Effectiveness of brief interventions as part of the Screening, Brief Intervention and Referral to Treatment (SBIRT) model for reducing the nonmedical use of psychoactive substances: a systematic review

Background The purpose of this systematic review is to assess the effectiveness of brief interventions (BIs) as part of the Screening, Brief Intervention, and Referral to Treatment (SBIRT) model for reducing the nonmedical use of psychoactive substances. Methods Bibliographic databases (including MEDLINE, Embase, The Cochrane Library, CINAHL, and PsycINFO to April 2012) and gray literature sources were searched. We included randomized controlled trials that opportunistically screened adolescents or adults and then provided a one-to-one, verbal BI to those at risk of substance-use harm. Of interest was the nonmedical use of psychoactive substances (for example, drugs prohibited by international law), excluding alcohol, nicotine, and caffeine. Interventions comprised four or fewer sessions and were compared with no/delayed intervention or provision of information only. Studies were assessed for bias using the Cochrane risk of bias tool. Results were synthesized narratively. Evidence was interpreted according to the GRADE framework. Results We identified 8,836 records. Of these, five studies met our inclusion criteria. Two studies compared BI with no BI, and three studies compared BI with information only. Studies varied in characteristics such as substances targeted, screening procedures, and BI administered. Outcomes were mostly reported by a single study, leading to limited or uncertain confidence in effect estimates. Conclusions Insufficient evidence exists as to whether BIs, as part of SBIRT, are effective or ineffective for reducing the use of, or harms associated with nonmedical use of, psychoactive substances when these interventions are administered to nontreatment-seeking, screen-detected populations. Updating this review with emerging evidence will be important. Trial registration CRD42012002414


PRIMARY OUTCOMES
Substance use n/a Not estimable 0 (0) n/a Not measured.

Frequency of use [1]
• Use of Cannabis / Other drugs 1

• Days abstinent
Past 30-day use using a modified timeline followback interview.

Follow-up: 1 and 3 months
Very serious limitations (high risk of bias) No serious limitations (1 study only)

No serious limitations Serious limitations
One study assessed outcomes (lack of scientific replication). Less than 400 participants in analysis.

No serious limitations
This study small and NS, not likely to be other studies ASSIST score (composite outcome 2 ) -All substances [2] Sum score. Higher score interpreted as higher substance involvement (range 0 to 27+ points).

Serious limitations (medium risk of bias)
No serious limitations ( statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias.

No serious limitations
Although a few number of studies located, mixed results in statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias.

Reduction of defined daily dosage -Any prescription drug
Patient's dose of a given prescription drug per day (in mg) divided by the product-specific WHO measure [4].

and 12 mo
Very serious limitations (high)

No serious limitations
(1 study only)

Serious limitations
Total population size <400. One study assessed outcome (lack of scientific replication).

No serious limitations
Although a few number of studies located, mixed results in statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias.

Change in drug level -Cocaine -Opioids
Biochemical hair analysis [5] 6 mo Serious limitations (medium)

No serious limitations
(1 study only)

No serious limitations Serious limitations
Total population size <400. One study assessed outcome (lack of scientific replication).
Information poorly reportedp value only provided. We do not know width of CI and therefore are uncertain how precise the data are.

No serious limitations
Although a few number of studies located, mixed results in statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias. Total events <300 and sample size <2000. One study assessed outcome (lack of scientific replication).

Use-related harms or negative consequences of use
3 mo: CI includes no effect and appreciable benefit.
12 mo: CI includes no effect and both appreciable benefit and harm.

No serious limitations
Although a few number of studies located, mixed results in statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias.

Drove a car after using cannabis.
Past 30 days [3].

and 12 mo
Serious limitations (medium)

No serious limitations
(1 study only)

No serious limitations Very serious limitations
Total events <300 and sample size <2000. CI includes no effect and both appreciable benefit and harm. One study assessed outcome (lack of scientific replication).

No serious limitations
Although a few number of studies located, mixed results in statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias. 3 mo: CI includes no effect and both appreciable benefit and harm.

Rode in a car
exists among them. We feel there is a low risk of publication bias.

Positive behaviour change
Tried to cut back on cannabis use. [3] 3 mo Serious limitations (medium)

No serious limitations
(1 study only)

No serious limitations Serious limitations
Total events <300 and sample size <2000. CI includes no effect and appreciable benefit.

No serious limitations
Although a few number of studies located, mixed results in statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias.

mo Serious limitations (medium)
No serious limitations (1 study only)

No serious limitations Serious limitations
Total events <300 and sample size <2000. CI includes no effect and appreciable benefit. One study assessed outcome (lack of scientific replication).

No serious limitations
Although a few number of studies located, mixed results in statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias.
3 mo: CI includes no effect and appreciable benefit.
statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias.

Other health measures
Felt unsafe in the past 30 days. [ Total events <300 and sample size <2000. One study assessed outcome (lack of scientific replication).
3 mo: CI includes no effect and appreciable benefit.

No serious limitations
Although a few number of studies located, mixed results in statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias.

Change in ASI composite score from baseline -Cocaine and/or heroin
Drug and medical subscales [5] 3 mo Serious limitations (medium)

No serious limitations
(1 study only)

No serious limitations Serious limitations
Authors only state not significant, no data are provided to calculate an effect estimate and confidence interval. Accordingly, we do not know how precise the results are.
One study assessed outcome (lack of scientific replication).

No serious limitations
Although a few number of studies located, mixed results in statistical significance for an important outcome (abstinence) exists among them. We feel there is a low risk of publication bias.