Outcome Assessment in Trials of Pharmacological Treatments for Alcohol Use Disorders: Fair and Strict Testing

Outcome assessment in the pharmacological treatment of alcohol use disorders (AUDs) faces specific challenges resulting from low adherence to treatment, high rates of dropout, and the susceptibility of patient self-reports to bias. This review discusses methodological issues in planning, conducting, and interpreting clinical trials on AUD treatment against the background of the principle of 'strictness and fairness' of testing. Threats to fairness include factors that limit the implementation of an intervention, such as low compliance and early treatment termination. In turn, fairness of testing is increased by factors that support the degree to which an intervention is implemented, such as the use of adequate pretreatments and the matching of psychosocial and pharmacological treatment strategies. Furthermore, selecting outcomes on the basis of an intervention's mechanism of action and including continuous outcomes as sensitive measures of drinking change further increases fairness by increasing the likelihood that the data will adequately reflect the effects of the intervention. On the other hand, strictness of testing is increased by all measures that limit the influence of confounders that could potentially lead to an overestimation of effects. The use of a side effect-mimicking placebo to prevent an unmasking of blinding and the repeated assessment of alcohol biomarkers to validate drinking self-reports might be valid strategies to further increase the strictness of testing by limiting risks of bias in trials of AUD treatment.