The measurement of harmful outcomes following drinking on licensed premises

The relationships between five potential indicators of alcohol-related harm following drinking on licensed premises in Perth were explored. These were annual purchases made by individual licensees of 'high' (>=3.8%) and 'low' (<3.8%) alcohol content drinks, the number of times a particular licensed establishment is cited by drivers as the last place of drinking prior to failing a roadside breath-test (including after accidents) and the annual number of assaults occurring either on or in the vicinity of particular licensed premises. The study area selected was a central part of the Perth metropolitan area with 367 licensed premises serving a residential population of 400 000. Highly significant correlations were found between each of the five variables. The correlations involving purchases of low alcohol drinks, however, were small. When purchases of alcohol were controlled, significant, though lower, correlations, were still evident between the other three variables. This suggests that there are risk factors other than extent of alcohol sales which further research will need to identify, and that these indicators of harm can be of value in monitoring the impact of future intervention strategies.