Alcohol use is a major risk factor for road traffic crashes worldwide. According to the Global Burden of Disease (GBD) 2015 update, 0.33 million road traffic deaths can be attributed to alcohol use globally in 2015 [1
]. In China, about 30% of road traffic deaths, or 93,750 fatalities, were ascribed to drunk-driving in 2015 [1
As a cost-effective intervention [2
], implementation of drunk-driving law was included by the United Nations in the list of recommended interventions for member countries in The Global Plan for the Decade of Action for Road Safety 2011–2020
]. In China, the Standing Committee of the National People’s Congress amended the Road Traffic Safety Law of the People’s Republic of China in 2011 to enhance the penalties for drunk-driving [5
]. At that point, drunk driving became a criminal offence, with severe punishments issued depending on the severity of the crime [7
]. Specifically, when drivers are detected having a blood alcohol concentration (BAC) of 0.02 to 0.08 g/100 mL, they are charged with drink driving and their license will be suspended for 5 years. When a driver is detected having a BAC > 0.08 g/100 mL, drunk driving is charged. Drunk driving is deemed as a crime in China and may cause imprisonment and suspension of a driving license for 10 years. In the case of a fatal traffic crash and a drunk driving charge, driving licenses may be suspended for life [5
Timely and rigorous evaluative studies are critical to monitor the effectiveness of implementation of drunk-driving law, but few publications are available [9
]. No published studies in English or Chinese examine the effectiveness of the implementation of drunk-driving law in China. Thus, we sought to accomplish this goal by examining data from two open access data sources, the GBD 2015 update and Chinese police data, that provide estimates of road traffic morbidity and mortality attributable to alcohol use for China. Multiple data sources offer stronger evidence than a single data source if they agree with each other. In this present study, therefore, we used both data sources to assess the effectiveness of drunk-driving law implementation in China. We hypothesized there might be sharp decreases in alcohol-related traffic injuries following implementation of the more severe drunk-driving law in China in 2011.
Our study offers three major findings, all of them intriguing but somewhat challenging to interpret. First, the GBD 2015 update and police data revealed significant disparities in the number of road traffic injuries attributed to alcohol use, as well as inconsistent patterns in rates of morbidity and mortality attributable to alcohol use after the implementation of more severe national drunk-driving law in 2011. Second, the GBD 2015 update displayed inconsistent changes in age-standardized morbidity versus mortality rates before and after the implementation of the more severe national drunk-driving law in 2011, with morbidity changing in the positive direction (13.28%) between 2010 and 2015 but mortality changing in the negative direction (−13.49%). Third, police data suggest a steady decrease in both crude morbidity and mortality before the implementation of a more severe drunk-driving law (from 2006 to 2010), but a gradual increase after 2011. We discuss each of these findings below.
The inconsistent findings between GBD 2015 update and police data are probably a result of the different strategies each dataset used to estimate injuries attributable to alcohol use. The GBD 2015 update applies sophisticated mathematical modeling based on various available data sources to generate estimates [10
], while police data are collected through a standardized questionnaire after each crash related to drunk driving is suspected by a police officer. Each data source offers both advantages and disadvantages, and it seems likely that neither is perfectly valid. The GBD 2015 update provides estimates based on risk factor attribution models and available research evidence. Because there is a lack of high-quality, continuous exposure data documenting fatal and non-fatal injury data in China [15
], the death estimates attributable to alcohol use in the GBD 2015 update may be biased to some extent. The police data consist of the actual numbers of fatal and non-fatal injuries attributable to crashes from drunk driving that are directly collected by road police. Police data are documented to have serious under-reporting problems for a variety of reasons [12
The fact that morbidity and mortality data showed inconsistent changes in the GBD 2015 update may primarily be a result of data input strategies for the estimation models. In the GBD 2015 update, the major input data to estimate morbidity in China are surveillance data from the National Injury Surveillance System (NISS). Because the NISS was established only in 2006 [16
], the GBD 2015 update employs a combination of advanced models and fragmented published data to estimate numbers and rates of non-fatal injury in China. In addition, NISS is a hospital-based surveillance system and its surveillance data are highly sensitive to health policy changes that lead to substantial changes in people seeking hospital-based medical care. For example, as the governmental investment increased, the coverage of social medical health insurance increased from 85% in 2008 to 95% in 2010, and 97% in 2015 [18
]. The rise of social medical insurance coverage led to a jump of patients who visited public hospitals and resulted in an impact on injury surveillance data [21
The result that police data indicate increasing drunk driving incidents following implementation of the more severe penalties but decreasing rates beforehand is bewildering. It may be caused by a combination of at least four factors: (a) rapid motorization and an increase in licensed motor vehicle drivers that occurred across China during that era, leading to more vehicles on the road and greater opportunity for alcohol-related road traffic crashes; (b) public education and enforcement that accompanied implementation of the more severe drunk-driving law, perhaps increasing recognition and recording of incidents; (c) other prevention efforts, including improvements of the road environment and introduction of new safety devices in motor vehicles, and/or (d) changes in data reporting that occurred, partly because of increased public education about and attention toward alcohol-related road traffic crashes nationwide during that time period. Data on enforcement seem particularly intriguing as an explanation. After implementing the severe drunk driving law, the traffic police division within the Ministry of Public Security of China greatly increased enforcement of drunk driving law [23
]. For example, the number of roadside breathing tests for drunk driving increased from 0.46 million to 0.78 million and the number of drivers whose licenses were temporarily suspended due to drunk driving grew from 0.33 million to 0.50 million from 2011–2015 [11
We conclude that the currently available data are insufficient to judge the effectiveness of implementing the more severe drunk driving law in China. Both the GBD 2015 update and police data suffer from some flaws; neither provides strong evidence of our hypothesized effect. The results highlight the importance and urgency of establishing a high-quality free and online data query system to support policy evaluation in China. China could replicate the successful establishment of the Fatality Analysis Reporting System (FARS) in the United States, which was created in 1975 and is now widely used to enact and evaluate road traffic laws including child restraint laws, state safety belt laws and Graduated Driving Licensing (GDL) laws [24
]. The data and results do, however, indicate the possibility of behavior change occurring as a result of the law. Enforcement may have increased, and publicity and government effort to detect dangerous driving are documented to have increased [11
]. Data from other nations offer clear evidence that reducing drunk-driving rates requires a multi-faceted effort, including safety education, institution of more severe penalties for risky driving under the influence of alcohol, and strong enforcement of law.