Challenges associated with drink driving measurement: Combining police and self-reported data to estimate an accurate prevalence in Brazil

Background Drink driving is an important risk factor for road traffic crashes, injuries and deaths. After June 2008, all drivers in Brazil were subject to a "Zero Tolerance Law" with a set breath alcohol concentration of 0.1 mg/L of air. However, a loophole in this law enabled drivers to refuse breath or blood alcohol testing as it may self-incriminate. The reported prevalence of drink driving is therefore likely a gross underestimate in many cities. Objective To compare the prevalence of drink driving gathered from police reports to the prevalence gathered from self-reported questionnaires administered at police sobriety roadblocks in two Brazilian capital cities, and to estimate a more accurate prevalence of drink driving utilizing three correction techniques based upon information from those questionnaires. Methods In August 2011 and January-February 2012, researchers from the Centre for Drug and Alcohol Research at the Universidade Federal do Rio Grande do Sul administered a roadside interview on drink driving practices to 805 voluntary participants in the Brazilian capital cities of Palmas and Teresina. Three techniques which include measures such as the number of persons reporting alcohol consumption in the last six hours but who had refused breath testing were used to estimate the prevalence of drink driving. Results The prevalence of persons testing positive for alcohol on their breath was 8.8% and 5.0% in Palmas and Teresina respectively. Utilizing a correction technique we calculated that a more accurate prevalence in these sites may be as high as 28.2% and 28.7%. In both cities, about 60% of drivers who self-reported having drank within six hours of being stopped by the police either refused to perform breathalyser testing; fled the sobriety roadblock; or were not offered the test, compared to about 30% of drivers that said they had not been drinking. Discussion Despite the reduction of the legal limit for drink driving stipulated by the "Zero Tolerance Law," loopholes in the legislation permit many drivers under the influence of alcohol to act with impunity. In this context the police/traffic officers are often powerless to enforce the law and thus drink driving continues to go unchecked. Conclusion Strong legislation and effective enforcement are necessary to reduce the prevalence of this dangerous behaviour. Correction techniques allow calculation of a truer prevalence of drink driving, which can assist police and policymakers alike to redirect resources and align strategies.