Randomized Controlled Trials in Evidence-Based Mental Health Care: Getting the Right Answer to the Right Question

Abstract

The purpose of clinical research is to answer this question: Would a new treatment, when added to the existing range of treatment options available in practice, help patients? Randomized controlled trials (RCTs)--in particular, double-blind RCTs--have important methodological advantages over observational studies for addressing this question. These advantages, however, come at a price. RCTs compare treatments using a particular allocation rule for assigning patients to treatments (random assignment) that does not mimic real-world practice. "Favorable" results from an RCT indicating that a new treatment is superior to existing treatments are neither necessary nor sufficient for establishing a "yes" answer to the question posed above. Modeled on an experimental design, RCTs are expensive in time and money and must compare simple differences in treatments. Findings have a high internal validity but may not address the needs of the field, particularly where treatment is complex and rapidly evolving. Design of clinical research needs to take account of the way treatments are allocated in actual practice and include flexible designs to answer important questions most effectively.

Occupational therapy for people with psychotic conditions in community settings: a pilot randomized controlled trial

“…The pragmatic trial investigated the effectiveness of the experimental intervention in normal practice settings. A heterogeneous sample in terms of diagnosis was chosen in order to replicate clinical practice and increase the external validity of the findings (20). Similarly, participants were identified according to the diagnosis on their medical records rather than through a structured clinical screening process.…”

Women, Co-occurring Disorders, and Violence Study: Evaluation design and study population

“…Cook and Campbell (1979) argued that by studying the effectiveness of interventions in their natural contexts, quasi-experimental designs can enhance the generalizability of the research findings. Dissemination of evidence-based practices that cannot be replicated outside a rigorous research environment may be less useful in the long run than identifying programs that look promising in realworld contexts vis-à-vis treatment as usual (Essock, Drake, Frank, & McGuire, 2003).…”

Turning Mental Health Nursing Problems into Quality or Research Projects

“…The appropriate steps to conduct a RCT would have to be followed (Essock, Drake, Frank, & McGuire, 2003;Mulder, Frampton, Joyce, & Porter, 2003;Turpin, 2005).…”