Generalizability of clinical trials for alcohol dependence to community samples

Abstract

There is a growing concern that results of tightly controlled clinical trials of individuals with alcohol use disorders may not generalize to broader community samples. To assess the proportion of community-dwelling adults with alcohol dependence who would have been eligible for a typical alcohol dependence treatment study, we developed a new, simple method: we applied a standard set of eligibility criteria commonly used in alcohol outcome studies to a large (n = 43,093) representative US adult sample interviewed face-to-face, the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). We found that approximately one-half (50.5%) of all individuals with a DSM-IV diagnosis of alcohol dependence (n = 1484) and 79.4% of those who sought treatment (n = 185) were excluded by one or more study criteria. Individual study criteria excluded from 0.9% to 48.2% of the overall sample and 0.8% to 43.7% of the treatment-seeking sample. For the overall sample, the lack of motivation/compliance and financial situation criteria excluded the largest percentage of individuals. In the treatment-seeking subsample, comorbid medical conditions and legal problems excluded the largest proportions of individuals. Our study provides a new method to assess the generalizability of clinical trials, and gives further evidence that typical clinical trials for alcohol dependence likely exclude most adults with the disorder in the community and under care, and support the notion that clinical trials recruit "pure" rather than "typical" patients. Clinical trials should carefully evaluate the effects of the selected eligibility criteria on the generalizability of their results.

Alcohol Consumption, Alcohol Dependence and Attributable Burden of Disease in Europe: Potential Gains From Effective Interventions for Alcohol Dependence

“…168 The samples in the randomized trials are different from community samples with AD; 169 but it is hard to predict what this means in terms of outcome. At this point, we can only indicate this potential bias in our estimations.…”

Representativeness of Patients Enrolled in Influential Clinical Trials: A Comparison of Substance Dependence With Other Medical Disorders

“…For alcohol dependence, which has been the most extensively studied, 79% of treatment-seeking individuals would be ineligible under at least one prevalent clinical trial criterion (Blanco et al, 2008). Other work has shown that African Americans are particularly likely to be underrepresented in alcohol-dependence trials (Humphreys, 2003;Humphreys and Weisner, 2000;Humphreys et al, 2007) and that, beginning in the 1970s, treatment outcome studies have become increasingly restrictive in their enrollment procedures (Humphreys et al, 2005).…”

Sildenafil for Erectile Dysfunction in the Middle East: Observational Analysis of Patients with Diabetes and/or Hypertension Treated in the Clinical Practice Setting

“…For example, in one large study, 75.8% of patients with major depression in the community would have been excluded from a typical RCT. 25 Treatment with sildenafil was well tolerated in the clinical practice population studied in the present observational analysis. No serious treatment-related cardiovascular events were reported.…”