Objective: This study examined whether data needed to inform evidence-based practice can be found in leading mental health journals.
Methods: Research studies described in articles that were published in 12 leading mental health journals in 1999 were examined to determine whether they evaluated clinical interventions, used rigorous designs, were conducted in routine practice settings, and included well-defined diagnostic groups and heterogeneous samples.
Results: Twenty-seven percent (N=295) of the 1,076 articles that were reviewed described research that evaluated interventions. Of these 295 articles, 64 percent evaluated pharmacologic interventions and 33 percent evaluated psychosocial or psychotherapeutic interventions. Of the articles that evaluated interventions, 60 percent described randomized designs, but samples were modest; 25 percent of the studies reported 31 or fewer participants. Of the 295 articles, 84 percent described studies conducted in specialty mental health settings; very few (4 percent) described studies conducted in public mental health or managed care environments, which are common practice settings. Most samples were diagnostically well defined, but evidence of treatments for diagnoses other than schizophrenia and mood disorders was limited.
Conclusions: This systematic review suggested that data needed to inform and advance evidence-based practice does not have the prominent place it deserves in leading journals. Only a quarter of the research studies that were examined evaluated clinical interventions, and articles that described pharmacologic interventions were published twice as often as articles that described psychosocial or psychotherapeutic interventions. Rigorous research designs predominated, but sample sizes were modest. Evidence was scarce on treatment effectiveness in routine practice settings.