CoChat
Sign Up
Cited by 110Substance Abuse Treatment and OutcomesDigital Mental Health InterventionsCannabis and Cannabinoid Research

Short- and long-term effects of digital prevention and treatment interventions for cannabis use reduction: A systematic review and meta-analysis

Open access
Boumparis, Nikolaos, Loheide-Niesmann, Lisa, Blankers, Matthijs, Ebert, David D., Korf, Dirk, Schaub, Michael P., Spijkerman, Renske, Tait, Robert J. +1 more·2019·Drug and Alcohol Dependence
DOI 10.1016/j.drugalcdep.2019.03.016
Open this paper in CoChat — free accountOpen PDF

Abstract

BACKGROUND: Frequent Cannabis use has been linked to a variety of negative mental, physical, and social consequences. We assessed the effects of digital prevention and treatment interventions on Cannabis use reduction in comparison with control conditions. METHODS: Systematic review with two separate meta-analyses. Thirty randomized controlled trials met the inclusion criteria for the review, and 21 were included in the meta-analyses. Primary outcome was self-reported Cannabis use at post-treatment and follow-up. Hedges's g was calculated for all comparisons with non-active control. Risk of bias was examined with the Cochrane risk-of-bias tool. RESULTS: The systematic review included 10 prevention interventions targeting 8138 participants (aged 12 to 20) and 20 treatment interventions targeting 5195 Cannabis users (aged 16 to 40). The meta-analyses showed significantly reduced Cannabis use at post-treatment in the prevention interventions (6 studies, N = 2564, g = 0.33; 95% CI 0.13 to 0.54, p = 0.001) and in the treatment interventions (17 comparisons, N = 3813, g = 0.12; 95% CI 0.02 to 0.22, p = 0.02) as compared with controls. The effects of prevention interventions were maintained at follow-ups of up to 12 months (5 comparisons, N = 2445, g = 0.22; 95% CI 0.12 to 0.33, p < 0.001) but were no longer statistically significant for treatment interventions. CONCLUSIONS: Digital prevention and treatment interventions showed small, significant reduction effects on Cannabis use in diverse target populations at post-treatment compared to controls. For prevention interventions, the post-treatment effects were maintained at follow-up up to 12 months later.

AI snapshot

Written from the paper’s full text.

TL;DR

This systematic review and two separate meta-analyses found that digital prevention and treatment interventions produced small, significant reductions in self-reported cannabis use at post-treatment compared with non-active controls. Prevention effects were maintained at follow-up up to 12 months, whereas treatment follow-up effects were no longer statistically significant.

Key findings

  • Prevention interventions: 6 studies, N = 2564, showed significantly reduced cannabis use at post-treatment (g = 0.33; 95% CI 0.13 to 0.54, p = 0.001); effects remained significant at follow-up up to 12 months (5 comparisons, N = 2445, g = 0.22; 95% CI 0.12 to 0.33, p < 0.001).
  • Treatment interventions: 17 comparisons, N = 3813, showed a small significant post-treatment effect (g = 0.12; 95% CI 0.02 to 0.22, p = 0.02); follow-up effects up to 12 months did not reach significance (9 studies, p = 0.10).
  • The systematic review included 10 prevention interventions targeting 8138 participants aged 12 to 20 and 20 treatment interventions targeting 5195 cannabis users aged 16 to 40.
  • Among treatment interventions, multi-session interventions such as CBT plus MI produced higher effect sizes (6 comparisons, g = 0.18) than one-session interventions such as PNF, MI, and BI (13 comparisons, g = 0.09), though subgroup differences were not statistically significant.

Methods

  • Systematic review with two separate meta-analyses; literature searched up to October 10, 2018, in PubMed, EMBASE, PsycINFO, and the Cochrane Central Register of Controlled Trials, yielding 419 abstracts (307 after duplicates), 70 full-text papers, 30 studies meeting review criteria, and 21 included in meta-analyses.
  • Included RCTs of internet- or computer-based prevention and treatment interventions; meta-analyses excluded active control comparisons and included only self-report cannabis use outcomes, assessed at post-treatment and furthest follow-up (3 to 12 months).
  • Random-effects meta-analyses calculated Hedges’s g, with I² heterogeneity, funnel plot/Egger’s test/trim-and-fill for publication bias, subgroup analyses, and univariable meta-regressions; risk of bias assessed with the Cochrane risk-of-bias tool.

Limitations

  • The paper states that many studies satisfied only a small number of Cochrane risk-of-bias criteria, so findings derive largely from low-quality RCTs; most included studies were at high risk of performance and detection bias, and only two RCTs fully blinded participants and personnel.
  • The paper states that limited public availability of RCT protocols or pre-registrations prevented evaluation of possible protocol violations, and that outcome measures varied in time frames and self-report questionnaires, reflecting the lack of an established gold standard for reporting cannabis use frequency or quantity.
  • The paper notes that subgroup analyses in conventional meta-analyses have limited power for moderator analyses; the text also makes evident that pooled effects are restricted to non-active control comparisons, and heterogeneity was high for prevention interventions and moderate for treatment interventions.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1016/j.drugalcdep.2019.03.016

Cites

92 of 92 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 11 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    Subgroup analyses did not show significant differences between groups for prevention interventions.

    “Subgroup analyses did not show significant differences between groups (see Table 4).”
    Quote in the paperNot a substantial claim
  • Findingown data

    Heterogeneity was moderate for the treatment post-treatment meta-analysis.

    “Heterogeneity was moderate (see Fig. 3 and Table 4).”
    Quote in the paperNot a substantial claim
  • Findingown data

    Heterogeneity was high for the prevention post-treatment meta-analysis.

    “Heterogeneity was high (see Fig. 2 and Table 4).”
    Quote in the paperNot a substantial claim
  • Findingown data

    Prevention interventions compared with active controls (4 studies, n = 6647) reported no significant differences in cannabis use at follow-up.

    “The prevention interventions that were compared to active comparison conditions (4 studies, n = 6647)and which were therefore not included in the first meta-analysis (see Table 1)reported no significant differences in Cannabis use between conditions at the follow-up assessments.”
    Quote in the paperSubstantial claim
  • Findingown data

    No publication bias was indicated for the prevention meta-analysis by funnel plot, trim-and-fill, and Egger's test (p = 0.58).

    “No publication bias was indicated via assessment of the funnel plot and Duval and Tweedie's trim-and-fill procedure.Egger's test did not indicate an asymmetric funnel plot (p = 0.58).”
    Quote in the paperNot a substantial claim
  • May combine two claims. 13
  • The quote only partly states this claim. 2
  • Classified as probably not a substantial claim. 5

  • May combine two claims.

    Findingown data

    The meta-analyses included 21 studies, 6 prevention and 15 treatment interventions.

    “Thirty randomized controlled trials met the inclusion criteria for the review, and 21 were included in the meta-analyses.”
    • Also stated in: § 3.1. Selection and inclusion of studies · plain
      “A total of 30 studies met all criteria for the systematic review (10 prevention interventions and 20 treatment interventions) and 21 studies for the meta-analyses (6 prevention and 15 treatment interventions).”
    Quote in the paperSubstantial claim

  • May combine two claims. The quote only partly states this claim.

    Findingown data

    The systematic review included 30 randomized controlled trials, of which 10 were prevention interventions and 20 were treatment interventions.

    “A total of 30 studies met all criteria for the systematic review (10 prevention interventions and 20 treatment interventions) and 21 studies for the meta-analyses (6 prevention and 15 treatment interventions).”
    Quote in the paperSubstantial claim

  • The quote only partly states this claim.

    Findingown data

    The treatment interventions in the systematic review targeted 5195 cannabis users aged 16 to 40.

    “The treatment interventions we identified for the systematic review (n = 20) recruited a total of 5195 participants (n = 3007 in the experimental conditions; n = 2188 in the comparison conditions; Table 2).The mean ages of the participants ranged from 16 to 40”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    In the prevention meta-analysis, the included studies satisfied few Cochrane risk-of-bias criteria: three studies fulfilled one criterion, two fulfilled two criteria, and one fulfilled three criteria.

    “In summary, three studies fulfilled one criterion of the possible seven, two studies fulfilled two criteria, and one study fulfilled three criteria.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    In the treatment meta-analysis, the included studies varied in Cochrane risk-of-bias criteria: two fulfilled one, two fulfilled two, three fulfilled three, four fulfilled four, two fulfilled five, and one fulfilled seven criteria.

    “In summary, two studies fulfilled one of the seven possible criteria, two studies fulfilled two criteria, three studies fulfilled three criteria, four studies fulfilled four criteria, two studies fulfilled five criteria, and one study fulfilled seven criteria.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For prevention interventions, meta-regression analyses did not reveal significant associations between effect sizes and intervention duration, number of sessions, or risk of bias.

    “The univariable meta-regression analyses did not reveal significant associations between the effect sizes of the digital interventions and (a) intervention duration in days (slope 0.003; 95% CI -0.004 to 0.01, p = 0.421), (b) number of sessions in the intervention (slope 0.015; 95% CI -0.03 to 0.06, p = 0.548), or (c) risk of bias of the studies (slope -0.03; 95% CI -0.34 to 0.28, p = 0.868).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    Risk of bias was independently evaluated by two authors using the Cochrane risk-of-bias tool, with consensus reached by discussion in cases of disagreement.

    “Risk of bias was independently evaluated by two of the authors (N.B. and L.L.); in the event of disagreement, consensus was reached by discussion.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation

    Prior meta-analyses concluded that digital interventions appeared to effectively reduce cannabis use in adolescents and adults at post-treatment.

    “The study by Tait and colleagues (Tait et al., 2013) concluded that digital interventions appeared to effectively reduce Cannabis use in adolescents and adults at post-treatment, albeit with a limited number of available RCTs”
    • Also stated in: § 1. Introduction · plain
      “The meta-analysis by Olmos and colleagues (Olmos et al., 2018) came to a similar conclusion after identifying 9 RCTs”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Tait et al., 2013

      Not checked in the free run.

    • cites Olmos et al., 2018

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    Digital prevention and treatment interventions are effective in reducing cannabis use at post-treatment, with prevention interventions producing a larger pooled effect size than treatment interventions.

    “We found that digital prevention and treatment interventions are effective in reducing Cannabis use at post-treatment, with prevention interventions producing a larger pooled effect size for Cannabis use reduction than treatment interventions.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The self-reported cannabis use outcomes from included studies, despite varying time frames and questionnaires, can be validly pooled in meta-analysis.
    • Rests on an assumption: The pooled effect sizes from the separate prevention and treatment meta-analyses are comparable enough to conclude that prevention interventions produce a larger effect than treatment interventions.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    Outliers were explored by assessing whether study effect sizes and 95% confidence intervals overlapped with the pooled 95% CI, and an additional analysis without outliers was conducted when outliers were observed.

    “The presence of outliers was explored by assessing whether effect sizes and 95% confidence intervals of the studies overlapped with the 95% CI of the pooled effect size; in cases where outliers were observed, we conducted an additional analysis without the outliers.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisereasoning

    Previous meta-analyses did not distinguish between prevention and treatment interventions, which might have complicated interpretation of pooled effect estimates of cannabis use reduction.

    “These previous meta-analyses did not distinguish in their inclusion criteria between prevention and treatment interventions, whichgiven the substantial differences, most notably in terms of target groupsmight have complicated the interpretation of the provided pooled effect estimates of Cannabis use reduction.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    The search and screening were performed by two authors independently, with consensus reached by discussion in cases of disagreement.

    “The search and screening were performed by two of the authors (N.B. and L.L.) independently; in cases of disagreement, consensus was reached by discussion.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    Digital interventions can expand accessibility of evidence-based health services by removing barriers such as distance and stigmatization, making prevention and treatment available independent of time and place.

    “Specifically, by removing barriers such as distance and stigmatization, such interventions can make prevention and treatment available independent of time and place, thus lowering the threshold to access care (Taylor and Luce, 2003)”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Taylor and Luce, 2003

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    Publication bias was assessed by funnel plot inspection, Egger's linear regression test, and Duval and Tweedie's trim-and-fill procedure.

    “We assessed publication bias by visually inspecting the funnel plot.To assess whether the bias captured by the funnel plot was significant, we used Egger's linear regression test of the intercept (Egger et al., 1997).Duval and Tweedie's trim-and-fill procedure (Duval and Tweedie, 2000) was used to account for potentially missing studies.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Egger et al., 1997 source

      Not checked in the free run.

    • cites Duval and Tweedie, 2000 source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data

    The post-treatment effects were sustained for up to 12 months for prevention interventions but not for treatment interventions.

    “Follow-up assessments indicated that the post-treatment effects were sustained for up to 12 months for the prevention interventions but not for the treatment interventions.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The self-reported cannabis use outcomes from included studies, despite varying time frames and questionnaires, can be validly pooled in meta-analysis.
15 of 15

  • The paper relies on this without support.

    The self-reported cannabis use outcomes from included studies, despite varying time frames and questionnaires, can be validly pooled in meta-analysis.

    “The applied intervention outcomes varied as to the time frames measured (with self-reported frequencies ranging from Cannabis use in the past week to the past six months) and the self-report questionnaires employed.”
    “The variation in outcome measures reflects the lack of an established gold standard for reporting Cannabis use frequency or quantity”

  • The paper relies on this without support.

    The pooled effect sizes from the separate prevention and treatment meta-analyses are comparable enough to conclude that prevention interventions produce a larger effect than treatment interventions.

    “We found that digital prevention and treatment interventions are effective in reducing Cannabis use at post-treatment, with prevention interventions producing a larger pooled effect size for Cannabis use reduction than treatment interventions.”

  • The paper relies on this without support.

    The effect sizes from digital interventions can be validly compared with effect sizes from face-to-face interventions reported in other meta-analyses.

    “digital prevention and treatment interventions still produce smaller effect sizes than face-to-face prevention (d = 0.58) and treatment interventions (g = 0.49) (Davis et al., 2015;Porath-Waller et al., 2010).”

  • The paper relies on this without support.

    The effects of digital interventions observed in the included studies generalize to a variety of target users and different settings.

    “we suggest that digital interventions have the potential to decrease Cannabis use in a variety of target users and different settings.”

  • The paper relies on this without support.

    Reducing Cannabis use leads to improved health outcomes.

    “They may have a small effect, but when implemented on a large scale at low cost they could produce substantial health gains.”
5 of 5

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • A meta-analysis of digital interventions for cannabis use disorder found that cannabis use was significantly reduced following both prevention and treatment interventions as compared with controls.

    “A meta- analysis of digital inter ventions for cannabis use disorder found that cannabis use was significantly reduced following both prevention and treatment interventions as compared with controls.”

    From Substance use disorders: a comprehensive update of classification, epidemiology, neurobiology, clinical aspects, treatment and prevention (2023) · cites this paper as [257]

    The Discussion (§4; page unavailable) reports that both digital prevention and treatment interventions reduced cannabis use at post-treatment, supporting the main finding. But the paper studies cannabis use reduction, not exclusively cannabis use disorder: the Introduction (§1) says prevention targets young people regardless of use status, and §3.5 says only one treatment study recruited participants based on a cannabis abuse or dependence diagnosis. (LLM, conf=0.88)

    Experimental — accuracy not yet measured

  • While effects of prevention interventions for cannabis use disorder remained significant at follow-ups up to 12 months, effects of treatment interventions did not.

    “However, while the ef fects of prevention interventions remained significant at follow- ups of up to 12 months, effects of treatment interventions did not 257 .”

    From Substance use disorders: a comprehensive update of classification, epidemiology, neurobiology, clinical aspects, treatment and prevention (2023) · cites this paper as [257]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Prevention-group counts exceed the stated total

    Minor

    The reported experimental and comparison groups total 8,418 participants, not 8,138.

    Reported 8138 · recomputed 8418

  • Prevention comparison categories do not total eleven

    Minor

    The stated six non-active and four active comparisons add up to ten, rather than eleven.

    Reported 11 · recomputed 10

  • Number of studies reporting baseline ASSIST scores differs

    Minor

    The treatment characteristics first say four studies reported mean ASSIST Cannabis scores, then say the baseline mean ASSIST score was based on five studies.

3 of 3

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. Schwinn, 2010, Preventing drug abuse among adolescent girls: outcome data from an internet-based intervention, Prev. Sci., 11, 24, 10.1007/s11121-009-0146-9 · DOI

Cited by

Sorted by citations

Related

Authors
Boumparis, N., Loheide-Niesmann, L., Blankers, M., Ebert, D. D., Korf, D., Schaub, M. P., Spijkerman, R., Tait, R. J., & Riper, H.
Affiliations
  • Amsterdam Public Health (NL)
  • Amsterdam UMC Location University of Amsterdam (NL)
  • Friedrich-Alexander-Universität Erlangen-Nürnberg (DE)
  • University of Amsterdam (NL)
  • University of Zurich (CH)
  • Parnassia Groep (NL)
  • Curtin University (AU)
Corresponding author
Boumparis, Nikolaos
ORCID
Venue
Drug and Alcohol Dependence
Publisher
Elsevier BV
Published
2019-07
Volume
200
Pages
82-94
ISSN
0376-8716 (ISSN-L)
DOI
10.1016/j.drugalcdep.2019.03.016
OpenAlex
W2945809486
Semantic Scholar
98cf294ded57ca214e1322d0e879cc89acd2b70d
Cited by
110
References
91
Citations by year
2026: 9 · 2025: 11 · 2024: 23 · 2023: 20 · 2022: 18 · 2021: 13 · 2020: 13 · 2019: 3
FWCI
1.71
Type
journal article
Language
en
Open access
Licence
Locations
  • Drug and Alcohol Dependence (publishedVersion, cc-by-nc-nd)
  • PubMed (publishedVersion)
  • Amsterdam UMC (VU Amsterdam) - Institutional Repository (publishedVersion, other-oa)
  • VU Research Portal (publishedVersion, other-oa)
  • Zurich Open Repository and Archive (University of Zurich) (acceptedVersion, cc-by-nc-nd)
  • UvA-DARE (University of Amsterdam) (submittedVersion, cc-by-nc-nd)
  • Radboud Repository (Radboud University) (submittedVersion)
  • Data Archiving and Networked Services (DANS) (submittedVersion, other-oa)
  • eSpace (Curtin University) (submittedVersion, cc-by-nc-nd)
  • Pure Amsterdam UMC (submittedVersion, other-oa)
  • Zurich Open Repository and Archive (University of Zurich) (submittedVersion, cc-by-nc-nd)
  • Radboud Repository (Radboud University) (submittedVersion, other-oa)
  • Zurich Open Repository and Archive (University of Zurich)
Topics
  • Health Sciences › Medicine › Epidemiology › Substance Abuse Treatment and Outcomes
  • Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
  • Health Sciences › Medicine › Pharmacology › Cannabis and Cannabinoid Research
Keywords
cannabis use reduction, digital interventions, cannabis use disorder, adolescents, young adults, maintenance of treatment effects, follow-up outcomes
MeSH terms
Adolescent, Adult, Behavior Therapy, Marijuana Abuse, Child, Humans, Marijuana Smoking, Therapy, Computer-Assisted, Time Factors, Randomized Controlled Trials as Topic, Treatment Outcome, Young Adult

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review