Developing a baseline assessment battery: balancing patient time burden with essential clinical and research monitoring.

Abstract

Objective: Baseline assessment in a multisite alcohol-dependence treatment study has several purposes: addressing inclusion/exclusion criteria and characterizing participants to illuminate subsequent efficacy and safety patterns of the interventions. Combination pharmacotherapy and behavioral therapy trials, however, require more complex assessments than single-modality studies. Medication trials require measures of initial safety for study drug as well as for subsequent side-effect and adverse-effect monitoring (e.g., physical examination, laboratory markers, somatic symptoms, medical conditions and concomitant medications). Behavioral therapy trials (and in some cases medication trials) warrant baseline measurement of mediators of therapy effect (e.g., prior treatment history, motivation for change, self-efficacy, other psychiatric conditions, treatment expectations and network supports). Measures may be needed to interpret interactions that may be discovered between these modalities.

Method: The National Institute on Alcohol Abuse and Alcoholism COMBINE Study evaluated the potentially overwhelming number of candidate instruments through an iterative process, using the following sequence to finalize a rational baseline assessment battery: key constructs, representative measures, determination of duration of assessment, risk of assessment reactivity, goal priorities, necessity for measure "pruning" and order of measure presentation. After selecting the draft battery, feasibility was evaluated in a pilot study prior to the main trial.

Results: The battery was feasible to administer and avoided unintended selection bias. Dropout was substantial, however, and differences across sites in baseline assessment completion rates reflected a tendency of a central intake model to function as an initial filter of dropouts, compared with a direct recruitment model.

Conclusions: This process holds several potentially useful lessons for investigators.