NDARC Technical Report no. 317 (2011)
Aims: Alcohol misuse causes substantial public health harm. Strategies have been proposed to reduce alcohol-related harm at the community-level, which requires suitable community-level measures to monitor changes over time and between communities. For alcohol-related crime, certain offences occurring at certain times that often involve alcohol have been used as a proxy measure. There is currently no adequate empirical rationale for identifying the most reliable proxy measure of alcohol-related crime. This report examines the suitability of three measures of alcohol-related crime.
Methods: Police records of reported incidents from twenty communities in NSW, Australia, that were involved in a community-wide randomised controlled trial to reduce alcohol-related harm were examined. Three measures were derived; i) serious assaults only, ii) a broader range of assaults and iii) assaults and public nuisance offences. Hierarchical linear models (HLM) account for various sources of variability and correlation of longitudinal data and were used to determine reliability estimates for model parameters and in the calculation of the intraclass correlations (ICC).
Results: The broadest measure of alcohol-related crime (assaults and public nuisance offences) was found to have the highest reliability estimates between communities at a given time point and over time. This measure also had the highest ICC, indicating relatively more variability in the measure can be attributed to differences between towns rather than changes over time.
Conclusions: The HLM approach gives more accurate reliability estimates than could be assessed using a repeated measures ANOVA. For the communities from where these data derive, the broadest measure is the most reliable for comparing rates of alcohol-related crime between them, and for assessing intervention effects over time.
Citation: Breen, C., Shakeshaft, A., Slade, T., D'Este, C. and Mattick, R. (2011) Alcohol-related crime: Finding a suitable measure for community-level analyses using routinely collected data, Sydney: National Drug and Alcohol Research Centre.