320 likes | 451 Views
A review of the effectiveness of addiction TCs. Prof. Dr. Wouter Vanderplasschen Ghent University, Belgium Wouter.Vanderplasschen@UGent.be. Overview. Introduction and background Findings from available reviews Aims of the study Methods Study findings Conclusion. 1. Background.
E N D
A review of the effectiveness of addictionTCs Prof. Dr. Wouter Vanderplasschen Ghent University, Belgium Wouter.Vanderplasschen@UGent.be
Overview • Introduction and background • Findings from available reviews • Aims of the study • Methods • Study findings • Conclusion
1. Background • Studycommissionedby the EMCDDA (European Monitoring Centre for Drugs and Drug Addiction) • Research period: October 2011 – August 2012 • Launchpublication: January 2014
2. Available reviews • At least 4 independent reviews of TCs in English literature • Lees, Manning & Rawlings (2004) • Smith, Gates & Foxcroft (2008) • De Leon (2010) • Malivert, Fatseas, Denis, Langlois & Auriacombe, 2012) • Divergent conclusions: • ≠ scope, objectives, selection criteria, analyticmethods • Few studies retained in all 4 reviews
29 controlled studies on TCs (8 RCTs) • DemocraticTCsforpersonality disorders, as well as addictionTCs • Strong positive effect of TCs comparedwith control interventions • Substantialstudyheterogeneity • Addiction TC outcomessignificantly more effectivethanoutcomes of democraticTCs (! More severelydisturbedpopulation)
7 RCTs of drug-free TCs, comparedwithvarying control conditions (day TC, community residence, short TC program, …) • Focus on substanceuseandretention • Few evidencethatTCs offer significant benefits comparedwithother types of residentialTx or other types of TCs • Poorevidenceduetolack of studies + itsmethodologicallimitations (high attritionrates, drop-out fromTx)
Critical evaluation of the assertion that TC effectiveness is not proven, based on RCT research • Non-exhausitve review of North American literature on addicitonTCs • Needtodistinguishbetweenvarious sources of evidence: • Field effectiveness studies : longitudinal, naturalistic design (single program or multi-program studies) • Controlled/comparison studies: comparison of twoconditions (TC vs TC or non-TC program), based on controlledgroupallocation • Meta-analyses: statistical analysis of data fromoriginal studies, resulting in an effect size • Cost-benefit analyses: Studies on costrelativetoTxoutcomes
Consistent evidence of TC effectiveness • numerous field effectiveness studies • controlled studies: betteroutcomes whencomparedto non TC-control condition • meta-analyses: 4 found small to moderate effect sizes, 2 found insufficientevidence • Cost-benefit analyses: in favor of TC treatment, in particularreducedcosts associatedwithcriminalityandgains in employment. • Most TCsroutinelyuseevidence-basedinterventionslike MI, CBT, … • Substantialevidencefor TC effectiveness: • Strong relationbetweenTxcompletionandoutcomes, betweendosageandsuccess • Consistent positivefindingsacross studies • TemporalassociationbetweenTx exposure andimprovements
Systematic review of 12 follow-up studies of TC effectivenessduringandafterTx • Studies on prison TCsexcluded • Txcompletion: 9-56%, program cessation most oftenafter 15-30 days • Decrease in substanceuseduring follow-up, still 21-100% used or relapsed • 20-33% re-enteredTx • Large differencesbetween studies in Txduration + length of follow-up period • Txcompletionandretentionidentified as robustpredictors of abstinence
3. Aims of the study • Despite 50 years of TC research, findings regarding its effectiveness are not unequivocal • Evidence merely associated with RCTs (cf. Cochrane collaboration) • However, (quasi)experimental designs have many limitations when evaluating complex and integrated multi-interventions • Small proportion of the effectiveness literature • Still, controlled studies offer the ‘best’ design to compare the effectiveness of one intervention with another, without denying evidence from other sources • Comprehensive review of the evidence resulting from (randomized and non-randomized) controlled studies. • Quasi-experimental designs included, as sound alternative for more restrictive and less naturalistic RCTs (role of self-selection and self-matching) • Available reviews limited to RCTs (Smith et al., 2006), ‘voluntary’ Tx (Malivert et al., 2012), studies published before 2000 (Lees et al., 2004)
4.Methods (eligibility criteria) • Eligibility criteria • Intervention: Drug-free TCs for the treatment of drug addiction • Target population: Adults addicted to illegal drugs • At least one of the following outcome measures was reported: • substance use (illicit drug use, alcohol use, …) • length of stay in Tx(retention, treatment completion/drop-out) • employment status • criminal involvement • health and well-being • family relations • quality of life • treatment status • mortality • …
4. Methods (eligibility criteria) • Type of studies Controlled studies • Randomized controlled trials and quasi-experimental studies comparing (prospectively) residents that followed TC treatment with • a control group that was treated in a usual care setting (‘treatment as usual’/standard of care) or another type of TC (e.g. shorter program/day TC) • or with a control group out of treatment (e.g. in prison/waitlist controls). • Studies needed to report findings on TC outcomes in adequate format and separately from other types of interventions (e.g. aftercare). • No language or country restriction was applied for selecting this type of studies.
4. Methods (search strategy) • Databases search (up to December 31st, 2011 ): • ISI Web of Knowledge (WoS) • PubMed • DrugScope • No language or publication year restrictions. • Key words : “therapeutic communit*” AND “drug* or addict* or dependen* or substance use” AND “outcome* or evaluation or follow-up or effectiveness” • Reference lists of retrieved studies and available reviews • The International Journal of Therapeutic Communities
5. Studyfindings • 30 publicationswithlongitudinal scope, based on prospectivecontrolledstudy design • Based on 16 original studies • 5 RCTs (truerandomisation) • Majority of studies performed in 1990s + allfrom US • At least 5 additional studies comparedwith Smith et al. 2006 (all in correctionalsettings)
5.1. Substanceuseandlegaloutcomes • Varying follow-up period (mostly 6-12 months, exc. >36 months) • Betweengroupdifferencesdiminished over time • ‘Substanceuse’ and ‘legalinvolvement’ most frequentlyassessed • 10/14 studies: ++ substanceuseoutcomes • 9/13 studies: ++ legaloutcomes • Multiple outcome indicators used: • seldom ≥ 2 significant outcomes in onecategory (cf. Prendergast, 2003) • Improvement in onecategorynotnecessarilyassociatedwithimprovement on otherdomains
Substanceuse • Relapse ratesbetween 25 and 55% after 12-18 months • 77 vs. 94% 3 yearsafter prison TC • Lower relapse ratesassociatedwithlonger time in Tx + participation in subsequentTx or aftercare • Relapse associatedwithseverity of dependence at Tx entry • Longer time to relapse among TC-participants • Substantialheterogeneity
Legal involvement • 1-year re-arrest andrecidivismratesaround 40-50% • Re-arrest: 63% after 3 years, 80% after 5 years • Reincarcerationratebetween 30-55% after 12-18 months (exc. Sacks et al., 2004, 2011: 9-19%) • Re-incarceration: 70% after 5 years • Time toreincarcertionhigher in TC-group • Txcompletionand Time in Txpredicted absence of recidivsm • Txcompletionassociatedwitholderage, being on paroleand single (instead of poly) drug dependence • Predictors of drug-free and no re-arrest status: • Participation in aftercare • Post-Txemployment • Olderage • Importance of Txcompletion (includingaftercare program) • Aftercarecompletersscoredbetterthanaftercare drop-outs, whoscoredbetterthan TC completersand TC drop-outs (Prendergast 2004; Wexler, 1999)
Otheroutcomes • 10/16 studies found sign. ≠ on at least 2 domains • Retentionandcompletionrates in TCslowercomparedtoother types of Tx • Lowercompletionrates in longer programs • 5/6 studies: betteremploymentrates • 5/7 studies: improvedpsychologicaloutcomes
Long-term outcomes • 6 studies looked at outcomesbeyond 12-18 months • 5/6: improvedlegaloutcomes • 3/6: lower drug use levels, but onestudy found higher alcohol useamong TC participants (Bale, 1980) • Comparisoncondition • 11/16: TC vs. usualcare • 5/16: TC vs. othertype of TC • Few differences, but bettersubstanceuseoutcomes at first FU (3/5), betteremploymentrates (2/5) andlesspsychologicalproblems (1/5) • ≠ only significant when most andleast intensive conditionwerecompared
Community vs. prison TCs • 7/16 studies in community TCs, 9 in prison TCs • Prison TC: different context (compulsorycustodyandconditionaltermsand privileges) • Community TCs: • Bettersubstanceuseoutcomes (5/6 studies) • Superior legaloutcomes (3/4 studies) • Prison TCs: • Substanceuseoutcomesonlymarginallybetter (4/7 studies) • Legal outocmesbetter in 6 studies (out of 9), but effectsmaintainedafter 2 (n=3) and 5 years (n=2)
6. Studylimitations • Only peer-reviewed studies • Non-English articles • Selection of reportedoutcomes • Studyheterogeneity: • Program characteristics, sampling methods, outcomemeasures • TC modifications: length, special needsgroups • Few replications in similarconditions • Program fidelity?! • Varying control conditions • Differingpopulations • Few ‘official’ or ‘objective data’
Most studies found significant differences, but only on one or twodomains • Low number of RCTs (n=5) + oftentruerandomization was compromised, or simplynotpossible (e.g. prison settings) nor advisable (e.g. self-selectionandself-matching hypothesis) • Poormethodologicalquality: • Small + convenience samples • High attritionrates
7. Conclusion • Someevidencefor TC effectiveness, at least in US • Lowersubstanceuseandrecidivismrates in > half of the selected studies • Positivefindingsacrosssettingsandregardless type of controls • 4 studies significantlydifferentialoutcomes in at least 3 domains • Needfor recovery-orientedTx • Gain more active control over theirlives (‘agency’) • A way of living a satisfying, hopefulandcontributing life, even with the limitationcausedbyillness (Slade et al., 2008) • Importance of subjectivequality of life + individuals’ strengthsand support systems • Stable recovery: socialparticipatonandhavingmeaningfulactivitiesandpurposes • Relapse is part andparcel of the recovery process; shouldbeseen as a learning moment • Importance of Txfidelity • Standards and goals • Training, education, role of recovered drug users
Conclusion • TC treatment: • Who benefits most from (what type of) Tc Tx at what point in the recovery process? • Notitshighereffectiveness, but ratherindividuals’ assetsand community resources andtheir personal needsand goals • Challenges: • Involve the outside community • More flexible + individualiszedapproach • Look at the costs of lengthyTx in relationtoitsbenefits • Bridge the gap between the TC andoutside society • Structural links between TC andotherTxmodalities + withaftercareandongoingsupport • RCTsandcontrolled designs are neededto prove the impact of TC treatment • Look formeasurementsthatretain as manyresidents as possible in the analyses tominimizeattrition • Use of in-treatment andprocessmeasures (dynamic variables) • Useof time to event outcomes, ratherthanbinominaloutcome variables • Needforconfirmationin a meta-analysis
Acknowledgements • Thanksto: TeodoraGroshkova (EMCDDA), Rowdy Yates (EFTC), George De Leon, Stijn Vandevelde, allkeyinformantsandmanyothers • Moren information: • Vanderplasschen, W., Colpaert, K., Autrique, M., Rapp, R.C., Pearce, S., Broekaert, E. & Vandevelde, S. (2013). Therapeutic communities for addictions: a review of their effectiveness from a recovery-oriented perspective. The Scientific World Journal, 2013, Article ID 427817. • Tobeinformedabout EMCDDA-publicationlaunch: • Wouter.Vanderplasschen@UGent.be • www.orthopedagogiek.ugent.be