Spatial-temporal analysis of drink-driving patterns in Hong Kong

Area of Science:

• Road safety
• Traffic behavior analysis
• Alcohol-related incidents

Background:

• Bars and restaurants are primary drinking venues, raising concerns about increased drink-driving and alcohol-related crashes in these areas.
• Effective drink-driving enforcement strategies require understanding the spatial and temporal patterns of impaired driving.

Purpose of the Study:

• To model the spatial-temporal distribution of drink-driving in Hong Kong.
• To identify factors influencing drink-driving prevalence across different geographical and temporal clusters.

Main Methods:

• Cluster analysis was employed to identify spatial-temporal patterns of drink-driving.
• Zero-inflated regression models were used to analyze contributing factors for each identified cluster.

Main Results:

• Six distinct spatial-temporal clusters of drink-driving were identified, including bar/restaurant areas, urban, and rural zones, across different timespans.
• Drivers in rural areas exhibited higher alcohol consumption compared to urban areas.
• Seasonal variations and vehicle class significantly impacted breath alcohol concentration (BrAC) levels.

Conclusions:

• Drink-driving patterns vary significantly based on location (urban vs. rural) and time (weekend-overnight vs. other timespans).
• Targeted enforcement strategies should consider geographical, temporal, and seasonal factors, as well as vehicle types, to effectively reduce drink-driving incidents.