Malivert M., Fatséas M., Denis C. et al.
European Addiction Research: 2012, 18, p. 1–11
Unable to obtain a copy by clicking title? Try asking the author for a reprint by adapting this prepared e-mail or by writing to Dr Auriacombe at firstname.lastname@example.org. You could also try this alternative source.
Methodological shortcomings in the original studies prevented this review from reaching a firm conclusion on the lasting benefits of residential therapeutic communities, though it was clear that while residents stayed, substance use was significantly reduced.
Summary Therapeutic communities are a type of drug-free residential service. Their objectives include maintaining abstinence and socially rehabilitating drug users. Unlike some other residential facilities with similar aims, in these communities the therapeutic system is based on treatment stages which afford residents increasing degrees of personal and social responsibility for the running of the service. Peer influence mediated through group processes is used to help individuals assimilate social norms and develop social skills, and social rehabilitation is promoted by daily activities. Residents have the opportunity to progress in the hierarchy to themselves managing group activities as a peer leader.
A previous systematic review found little evidence that therapeutic communities significantly improved on outcomes from other types of residential treatment, or that one type of therapeutic community was more effective than another. However, this review included services in prison or offered as an alternative to prison, possibly not comparable to those voluntary entered outside prison. It was also limited to studies which randomly allocated participants to therapeutic communities versus alternative or no treatments, and did not seek to establish which types of substance users might be best suited to therapeutic communities.
Using a similar methodology, the featured review focused on therapeutic communities for adult substance users outside the criminal justice system. Aims were to assess their effectiveness in terms of completion and retention in treatment as well as substance use, and to determine if any characteristics were associated with greater success in achieving abstinence. Studies had to report relevant outcomes and be available in English or French up to the end of January 2011, but not necessarily to have randomly allocated participants to therapeutic communities versus an alternative.
Twelve such studies were found investigating outcomes from 61 therapeutic communities and 3271 participants. All but two studies were conducted in the Americas, and in all but two cocaine was the dominant problem drug. Differences between the studies were such that it was not appropriate to combine their findings in a meta-analysis. A study which uses recognised procedures to combine quantitative results from several studies of the same or similar interventions to arrive at composite outcome scores. Usually undertaken to allow the intervention's effectiveness to be assessed with greater confidence than on the basis of the studies taken individually.
Reported in all the studies, average retention in the communities ranged from 38 to 180 days, representing 30% of the expected programme duration. Six studies also assessed the proportion of residents who completed their programmes, ranging from 9% to 56%; 27–70% stayed at least half the expected time.
Follow-up periods during which outcomes were assessed varied from six months after entering the communities to six years after discharge. In all the studies substance use decreased during the programme or after discharge. Nevertheless, during follow-up periods, 21–100% of subjects had used substances or met criteria for relapse.
The post-discharge period was reported on in eight studies, most often (three studies and four communities) the six months after leaving. Over this period, in one study 34% of former residents had used substances in the previous 30 days, and cannabis use and drinking had both fallen significantly. In another, after staying in a community with a three-month programme, 48% of former residents had relapsed, compared to 41% when the programme was six months. Lastly, over a three-month window during the six-month follow-up, 33–41% of former residents of another community reported use of cocaine, 34–35% alcohol, 16–18% cannabis and 9–15% heroin.
The longest post-discharge follow-ups were four years in one Australian study, during which all the former residents had relapsed at least once at some time, and in Spain six years, during which 46% of former residents had relapsed, defined as substance use more than three times in two months, most commonly alcohol, cannabis and/or cocaine.
When residents dropped out, it was usually within the first month. Generally, residents dependent on heroin but not other substances were more likely than other substance users to complete treatment. Older residents and those in shorter programmes too were more likely to complete. Psychiatric disorders were unrelated to completion.
Having stayed longer in the programme was the variable most consistently associated with abstinence during follow-up periods. In one study too, communities with longer intended programmes were associated with better substance use outcomes at follow-up. In one US study, relapse was more common among patients who had been employed in the three years before starting treatment, had a history of drug injecting, or lived with a partner who also used substances. In another US study, cocaine dependence alone was more predictive of relapse than heroin combined with cocaine dependence.
In the three studies to have assessed this, after leaving 20–33% of former residents re-entered treatment of some kind during follow-up periods.
Depending on the length of the treatment period, this review documented positive outcomes in the form of significant decreases in substance use during therapeutic community stays, though whether this was also the case after leaving was obscured by methodological issues. Most studies found low treatment completion rates, and that residents who left usually did so soon after starting the programme.
In more detail, all studies found decreased substance use during the programme and after discharge. During follow-up periods, 21–100% of subjects had used substances or met criteria for relapse and 20–33% had started another treatment episode. Longer retention best predicted abstinence at follow-up. Completion was most likely among older residents and in shorter programmes but was unrelated to psychiatric disorders.
It was clear that substance use was depressed during stays in the communities, but whether relapse continued to be prevented in the follow-up periods was unclear because of different definitions, and an inability to determine whether relapse was to the main problem drug or to another substance. These methodological limitations may explain the great variability of relapse rate across studies, casting doubt on the degree to which the results can be relied on as indicators of the benefit to be expected from such programmes.
Compared to other types of treatment, therapeutic communities do not appear to offer significant extra benefits, but might represent a better option for patients with severe psychosocial problems, depending on whether they stay long enough in the programme.
commentary This review attempted to clarify the indecisive verdict of an earlier review by including non-randomised studies and eliminating a presumed major source of variability – whether the community was in the context of a prison or other sentencing option. Still, it too was indecisive in respect of lasting impacts, firmly concluding only that while residents stay, they use substances less often than before they entered. This in itself is a worthwhile achievement, but one considerably diluted by the review's finding that typically stays are short because residents quickly leave. This too seemed the major limitation on the effectiveness of English residential rehabilitation services in an audit of the progress of residents in 2010–11. Reporting on that audit, England's National Treatment Agency for Substance Misuse stressed that residential rehabilitation works in concert with non-residential services, typically taking its residents after they have been prepared by other services, which also continue the treatment of many residents after they leave. From the featured review it seems that internationally it is also the case that for many residential rehabilitation is not the end of a treatment and addiction career but an episode within it, making it difficult to isolate the contribution of the residential element in the treatment journey.
Compared to the earlier review which it sought in some ways to improve on, the featured review's search for studies was very limited – to, it seems, just one database, though a major one. The earlier review found little evidence that residential therapeutic communities were, in outcome terms, preferable to residential or non-residential alternatives, a result partly due to significant shortcomings in the studies, including high proportions of participants who were not or could not be included in the analyses.
None of the randomised studies in that review addressed the issue of whether the therapeutic community model is preferable to other ways of structuring residential care outside prison. Just one of the studies directly addressed another key issue – whether residential care is critical to the success of therapeutic communities. If it is not, costs can be reduced and/or more people treated. The study capitalised on the decision in 1990 of a residential service in San Francisco to introduce a parallel day programme, also based on therapeutic community principles and scheduled to run for a year followed by aftercare. The study exemplified the main weakness of randomised trials of residential versus non-residential care: such studies must select patients who can safely and practically be sent to either option and who are willing to leave the choice to chance, yet any advantages of residential care are likely to be most apparent among homeless clients, those whose vulnerability makes non-residential care unsafe, or those with strong preferences. Given this winnowing of the caseload, not surprisingly, in San Francisco residential care conferred few long-term advantages. In the first six months (when they were at least partly protected by the residential environment) residential clients were significantly less likely to relapse, but over the next year the benefits from residential care dissipated while relapse rates among day clients remained steady. A year to a year and a half after entering treatment, about half of both groups had remained abstinent and about a quarter had for a period relapsed to using at least four times a week.
Drug and Alcohol Findings has also summarised studies comparing residential and non-residential care. From these studies it seemed that residential settings help extricate residents from particularly damaging environments but the added benefits can fade after discharge back into the community. Those who particularly benefit have been people at risk of suicide and clients with relatively severe psychiatric problems, in some cases combined with severe employment or family problems. These and other studies support the general contention that more severely dependent and problematic clients differentially benefit from residential care. Where studies have found no added benefit for more severe cases this may have been because the service's caseload was limited in severity, or because the study set severity limits so that all the subjects could safely be allocated to residential or non-residential care.
Other attributes found in some studies to favour residential care include low cognitive functioning, homelessness, low social support, and poor employment prospects. What matters in any particular situation will depend on the range of problems in the caseload and the alternative treatments on offer. For example, if very severe cases are admitted beyond the capacity of any of the options, or if the caseload is unproblematic enough to do well whatever the treatment, then none will seem preferable. Similarly, where these are available, intensive non-residential options (but not routine outpatient care) may almost match residential settings even for severe cases.
For more on residential rehabilitation see this introduction and one-click search for relevant Findings analyses.
Last revised 08 August 2012. First uploaded