sufficient adjustment factors (age, sex, crash role, perceived danger in crash, pre-injury health, pre-injury EQ-5D, recruitment source), and then more extensive adjustments including post-injury factors. The presence of mediating pathways for SF-12 Physical Component Summary (PCS) and Mental Component Summary (MCS) via post-injury factors was evaluated.ResultsBased on hospital length of stay (LOS), 25 and 10% of participants sustained moderate and severe injuries, respectively, while 43 and 4% had these injuries based on ISS. Twelve months post-injury LOS ≥7 days versus ≤1 day was associated with an estimated 9 units lower mean SF12 PCS using a minimally sufficient adjustment model, and LOS ≥ 7 days was associated with a 3 units lower mean SF12 MCS score. Mediation analyses (LOS ≥ 7 days vs ≤1 day) found for SF12 MCS outcomes, effects of injury severity were small and mostly indirect (direct effect - 0.03, indirect effect - 0.22). Whereas for SF12 PCS outcomes the effect of having a more severe injury rather than mild were both direct and indirect (direct effect - 0.50, indirect effect - 0.38).ConclusionsIndividuals with severe injuries (those with LOS ≥ 7 days and ISS 12+) had poorer recovery 12 months after the injury. In addition, post-injury mediators have an important role in influencing long-term health outcomes.Trial registrationAustralia New Zealand Clinical trial registry identification number - ACTRN12613000889752 .