Send email for updates


About updates
Matrix cell
Supported by Alcohol Research UK web site Society for the Study of Addiction web site Skills Consortium web site. Opens new window


Title and link for copying Comment/query to editor Drug Matrix Alcohol Matrix

S Seminal studiesK Key studiesR ReviewsG Guidancemore Search for more studies

Links to other documents. Hover over for notes. Click to highlight passage referred to. Unfold extra text Unfold supplementary text

Alcohol Matrix cell E4: Treatment systems; Psychosocial therapies

S 0% mutual aid group attendance with usual procedures, 100% with systematic encouragement (1981). Early and tiny trial pioneered procedures to link alcohol treatment patients to mutual aid groups. None offered just information attended in the following week, but everyone systematically encouraged by procedures which included a phone call with an existing member, who accompanied the patient to their first meeting. Similar procedures have been trialled in contemporary UK and US studies. Discussion in bite’s Issues section.

K Difficult, but system change can be engineered and extend service delivery (2012). Results in US states included extending anti-relapse alcohol medication to more clinics and the implementation of psychosocial approaches which promoted continuing care by facilitating the transfer of patients between services. Siting case managers at detoxification services has been (2006) one successful transfer tactic, discussed in cell E3’s bite.

K De-individualisation forced by cost-containment associated with worse treatment outcomes (2008). Advanced US analysis of comprehensive data from a national treatment study found substance use outcomes were relatively poor after treatment at centres constrained by funders in the services they could offer and in their ability to individualise treatment, but relatively good in services subject to quality accreditation.

K Promoting SMART Recovery mutual aid (2010). What worked in disseminating a cognitive-behavioural alternative to 12-step mutual aid in England. Key theme was the tension between being supported versus being co-opted by treatment services. Much of the research relates to 12-step groups but may also apply to other mutual aid approaches. For this work see bite’s Issues sections (1 2).

K Active referral promotes NA/AA attendance in UK but abstinence not significantly improved (2012). Systematic encouragement from either a doctor or a peer (especially the latter) substantially improved attendance at 12-step groups, but impacts on abstinence were smaller and not statistically significant. Corresponding US study below. Discussion in bite’s Issues section.

K In USA active and persistent attempts to link patients to mutual aid groups helps reduce substance use (2007). Among patients treated by the US medical service for ex-military personnel (the Veterans Affairs Health Care System), showed that persistent and practical efforts can strengthen 12-step group involvement after treatment and modestly improve substance use outcomes. Corresponding British study above. Discussion in bite’s Issues section.

K Text chat therapy (2011). Dutch trial of internet-based therapy for problem drinking via text-chat conversations found this improved on an automated self-help option. Impacts of this programme were included in a simulation study (2011) which suggested health would improve and/or costs reduce if across a country on-line interventions supplemented or replaced conventional care.

R Ways to continue to (after)care for patients (2011). Psychosocial strategies to help ensure patients who need it receive long-term care or aftercare. Discussion in cell D2’s bite.

R Peer-based recovery support services (2009). Compendious review discussed in bite’s Highlighted study section. Will inform (among others) planners and commissioners of services aiming to improve linkage to mutual aid groups and other peer-based recovery support resources, interventions discussed in bite’s Issues section. Also reviews literature on peer-based support itself, for which see discussion in bite’s Issues section.

R Attending AA meetings after treatment helps sustain drinking reductions (2014). That was the implication of this innovative analysis, but does that mean treatment services improve drinking outcomes by actively promoting participation in 12-step fellowships? Our commentary tackles this issue head on and finds the evidence equivocal. Relevant discussions in bite’s Issues sections (1 2).

R Varieties and impacts of case management (2006). An expert Euro-US collaboration examines a core mechanism for transforming isolated episodes of care into coherently staged and comprehensive reintegration programmes – a ‘case manager’ who remains a stable hub orchestrating different phases of service delivery. See also US guidance below. Discussion in bite’s Issues section.

R Severe cases differentially benefit from residential care (2003). Notes from Drug and Alcohol Findings on studies comparing residential with non-residential treatment. Concludes that for patients who accept and can safely be sent to either, there is little to choose between them. Severe cases may however differentially benefit from residential care. See also a review (2006) limited to randomised trials of therapeutic communities.

G Commissioning an alcohol intervention and treatment system ([UK] National Institute for Health and Care Excellence, 2011). UK’s official health advisory body on how commissioners should organise and procure treatment services across an area which implement related national guidance and satisfy policy requirements. On psychosocial interventions NICE took its lead from associated clinical guidelines ([UK] National Institute for Health and Care Excellence, 2011). Discussion in bite’s Where should I start? and Issues (1 2) sections.

G Setting up a local treatment system ([UK] Department of Health and National Treatment Agency for Substance Misuse, 2006). Guidance for local health organisations and their partners on delivering a planned and integrated treatment system for adult problem drinkers.

G Integrated care for drug and alcohol users (Report Produced for the Scottish Advisory Committee on Drug Misuse, 2008). Treatment system guidance for Scotland.

G Organising holistic and continuing care ([US] Substance Abuse and Mental Health Services Administration, 1998). US consensus guidance on case management to orchestrate the range of services often needed to promote lasting and multi-faceted recovery. Related review above and discussion in bite’s Issues section.

more This search retrieves all relevant analyses.
For subtopics go to the subject search page and hot topics on computerised therapy and advice and on commissioning services.

Matrix Bite Open Matrix Bite guide to this cell Open Matrix Bite guide and commentary

Close Matrix Bite guide and commentary Close Matrix Bite Close Matrix Bite guide and commentary


Links to other documents. Hover over for notes. Click to highlight passage referred to. Unfold extra text Unfold supplementary text

What is this cell about? The roles of ‘psychosocial’ therapies in systems for treating problem drinking implemented across an area. In particular, their role in creating a cost-effective mix of services ranging from brief advice and counselling to extended therapies and residential care which offers patients/clients attractive access points and appropriate options for moving between services or using them in parallel. In this they are critical because in Britain treatment for alcohol dependence usually consists entirely of psychosocial support. Drugs are almost universally used to ease withdrawal in inpatient units, but in 2013/14 in England, of the 101,782 drinkers treated in non-residential community settings, just 16% were prescribed a medication.

Creating a local treatment system involves commissioning, contracting and purchasing decisions to meet local needs in the context of resource constraints and national policy. At this level, research is rarely of the ‘gold standard’ randomised controlled trial format. Whole treatment systems are not easily manipulated by researchers based on a toss of a coin, leaving us largely reliant on studies of how things work out naturally. These have to grapple with multiple influences and differences between areas which obscure the impacts of the features researchers are attempting to pin down as causal factors in outcomes. Such studies will often include but not focus on psychosocial services, making the corresponding cell on treatment systems as a whole of relevance also to this cell.


Where should I start? Anyone in Britain responsible for alcohol treatment systems is likely to start with this guidance from the National Institute for Health and Care Excellence (NICE), the UK’s official source of national guidance and advice to improve health and social care. Based on associated clinical guidelines, NICE says commissioners “should consider commissioning psychological interventions ... as a key component of their specialist alcohol service”, and should go so far as to “specify [their] structure and duration ... including the length, number and frequency of sessions”. Though for us these guidelines were an “impressive” achievement, we can use this bit of their advice to illustrate a general point about being too doggedly ‘evidence-based’ to the point of being ‘evidence-bound’.

Look at the relevant recommendations (section 6.24) in the full version of the guidelines. Should commissioners really stipulate that cognitive-behavioural and behavioural therapies “usually” occupy an hourly session a week, and social network and environment-based therapies eight 50-minute sessions, all spanning 12 weeks? Our analysis of the guidelines warned that “the expert group was vulnerable to seeing what researchers have chosen to study for research purposes as the way practitioners should do things”. Now the mystery of why 12 weeks becomes clear: that span is typically chosen by researchers. Yet as cautioned in cell A4’s bite, there is no reason to believe this is also how patients should be treated – that there is any “usual” span, frequency or session length for what should be a highly individualised endeavour. What we have is a cascade from how researchers standardise for research purposes, to clinical guidelines, to commissioners, who on this basis are advised to embed these practices in services. On the basis of this research among others, we should question any advice or service-delivery system which could have the effect of de-individualising treatment. Related issue discussed below.


Highlighted study William White’s monograph on peer-based recovery support through mutual aid groups and other resources is typically meticulous and comprehensively researched, yet cogently written. Its importance is that these low-cost or free resources seem the only feasible way to square the circle of doing more with less – reconciling diminished resources with bringing more patients to the point where they can leave treatment, achieve wide-ranging, whole-life recovery, not just remission from dependence, yet avoid life-, health-, and crime-threatening relapse. Apart from cost, peer support’s virtues include potentially life-long support available 24 hours a day, and possibly too self-empowerment not naturally promoted by the roles of ‘patient’ or ‘client’.

The monograph comes from (see his collected writings) someone who more than any other has promoted the recovery movement and provided its scholarly underpinnings, helping contextualise treatment as often merely the first step to the prolonged “recovery maintenance” advocated by the US expert chosen to advise Public Health England on addiction treatment. In this vision, the focus shifts to systems around the clinic within which the patient must eventually reshape their life in community with others who have done or are trying to do the same, secured by ties to family, community, and work. Peer-based support is an essential element within what William White has described as “recovery-oriented systems of care”.

In a more modest and selective way, below we also look at the evidence on mutual aid as an aftercare resource and as a route to recovery in its own right, asking whether its clear potential to advance and sustain recovery can profitably be drawn on by treatment services and commissioners and integrated into a local treatment/recovery system.

Issues to consider and discuss

So fundamental it needs no evidence? Here’s another recommendation from NICE’s guidance for commissioners: “Care coordination should be part of the routine care of all service users in specialist alcohol services”. Again the source for this recommendation is the associated clinical guidelines. What these mean by ‘care coordination’ is that someone should “assure all agreed elements of the care package are linked together and communicated in a clear and effective manner”. According to the guidelines, another manifestation of this ‘anchor’ function is the ubiquitous keyworker role. But for this core function, the guidelines offer no evidence at all. At the heart of the way addiction treatment is practised in the UK lines lies an evidence black hole.

Is it that the need for care coordination/keyworking is so obvious that it no more requires a randomised controlled trial than does strapping on a parachute before leaping from a plane? Before accepting this, remember that trials have overturned other seemingly self-evident truths – such as that fully-fledged treatment must be better than brief advice which then leaves ‘alcoholic’ patients to sink or swim on their own.

In contrast to the core care coordination role, the similar but more active and specialised case management role (recommended only for harder-to-hold and more complex cases) is supported by at least some research evidence in NICE’s clinical guidelines and from entries in this cell (1 2). Should the variously termed care management, care coordination or keyworking also be subject to validation through research trials, and how would such a study be done? Without these trials, can we really say care coordination should be mandatory for everyone in specialist treatment?


How much should be left to services? What this bite’s Where should I start? section hinted at, here we tackle here head on: what should commissioners specify in their tenders for bids to run services and in their agreements with service providers, and what should they leave to the discretion of managers and clinical staff in the services? That is, of course, an issue for treatment of any kind, but it is most pertinent in psychosocial therapies. Medications have to be adjusted to the individual and the situation, but regulatory authorities and pharmaceutical companies set clear expectations about what conditions and what types of patients the drugs are for, there are protocols for induction, dosing and treatment termination, and documented effects and side effects. Qualified and professionally regulated experts in the form of doctors and pharmacists interpret these rules and unqualified personnel are not allowed to prescribe prescription-only medications.

But for psychosocial approaches, beyond the generalities of the ‘common factors’ introduced in cell A2’s bite – and qualities like empathy are hard to contractually require – there are no rules, or none that should not sometimes be broken; it is not even the case that qualified ‘experts’ do better on average than ‘counsellors by experience’. If commissioners leave it all to the service, they have no direct control over the quality and nature of the services they are responsible for. Services could legally use unqualified counsellors and therapists and deliver programmes which might be recognised therapies, but also might be idiosyncratic or simply those which appeal to the manager or the therapist. The risk is that patients will be short-changed and do less well than they might, though that is hard to predict. Under-dosing a powerful medication or substituting a placebo may predictably worsen outcomes (though the placebo effect should not be underestimated), but the same cannot be said of psychosocial therapies. Nevertheless, commissioners may feel more confident in specifying accepted and research-tested interventions delivered by qualified personnel.

Leave it to the service and risk idiosyncratic interventions, or tie the service down to the tried and tested?

The alternative risk is that if commissioners specify too tightly or on inappropriate dimensions, they may counterproductively limit the service’s responsiveness to the needs and preferences of individual patients, and that too could be damaging. One way out is to specify (and pay for) only the required outcomes and leave most of how they get to those outcomes to the service – but this strategy is itself largely unevidenced, and what evidence we do have from the UK substance misuse sector is not encouraging. Where do you stand on this dilemma? Leave it to the service and risk idiosyncratic interventions from unsuitable personnel, or tie the service down to the tried and tested and risk de-individualisation and discouraging innovation?


Build your aftercare system on mutual aid? Without doubt the biggest challenge facing many authorities responsible for commissioning substance use treatment services in the UK is dramatically reduced funding for these services. At the same time, expectations of holistic recovery – not just the remission of dependent substance use – have effectively led to demands to do more with less. Usually freely available, mutual aid groups offer a potential way to square this circle. Nearly all the research has focused on groups based on the 12 steps of Alcoholics Anonymous and in particular on their impacts in the US context. Though reflecting this balance in the research, we do not intend to dismiss the contributions of other mutual aid approaches such as SMART Recovery, particularly important in the UK context where 12-step approaches are not as dominant as in the USA.

In evidencing whether the potential of these resources can be realised, much has been built on a US study which showed that if they try hard and smartly enough, services can profitably direct patients to continuing support from mutual aid groups. Conducted by the US medical service for ex-military personnel (the Veterans Affairs Health Care System), it was the only study cited by NICE in its clinical guidelines to directly test whether investing in relatively intensive encouragement and direct help to attend 12 step groups gained better outcomes. The intervention entailed counsellors linking patients to 12-step volunteers who would accompany patients to meeting, and using 12-step journals to check on meeting attendance and to re-encourage attendance if needed. Largely on the basis of these findings, NICE’s alcohol services commissioning guidance recommended such efforts “for all people seeking help for alcohol misuse”. Specifically, they were to be given “information on the value and availability of community support networks and self-help groups (for example, Alcoholics Anonymous or SMART recovery)” and helped to participate “by encouraging them to go to meetings and arranging support so that they can attend” – broadly the approach tested in the US study. Check whether NICE’s evidence stacks up by Unfold supplementary text unfolding and working through the supplementary text below.

% interviewed patients attended groups or abstinent

Not at the time available to NICE were the results of a British study (listed above) which might have tempered its enthusiasm for encouraging referral to 12-step groups. The general picture was that being encouraged by either a doctor or a peer (especially the latter) substantially improved attendance at 12-step groups, but resulting impacts on abstinence were much smaller and not statistically significant chart right. This contrast calls in to question whether post-treatment attendance ‘artificially’ elevated by special efforts during treatment has a substantial impact on later abstinence. Rather than ‘naturally occurring’ 12-step attendance fostering abstinence, the results seem more consistent with its being a marker of pre-existing ability and determination to sustain abstinence.

The study was, however, not intended as a definitive test of what intensive intervention might be able to achieve in Britain, so neither are its results a definitive verdict on such interventions. It trialled a very modest single-session intervention, and assessed its impact with a single, limited measure of substance use among a small group of patients from one unit over a short time period. Though falling short of statistical significance, the small increase in the proportions abstinent might (if further studies suggest this was not a chance finding) be considered clinically worthwhile, especially if associated with broader welfare gains.

The final strand of evidence we offer derived from an innovative synthesis of US trials of encouraging patients to attend AA meetings after treatment. Listed above, the analysis was not intended to establish whether treatment services safeguard treatment gains by actively promoting attendance, but our commentary used the same trials to tackle this issue. Relative to alternative treatments, the general pattern was that despite often substantially increasing attendance, interventions based on facilitating involvement in 12-step groups have generated no or only modest statistically significant impacts on drinking. Some studies have found abstinence modestly boosted relative to alternative approaches, but generally without any extra impacts on heavy drinking or its adverse consequences.

What then is the answer to the question we began with – whether to build your aftercare system on mutual aid? As elsewhere in this commentary, it seems desirable to have a range of aftercare options, permitting a more individualised offer to patients. If the default is mutual aid, though some patients will benefit, even extensive attempts to encourage patients to attend will reap at best modest impacts across the whole caseload. When groups are abstinence-based, such efforts may lead more patients to recover via abstinence, but sometimes only instead of non-abstinent routes. Though that seems as far as the evidence can take us, absence of strong evidence for ‘x’ does not mean, ‘Do not do x,’ if it makes sense, costs little, may help some, and is unlikely to cause harm. Referral to 12-step groups seems to fit in this bag, as long as it is not pursued so determinedly that other options which suit some patients better are swept aside, and unsuitable patients and/or groups are pressured into contact. Given what we know from everyday life and from studies of coerced AA attendance ( below), it would be extraordinary if the right kind of social support failed to help some drinkers, and extraordinary too if the same but for them ‘wrong’ kind failed to be counterproductive for others.


Support mutual aid says NICE The section above addressed the issue of whether treatment services should intensively refer patients to mutual aid groups, but these groups also gain adherents outside the context of and instead of treatment, forming a standalone recovery resource. For NICE, this too should be encouraged. NICE’s guidance recommended that commissioners consider “supporting the establishment of new groups and networks [and] encourage local providers to make rooms available within treatment centres for meetings [and] ensure that case management is available to people with complex needs, so they receive additional support and/or transport to make it easier for them to attend meetings.” These activities should foster mutual aid as a standalone resort for problem drinkers who do not opt for treatment, as well as an extra resource for those who do. If they can attract and work well at least for an appreciable proportion of dependent drinkers, the return on investment for a commissioning system would be substantial.

Establishing whether attending Alcoholics Anonymous mutual aid groups really does help was the main purpose of the synthesis of US studies referred to in the section above. It concluded that their amalgamated results showed that attending AA meetings after treatment really does help sustain drinking reductions – it was not just a case of patients who are in any event going to do well also attending the groups.

But even this sophisticated analysis did not answer the question of whether outside of a treatment context attending mutual aid groups is beneficial. To get closer to that we can turn to an earlier analysis led by the same author and described in our commentary on his later study in the paragraph starting, “A second study”. You will see that among drinkers who despite contacting gateways to treatment remained untreated, attending AA groups was estimated to help reduce the severity of drinking, and that rather than these groups attracting more promising problem drinkers, they attracted those less likely to do well on their own. It is, however, worth noting that while drinkers who attended AA were much more likely to be abstinent than those who chose to go it alone, there was no substantial or significant difference in their experience of drink-related problems. Another analysis derived from the same study found that those who initially opted for AA rather than outpatient treatment cost health services much less, but did no worse in respect of their drinking and related problems. Cost-savings were also apparent when the health-care costs of patients treated in US 12-step-based treatment programmes were compared with those treated in cognitive-behavioural programmes.

These rather convoluted attempts to disentangle the effects of attending mutual aid groups from self-selection bias are critical because the best way to achieve this separation – a randomised trial – has proved unsuitable. All three to date were deeply flawed as assessments of AA as usually accessed and attended, and in two of the trials methodological features meant they were poor indicators of relative impacts on drinking. But the results of these trials do act as a caution that people forced to attend AA do worse than when coerced instead into professionally run treatments or simply left to their own devices. Focusing on the latter, a synthesis of the findings found comparisons with no treatment were not statistically significant, but the trend was for people coerced into AA to do worse.

Though randomised trials seem inappropriate, evidence gathered through other kinds of analyses is subject to too many caveats to be convincing. Despite the analyses referred to above, the evidence remains weak and apart from abstinence as an outcome, almost entirely lacking in analyses which have focused on adjusting for self-selection bias. Rather than these weaknesses being due to weak or no impacts, they may be due to the difficulty of finding a way to convincingly demonstrate those impacts. Fostering mutual aid as recommended by NICE can be a low-resource strategy to widen access to recovery resources, so the evidence bar can be correspondingly lower than for a costly strategy. Worth prioritising, worth trying among other things, or too uncertain and best left to develop on its own – what would your assessment be?

Close Matrix Bite guide and commentary Close Matrix Bite Close Matrix Bite guide and commentary