background Results from randomised controlled trials (rcts) are considered to provide the most reliable information on treatment efficacy for the treatment of major depression . However, rcts probably include patients that differ in many respects from ‘real life’patients . Hence, generalisability of the results might be jeopardized, since most ‘real life’ patients do not meet the selection criteria.
aim We explored the use of eligibility criteria in antidepressant and psychotherapy trials for major depression. Furthermore, we explored the socio-demographic and socio-economic features of trial participants. Subsequently, we compared participants in rcts to real life patients. We investigated the influence of the eligibility criteria, and the socio-demographic and socio-economic features on treatment outcome in daily practice.
method To obtain data on common eligibility criteria in trials we used the model of Zimmerman and colleagues (2002). For the eligibility criteria in psychotherapy trials as well as the sociodemographic/socio-economic features of rct participants, we performed an extensive literature review. Data on ‘real life patients’ were derived from chart review and routine outcome monitoring.
results Daily practice patients are frequently (> = 75%) ineligible for rcts due to not meeting the criterion of a minimum baseline severity, suicidality, and comorbid Axis I and II disorders. Exclusion of patients with mild depression resulted in higher proportion of remitters (or 2.0, ci 1.3-3.1). Overall the influence of eligibility on treatment outcome was small. Furthermore, rct-participants were remarkably more often employed than real life patients (68%versus 34%). Having a paid job contributes positively to treatment outcome (or 1.9, ci: 1.2-2.8).
conclusion Results from rcts might not be generalisable to the outpatient population who often suffer from milder depression, and are frequently unemployed. However, these features do not explain all differences in treatment outcome between trial setting and routine clinical practice, since the influence of the found differences on treatment outcome is relatively small.