Need for Informed Consent in Substance Use Studies—Harm of Bias?

Abstract

The aim of this study was to examine the differences between those who gave informed consent to a study on substance use and those who did not, and to analyze whether differences changed with varying nonconsent rates. Cross-sectional questionnaire data on demographics, alcohol, smoking, and cannabis use were obtained for 6,099 French- and 5,720 German-speaking 20-year-old Swiss men. Enrollment took place over 11 months for the Cohort Study on Substance Use Risk Factors (C-SURF). Consenters and nonconsenters were asked to complete a short questionnaire. Data for nearly the entire population were available because 94% responded. Weekly differences in consent rates were analyzed. Regressions examined the associations of substance use with consent giving and consent rates and the interaction between the two. Nonconsenters had higher substance use patterns, although they were more often alcohol abstainers; differences were small and not always significant and did not decrease as consent rates increased. Substance use currently is a minor sensitive topic among young men, resulting in small differences between nonconsenters and consenters. As consent rates increase, additional individuals are similar to those observed at lower consent rates. Estimates of analytical studies looking at associations of substance use with other variables will not differ at reasonable consent rates of 50%-80%. Descriptive prevalence studies may be biased, but only at very low rates of consent.