Table 5.
Evidence table for brief intervention (BI) versus written information in participants screened for at-risk substance use
| Outcome a | Follow up | Event rates BI versus info | Effect estimate (95% CI) | Studies (people) | Quality of evidence | Comments |
|---|---|---|---|---|---|---|
|
PRIMARY OUTCOMES | ||||||
|
Substance use | ||||||
|
Abstinence - All substances |
3 mo |
Range 14 to 18% versus 9 to 13% |
RR 1.12 (0.41 to 3.09) |
2 (223) |
Very low |
Two studies not statistically significant. |
| RR 2.08c | ||||||
| Cannabis (Self-reportb, past 30 d, [34]). |
See Comments |
|||||
| Sedatives/hypnotics/opioidsd (NR, period not provided, [37]) | ||||||
|
Abstinence - Cocaine/heroin |
6 mo |
17% versus 13%f |
Adj RR 1.41 (0.98 to 1.95)gh |
1 (778) |
Low |
|
| Objectivee, past 30d [17] | ||||||
|
Abstinence - All substances |
12 mo |
Range 25 to 45% versus 20 to 22% |
RR 2.05 (1.13 to 3.70) |
2 (228) |
Very low |
Mixed results between studies. |
| Adj RR 1.30cgi | ||||||
| Cannabis (Self-reportb, past 30d, [34]) | ||||||
| See Comments | ||||||
| Sedatives/hypnotics/opioidsd (NR, assessment period NR, [37]) | ||||||
|
High on cannabis |
3 mo |
36/42 (86%) versus 46/55 (84%)f |
Adj RR 1.05 (0.82 to 1.15)gj |
1 (102) |
Low |
|
| Self-reportb, past 30d [34] | ||||||
| 12 mo |
25/47 (53%) versus 41/55 (75%)f |
Adj RR 0.72 (0.45 to 0.97)gj |
1 (102) |
Very low |
|
|
|
Reducing use >25% - Sedatives/hypnotics/opioids
d
|
3 mo |
29/56 (52%) versus 21/70 (30%) |
RR 1.73c |
1 (126) |
Very low |
Results favor BI over control. |
| NR, period of assessment not provided [37] |
12 mo |
28/56 (50%) versus 34/70 (49%)f |
Adj RR 0.96cgi |
1 (126) |
Very low |
Results NS |
|
Frequency of use | ||||||
|
Change in cannabis consumption. Mean change from baselinek, self-reportb, past 30 d [34] |
3 mo |
-5 versus -0.8 d (fewer d at 3 mo) |
MD -4.2 (-8.1 to -0.3) |
1 (95) |
Low |
(−) value for MD indicates fewer d consumption with BI |
| 12 mo |
-7.1 versus -1.8 d (fewer d at 12 mo) |
MD -5.3 (-0.6 to 10) |
1 (102) |
Low |
||
|
Quantity of use | ||||||
|
Defined daily dosage - Sedatives/hypnotics/opioids
d
|
3 mo |
0.42 versus 0.12 (dosage higher at 3 mo) |
MD 0.30c |
1 (126) |
Very low |
Results NS |
| Mean change from baselinek, NR | ||||||
| Patient’s dose of a given prescription drug per day (in mg) divided by the product-specific WHO measure [37] |
12 mo |
Not provided |
See Comment |
1 (126) |
Very low |
Authors state no significant difference between groups, P = 0.330 |
|
Change in drug level - Cocaine |
6 mo |
-180 versus -21 ng/ 10 mg (less at 6 mo) |
See Comment |
1 (376) |
Low |
Authors state adjusted P = 0.058, likely representing multiple adjusted analysesm |
| Change in mean from baselinel, objectivee[17] | ||||||
|
Change in drug level - Opioids |
6 mo |
-7.6 versus -7.8 ng/ 10 mg (less at 6 mo) |
See Comment |
1 (189) |
Low |
Authors state adjusted P = 0.186, likely representing multiple adjusted analysesm |
| Change in mean from baselinel, objectivee[17] | ||||||
|
Use-related harms or negative consequences of use | ||||||
|
Carried a weapon (gun, knife, club) |
3 mo |
5/42 (12%) versus 17/55 (31%)f |
Adj RR 0.44 (0.15 to 1.09)gn |
1 (97) |
Very Low |
|
| Self-report, past 30 d [34] | ||||||
| 12 mo |
5/47 (11%) versus 11/55 (20%)f |
Adj RR 0.62 (0.20 to 1.60)gn |
1 (102) |
Very Low |
|
|
|
Drove a car after using cannabis. Self-report, past 30 d [34] |
3 mo |
6/42 (14%) versus 10/55 (18%)f |
Adj RR 0.85 (0.28 to 2.08)gn |
1 (97) |
Very Low |
|
| 12 mo |
8/47 (17%) versus 13/55 (24%)f |
Adj RR 0.67 (0.26 to 1.48)gn |
1 (102) |
Very Low |
|
|
|
Rode in a car with a person drunk/high after cannabis use. Self-report, past 30 d [34] |
3 mo |
11/42 (26%) versus 13/55 (24%)f |
Adj RR 1.01 (0.46 to 1.88)gn |
1 (97) |
Very Low |
|
| 12 mo |
10/47 (21%) versus 13/55 (24%) |
Adj RR 0.85 (0.37 to 1.67)gn |
1 (102) |
Very Low |
|
|
|
Physical fight. Self-report, past 30 d [34] |
3 mo |
9/42 (21%) versus 14/55 (25%)f |
Adj RR 0.91 (0.39 to 1.76)gn |
1 (97) |
Very Low |
|
| 12 mo |
6/47 (13%) versus 19/55 (35%)f |
Adj RR 0.35 (0.12 to 0.87)gn |
1 (102) |
Low |
|
|
|
Positive behavior change | ||||||
|
Tried to cut back on cannabis use. Self-report, past 3 and 12 mo [34] |
3 mo |
29/42 (69%) versus 28/55 (51%)f |
Adj RR 1.36 (0.96 to 1.64)gn |
1 (97) |
Low |
|
| 12 mo |
34/47 (72%) versus 33/55 (60%)f |
Adj RR 0.96 (0.91 to 1.45)gn |
1 (102) |
Low |
|
|
|
Tried to stop using cannabis. Self-report, past 3 and 12 mo [34] |
3 mo |
23/42 (55%) versus 19/55 (35%)f |
Adj RR 1.58 (1.01 to 2.12)gn |
1 (97) |
Low |
|
| 12 mo |
25/47 (53%) versus 21/55 (38%)f |
Adj RR 1.42 (0.92 to 1.90)gn |
1 (102) |
Low |
|
|
|
‘Tried to be careful about situations you got into when using marijuana’ |
3 mo |
32/42 (76%) versus 38/55 (69%)f |
Adj RR 1.13 (0.84 to 1.30)gn |
1 (97) |
Low |
|
| Self-report, past 3 and 12 mo [34] |
12 mo |
34/47 (72%) versus 38/55 (69%)f |
Adj RR 1.05 (0.76 to 1.24)gn |
1 (102) |
Low |
|
|
Decision to attend treatment | ||||||
|
Abstinence obtained by substance use treatment, including detox [[17]] |
6 mo |
n/a |
Not estimable |
1 (118) |
n/a |
Data poorly reported, not provided by group. |
|
SECONDARY OUTCOMES | ||||||
|
Use of different substances | ||||||
|
Change in type of drug from baseline to follow-up - Cocaine/opioids |
6 mo |
n/a |
Not estimable |
1 (118) |
n/a |
Poorly reported by authors. |
| Change from baseline [17] | ||||||
|
Intention to reduce use | ||||||
|
Not reported in any studies |
n/a |
n/a |
Not estimable |
0 (0) |
n/a |
|
|
Other health measures | ||||||
|
Felt unsafe |
3 mo |
14/42 (33%) versus 25/55 (45%)f |
Adj RR 0.67 (0.33 to 1.16)gn |
1 (97) |
Low |
|
| Self-report, past 30 d [34] | ||||||
| 12 mo |
11/47 (23%) versus 29/55 (53%)f |
Adj RR 0.35 (0.16 to 0.72)gn |
1 (102) |
Low |
|
|
|
Change in ASI composite score from baseline - Cocaine and/or heroin |
3 mo |
Not reported |
Not estimable; See Comment |
1 (854) |
Low |
Authors state not statistically significant |
| Change from baseline |
Drug - 6 mo |
49% versus 46% reduction from baseline |
Not estimable; See Comment |
1 (562) |
Low |
Authors state P = 0.06 |
| Drug and medical subscales [17] | ||||||
| Med - 6 mo |
56% versus 50% reduction from baseline |
Not estimable; See Comment |
1 (562) |
Low |
Authors state P = 0.055 |
|
|
Other adverse outcomes | ||||||
| Not reported in any studies | n/a | Not estimable | 0 (0) | n/a | ||
aFor change from baseline data, means for baseline and follow-up timepoints are shown in Table S5, where possible.
bSelf-report using Timeline Followback Calendar.
cConfidence interval not presented due to unit of analysis error.
dA small proportion (1.5%) of participants were assessed for caffeine use in this study.
eObjective measure by biochemical hair analysis.
fUnadjusted event rates.
gAdjusted RR calculated from authors’ adjusted OR.
hAdjusted for health insurance and homelessness.
iAdjusted for prescription drug dependence according to the Structured Clinical Interview for DSM-IV Axis I disorders.
jUnclear what variables were adjusted for.
kMean change analysis first calculates the change between follow-up and baseline values for each participant and then computes the mean across those data.
lChange in mean analysis calculated the reported mean at follow-up minus the mean at baseline.
mUnclear but likely adjusts for gender, race, age, EuroQol scores, previous psychiatric history, randomization status, education level, drug route, drug problem severity (Drug Abuse Severity Test score at baseline, polydrug use, injection drug use, baseline Addiction Severity Index drug score, number of previous treatment episodes) and readiness to change.
nUnclear what variables adjusted for. adj, adjusted; ASI, addiction severity index; BI, brief intervention; CI, confidence interval; d, days; info, information; med, medical; MD, mean difference; mo, months; NS, not significant; RR, risk ratio.