Note from the National Guideline Clearinghouse (NGC): A staged review of the literature was prepared by Oregon Evidence-based Practice Center (EPC) for use by the U.S. Preventive Services Task Force (USPSTF). Following this, AHRQ staff conducted a separate review of recent literature on the accuracy, reliability, and clinical utility of instruments designed to screen for drug use among adolescents, adults and pregnant women (see the "Availability of Companion Documents" field).
Staged Systematic Review
To update this topic, EPC staff utilized an analytic framework (see Figure 1 in the Evidence Synthesis [see the "Availability of Companion Documents" field]) with eight Key Questions (KQs) (see "Major Outcomes Considered" field).
For this report, EPC staff used a staged review approach that focused first on the evidence for the following five critical key questions (KQ 1, 4, 5, 5a, and 7) oriented toward the health benefits of treatment and on an overarching question determining whether there is direct evidence of benefit from screening patients for treatment.
In the logic of the staged review, if the evidence for these critical key questions is insufficient to establish the links between drug misuse identification through screening, treatment, and clinically-meaningful health benefits, further systematic review to include the other key questions in the analytic framework is unwarranted. Insufficiency of evidence for these critical key questions indicates that the overall body of evidence is insufficient for a USPSTF recommendation for drug misuse screening as a clinical preventive service in primary care. Indication of sufficient evidence for critical key questions 4, 5, 5a, and 7 indicates that a full systematic review of all key questions would be warranted.
Literature Search and Strategy
This staged review is intended to update the previous USPSTF report on drug misuse, which was based on an authoritative, but non-systematic, research review. Consequently, EPC staff conducted literature searches to systematically locate relevant literature for their critical key questions as follows (see Appendix A–Search Strategies in the original guideline document).
For KQ 1, Ovid MEDLINE for the time period 1994-April 2006 was searched. Randomized controlled trials (RCTs), controlled clinical trials, and longitudinal cohort studies were included. No relevant articles were identified for this key question.
For KQs 4, 5, and 5a, EPC staff conducted a two-stage literature search to locate high-quality, relevant systematic reviews, supplemented by bridge searches as necessary. They also retrieved all potentially relevant treatment research or trials cited in the previous 1996 USPSTF report. Relevant systematic reviews were identified from four distinct searches of Ovid MEDLINE, the Cochrane Database of Systematic Reviews (CDSR), the Database of Abstracts of Reviews of Effectiveness (DARE), and PsycINFO for the time frame 1994-January 2006. They identified 14 high-quality systematic reviews that addressed treatment for one or more of the illicit drugs addressed in this report (heroin, cocaine, marijuana, multiple drugs). Those systematic reviews were used as sources of relevant trials for this review, supplemented by two additional searches for randomized or controlled clinical trials in Ovid MEDLINE and PsycINFO from 2001-April 2006. Additional articles were obtained from comparing reference lists of related reviews, studies, editorials, reports, websites, and by consulting experts. Seventeen relevant articles were identified for these key questions.
For KQ 7, Ovid MEDLINE was searched for the time period 1994-April 2006. RCTs, controlled clinical trials, and longitudinal, cohort studies were included. All potentially relevant articles cited in the 1996 USPSTF report were also retrieved. Eleven relevant articles were identified for this key question.
All studies were managed in an electronic database (Reference Manager®).
Inclusion and Exclusion Criteria
Two investigators reviewed identified abstracts for potential relevance to all critical key questions and determined eligibility by applying inclusion and exclusion criteria specific to each critical key question (see Appendix B–Inclusion and Exclusion Criteria in the Evidence Synthesis [see "Availability of Companion Documents" field]). Full-text articles for included abstracts, articles from the previous USPSTF report, and articles located from existing systematic reviews were examined for relevance. Eligible studies provided data relevant to the critical key questions for marijuana, cocaine, opiates, or multiple substances, and were English-language, primary care feasible or referable (defined in Appendix B in the Evidence Synthesis [see "Availability of Companion Documents" field), conducted in the U.S. (or applicable country), and examined adolescents/teens ages 12-17, young adults ages 18-25, adults ages 26+, or pregnant women. Studies of detoxification/withdrawal, comparative treatment effectiveness, and animal studies were not included.
For KQ 1, RCTs, controlled clinical trials, and longitudinal cohort studies were included. For KQs 4, 5, and 5a, RCTs and controlled clinical trials were included. For KQ7, RCTs, controlled clinical trials, and longitudinal cohort studies were included.
Supplemental Review: Assessment of Screening Instruments
Literature Search
AHRQ undertook a systematic review of documents identified as of August 2006, from a number of databases. The aim was to identify appropriate, validated screening instruments for the detection of drug misuse among asymptomatic patients seen in ambulatory general medical settings.
Inclusion/Exclusion Criteria
AHRQ staff first searched the Substance Abuse Screening and Assessment Instruments database (http://lib.adai.washington.edu/instruments/) maintained by the University of Washington's Alcohol and Drug Abuse Institute. Information on each questionnaire in the database was examined and questionnaires were eliminated from further consideration using the following exclusion criteria:
- Instrument is designed to detect misuse of alcohol only, or of a single drug.
- Instrument is designed primarily for diagnostic purposes or for assessment of those already known to have a substance abuse problem.
- Instrument is not available to the public (not yet published, or subject to a fee for reproduction or downloading)
- Instrument requires specific training to administer or to score/interpret results.
- Instrument contains more than 20 items or takes more than 5 minutes to administer and score.
Using the title or acronym of each remaining questionnaire (i.e., those not excluded using the above criteria), AHRQ staff searched Ovid Medline and PsychINFO, for the period from 1980 through August 2006, for published evidence in English of the instrument's validity, reliability, and clinical utility. Abstracts of identified articles were screened and rejected if they met the following exclusion criteria:
- Not a study of the specified screening instrument
- Editorial, letter, or other opinion piece
- Study conducted using only a non-English version of the instrument
- Study that examined use of the instrument for a purpose other than screening
Results
After exclusion criteria were applied to all instruments described in the SASAI database, nine instruments were identified as potentially useful for screening for drug misuse in primary care practice settings:
- Alcohol, Smoking and Substance Involvement Screening Test (ASSIST)
- Cut down, Annoyed, Guilty, Eye-opener - Adapted to Include Drugs (CAGE-AID)
- Car, Relax, Alone, Forget, Friends, Trouble (CRAFFT)
- Drug Abuse Screening Test (DAST)
- Drug Use Disorders Identification Test (DUDIT)
- Relax, Alone, Forget, Friends, Trouble (RAFFT)
- Reduce, Annoyed, Guilty, Start (RAGS)
- Rapid Drug Problems Screen (RDPS)
- Simple Screening Instrument for Substance Abuse (SSI-SA)