Risk of injury from drinking: the difference which study design makes

Area of Science:

• Epidemiology
• Public Health
• Alcohol-Related Research

Background:

• Emergency department (ED) studies show varied injury risk estimates associated with alcohol consumption.
• The influence of study design on these risk variations remains unclear.

Purpose of the Study:

• To investigate how different study designs impact the estimated risk of injury related to alcohol consumption.
• To determine the role of control data in refining risk estimates.

Main Methods:

• Patient interviews collected data on recent and usual alcohol consumption.
• Risk estimates were calculated using case-control and case-crossover analyses (pair-matched and usual frequency).
• Control-crossover analysis was employed to adjust case-crossover estimates.

Main Results:

• Case-control analysis yielded an odds ratio (OR) of 2.7 (1.9-3.8), lower than adjusted case-crossover estimates.
• Adjusted case-crossover analysis, particularly at higher consumption levels (≥7 drinks), showed substantially increased risk (OR=7.1; 2.2-22.9).
• Usual frequency case-crossover analysis (OR=10.7; 8.0-14.3) was also higher than case-control, but adjusted control-crossover analysis (OR=2.2; 1.5-3.3) provided a different perspective.

Conclusions:

• Case-control analysis for alcohol-related injury risk may be subject to bias.
• Control data are essential for adjusting case-crossover estimates, improving accuracy.
• The importance of control data adjustments is most pronounced at higher levels of alcohol consumption preceding an injury.