Transporting efficacious treatments from academia to community: barriers and opportunities

Abstract

Over the past few years the chasm between substance dependence research findings and actual clinical practice has grown strikingly apparent to experts on both sides of the divide (Lamb et al., 1998; Marinelli-Casey et al., 2002). Many novel approaches demonstrated to have efficacy in rigorous research protocols languish on the shelf without ever achieving widespread clinical application. Naltrexone for opioid (Carroll et al., 2001) or alcohol dependence (Garbutt et al., 1999) stands out as an example of this problem in pharmacotherapy research, and contingency management for reduction in use of a variety of substances (Budney et al., 2000; Robles et al., 2002; Silverman et al., 1999) offers a salient example from behavioral research. The paper by Rounsaville et al. (2003), ‘Single versus Multiple Drug Focus in Substance Abuse Clinical Trials Research,’ in this issue of Drug and Alcohol Dependence highlights and makes noteworthy observations about a fundamental predicament in the design of clinical research trials which perpetuates this gap between research and practice: in deference to methodological purity, many clinical trials of previously untested interventions are structured to examine the effects of the intervention upon the use of a single substance when, in the real-world, most patients with substance dependence use multiple substances. Although many researchers have already gained awareness of and grapple with the theoretical and practical dilemmas that contribute to this disjunction, the paper by Rounsaville et al. serves to generate more public discourse about the topic and conveys a thorough, thoughtful, well organized delineation of these dilemmas. After describing some of the potential advantages of a single drug focus that explain its ascendancy in clinical trials work, the paper underscores the ripple effects, including limited feasibility and generaliziability, that ensue when single drug using research samples are unrepresentative of broader clinical populations. Without attempting to be doctrinaire, the paper provides one possible set of logical suggestions about how to resolve this conundrum and acknowledges the challenges such as large sample size requirements, flexibility in outcome measures, and alternative statistical approaches that multiple drug focus research would pose. An important background theme that emerges from a reading of the paper is that nearly every decision made regarding experimental design in substance dependence clinical trials involves a trade-off. Narrowing the focus of the study and the characteristics of the sample to preserve internal validity will entail sacrifices in external validity and vice versa. Rounsaville et al. make some distinctions between pharmacotherapy studies and behavioral intervention studies. While noting that certain pharmacotherapies, such as disulfiram and naltrexone, may have efficacy across different substances of dependence, they also state that ‘. . . Most pharmacological drug abuse treatments are hypothesized to affect a single class of abused substances while behavioral treatments tend to target processes common to many types of substance abuse.’ In theory, at least, their argument about behavioral treatments could be extended to pharmacotherapies as well. If, as current neurobiological theories of addiction presume, most substances of dependence alter brain reward mechanisms via direct or indirect increases in dopamine activity with modulation by other systems including endogenous opioids, serotonin, acetyl choline, norepinephrine, gamma amino butyric acid, and excita* Tel.: /1-206-764-2782; fax: /1-206-764-2293. E-mail address: andrew.saxon@med.va.gov (A.J. Saxon). Drug and Alcohol Dependence 70 (2003) 127 /129