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 E2: Treatment systems; Generic and cross-cutting issues

K Local area strategies in England lack vision, clarity and commitment ([UK] Department of Health, 2011). When from 2006 to 2011 government alcohol policy support teams made more than 480 visits to local strategic partnerships, they found improving but often muddled and uncoordinated attempts to improve public health through alcohol-related interventions, which lacked consistent commitment.

K What the local treatment system feels like for the patient (2015). Interviews with alcohol patients at an NHS addiction service in England revealed that for them the treatment journey was often “fragmented, with input from a number of different staff in different settings and an overall lack of clarity around the role and remit of each”.

K Adequacy of service provision in Scotland (NHS Health Scotland, 2014). Evidence that in 2012 Scotland’s alcohol treatment caseload equated to about 1 in 4 of the country’s alcohol-dependent adults – better than in England. Evidence too of a peer-based recovery orientation taking root. Discussion in bite’s Issues section.

K Systems change helped improve access to and retention in treatment (2008). US NIATx programme halved waiting times and extended retention partly by fostering a self-sustaining inter-service improvement network and a performance analysis system linked to funding. See also this later extension (2012) to the programme. Related NIATx study and web site listed below.

K Expert coaching helps services improve patient access and retention (2013). Randomised trial tested the improvement collaborative model developed by the US NIATx quality improvement resource. Arrangements for services to learn from each other were less effective and less cost effective at improving patient access and retention than assigning each clinic a quality improvement expert to ‘coach’ a service through the process. Related NIATx study listed above and web site below. Discussion in bite’s Highlighted study section.

K Disappointing early results from English payment-by-results schemes ([UK] Department of Health, 2013). Among the 3081 alcohol patients treated in the pilot schemes there was no indication of elevated abstinence rates and the proportion exiting treatment free of dependence was lower than in the rest of England and lower than in the same areas before the pilots. Discussion in bite’s Issues section.

K In Delaware paying for results led to rapid increase in drug-free treatment exits (2008). Rather than specifying treatment inputs like numbers of counselling sessions, the US state of Delaware incentivised patient recruitment, engagement, and drug- and alcohol-free treatment completions; the result was more patients, more engaging treatment, and a rapid increase in satisfactory treatment completions. But there were signs too that services focused on doing enough to earn the rewards without seeking to excel in these or in other ways. Discussion in bite’s Issues section.

K How much should treatment systems rely on residential rehabilitation? (2007). Rare randomised trial confirmed that unless there are pressing contraindications, intensive day options deliver outcomes equivalent to residential care. Often of course, there are pressing contraindications. See also this informal Findings review.

R Recovery-oriented systems of care (2008). Creating a recovery-friendly environment is the best way to sustain resolution of substance use problems argues this (as we described it) “sweeping, learned but practice-oriented tour-de-force”. Discussion in bite’s Where should I start? section.

R Research supporting components of a recovery system ([US] Substance Abuse and Mental Health Services Administration, 2009). Evidence for key elements of recovery-oriented systems of care such as continuity of care anchored in the community and delivered by integrated services on the basis of system-wide education and training. See also associated implementation case studies.

R Policy strategies for improving outcomes (2011). Two of the world's most respected addiction researchers also with top-level policymaking experience set out the options for improving treatment systems.

R Funding mechanisms for substance use treatment (Report for the Australian Department of Health, 2014). Chapter 6 comprehensively reviews funding mechanisms including payment by results, for which it finds no peer-reviewed evidence that it has improved post-treatment alcohol or drug client outcomes. Part 2 of the report makes recommendations for Australian service planning and commissioning which may in parts be applicable to the UK. Discussion in bite’s Issues section.

G Commissioning alcohol treatment systems in England (National Institute for Health and Care Excellence, 2011). UK’s official health advisory body on organising and procuring treatment services across an area which implement national guidance and satisfy policy requirements. Incorporated in broader commissioning guidance from Public Health England. Expected impacts on drinking built into a spreadsheet (NICE, accessed 2016) enabling commissioners to estimate health gains from a portfolio of interventions.

G Commissioning integrated alcohol harm-reduction systems in England (Public Health England, 2016). Key principles and associated action-prompts for developing an integrated local system to reduce alcohol-related harm, including treatment services which meet NICE standards. One of a suite of commissioning guidance and resources. Supported by a spreadsheet (Public Health England, accessed 2016) enabling commissioners to estimate social benefits and effects on performance indicators.

G Scotland’s vision of a high quality treatment system (Scottish Government and Convention of Scottish Local Authorities, 2014). What for the Scottish Government ‘quality’ consists of in substance use services. Intended to ensure commissioning of the quality of treatment and support services needed to meet the needs and aspirations of a local population. Builds on report (2011 ) on commissioning alcohol services. See also more provider-oriented English guidance.

G Integrated care for drug or alcohol users (Report Produced for the Scottish Advisory Committee on Drug Misuse, 2008). Guidance for Scotland on implementing a treatment system which aims to combine and coordinate all the services required to meet the assessed needs of the patient.

G Quality standards for alcohol and drug services ([Irish] Health Service Executive and Ana Liffey Drug Project, 2013). Update adopted by the Irish government of the QuADS standards developed for UK drug and alcohol services. Checklist of practices which for different types of services constitute quality, and which could be used by commissioners to evaluate services and promote improvements.

G US NIATx system change tools ([US] NIATx, accessed 2016). Web-based service provided by the University of Wisconsin and supported by US government, offering practical strategies for commissioners and planners to promote change across a treatment system including engaging services in mutual leaning and support, tested in this study. Specific aims include reducing waiting times and improving retention (see this study), and increasing admissions and reducing no-shows (see this study). Discussion in bite’s Highlighted study section.

more This search retrieves all relevant analyses.
For subtopics go to the subject search page and hot topic on evidence-based commissioning; see also this on-line library of papers related to recovery-oriented systems of care.

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? How across an administrative area to engineer an effective and cost-effective mix of services which offers patients/clients appropriate options for entering and moving between services or using them in parallel. Involves commissioning, contracting and purchasing decisions to meet local needs in the context of resource constraints and national policy. Activities include: needs assessment; restructuring or re-tendering services; contractual requirements on services to demonstrate evidence-based practice, meet standards, and implement performance monitoring; and financial or other rewards/sanctions linked to activity, quality or outcomes. At this distance from the preoccupation with intervention effectiveness, research is scarce, and research on whole treatment systems is rarely of the ‘gold standard’ randomised controlled trial format. Instead researchers often have to interpret how things happen in the messy real world, attempting to isolate what may have been the active ingredients among a complex set of variables not under their control. The key limitation of such methodologies is the difficulty of being sure what (if anything) was cause, and what effect.


Where should I start? William White’s monograph could form not just the start, but the middle and end of your investigation of the recovery transformation in treatment. It comes from (see his collected writings) an authority who more than any other has promoted and provided the scholarly underpinning to the new recovery eras in both the UK and the USA. For him ‘recovery’ entails a shift from isolated bouts of professional care for a problem which has become intolerably severe or attracted attention, to on the one hand, intervening before things have descended to this point, and on the other, locating treatment as often merely the first step to (as advocated by a US expert) extended monitoring and care – “recovery maintenance”. Accordingly the focus shifts from inside the clinic 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, sustained by ties to family, community, and productive activities. Explore this monograph to appreciate what this means for the roles of commissioners, services and treatment staff.

Optimistic and enthusing though it is, some interpretations of ‘recovery’ have their less clearly positive sides, including the implication that only patients who have left treatment can be recovered, an associated push to limit treatment durations rather than letting patient and clinician decide, a seeming demand that problem substance users develop lives more fulfilling than many who never had these problem, and a tendency to elevate abstinence as the prime objective and to relegate moderation and harm reduction to secondary importance. In the UK, however, these features seem to have been targeted more at the treatment of problem drug users rather than the less politically sensitive treatment of problem drinkers.


Highlighted study Our highlighted study tested the key methodology of the US NIATx partnership featured in cell D2’s bite for improving patient access and retention across a treatment system – the learning collaborative model. In the process it revealed the great strength of randomised trials; they can generate truly surprising results which by eliminating extraneous influences also eliminate alternative explanations. The result can be to force the researcher to re-evaluate the expectations which led to the study. Lead researcher in this case was NIATx director Dave Gustafson, whose organisation promotes the model he tested.

Check the free source for the study and you will see that the model’s cheapest method – monthly, expert-led teleconferences between staff from different clinics – made no significant difference to any of the processes it was intended to improve. The other way services could learn from each other – face-to-face versions of the teleconferences – were associated with improvements in waiting time for treatment, but not in retention or patient numbers. Given the weak performance of these methods, it comes as no surprise that adding them to the mix did not improve on just having an expert quality improvement coach to support and guide clinic staff. In other words, the collaborative bit of the learning collaborative model rarely generated improvements, and created no added value over and above the non-collaborative approach of assigning each clinic an expert guide. Coaching was also much cheaper than both face-to-face conferences and the combined intervention.

The message for commissioners seems inescapable: if you are responsible for treatment provision across an area, employ a quality improvement expert and set them to work with each service; don’t waste money getting services to learn from each other. That way you will at least give more patients a greater chance of getting better, though whether the process actually improved substance use outcomes is unknown. You might counter that in the US context, the clinics were profit-making businesses hardly likely to share tips, but in fact, none were. Apart from the usual caveats, notice that the ‘collaboratives’ were not natural networks, like services with the same catchment area, doing the same kind of work, or seeing the same kinds of patients. Within each US state, each clinic was randomly allocated to the different improvement methods. And there you have the great weakness of this kind of randomised trial: by eliminating ‘extraneous’ influences, it risks eliminating some which are not extraneous at all, but essential to the intervention working – perhaps in this case, a common interest across collaborating clinics.

Issues to consider and discuss

How many drinkers should be in treatment? A complex issue to which we have devoted a hot topic essay, abridged here. It argued that depending on where you draw the line, England’s performance in ensuring needy drinkers enter treatment can look anywhere from an abysmal 7% to an excellent 44%. Those absolute figures may be wrong, but the point here is to show how the figure can differ under different assumptions of what counts as being ‘in need of treatment’.

The proportion of drinkers in England in treatment of those in need of treatment under different assumptions

Let’s start with how many people are in specialist treatment – in England, about 115,000 adults during 2013/14. Based on a 2007 survey, this amounts to about 7% of all 1.6 million drinkers experiencing harm from their drinking. This can be narrowed down to the approximately 1 million adults who also score as at least mildly dependent on alcohol. Now the treatment caseload represents about 11% of dependent drinkers who might need this help. Narrowing down further, in 2011 NICE calculated that 260,000 adults were moderately dependent or worse. Now the caseload represents 44% of those in need. The lower figure can be justified as the percentage of all those who might need help, the higher as perhaps closer to those who really do need and will enter (1 2) specialist treatment to overcome their dependence.

The population in need of treatment is further constricted if we accepted the view that diagnosing an alcohol use disorder requires not just harm from drinking, but evidence of a pathologically impaired ability to control drinking. Compared to standard clinical criteria, applying this ‘harmful dysfunction’ diagnosis to US figures slashed the proportion whose need for treatment had never been met from 34% to just 4%.

So while we may suspect that capturing 115,000 of England’s problem drinkers in treatment is not enough, there is no clear way to determine the degree to which this is the case. It does however seem that Scotland is doing much better. There it has been calculated that in 2012 the treatment caseload represented about 1 in 4 alcohol-dependent adults. The closest parallel calculation for England was that in 2012/13 the numbers in specialist treatment represented about 1:14 of all possibly dependent adult drinkers. ‘Closest’ left several differences in the way the figures were calculated, but probably not enough to vitiate the conclusion that a higher proportion of drinkers in need of treatment get it in Scotland than in England.

Where would you draw the line? Harmful drinking, at least mildly dependent, moderately dependent, or severely dependent? If for the benefits they feel they get, the drinker has freely chosen to shoulder the harm they experience and their dependent state, can they be considered in need of treatment? Does where you draw the line depend on whether you want to portray Britain’s treatment-access performance as abysmal or as excellent, perhaps depending in turn on whether you wish to argue for more resources or contain expenditure? Is severity (of drinking and/or dependence) the right criterion? How about the duration of heavy drinking, whether the drinker wants treatment, or how many patients we want to afford to treat?


Is payment by results the way to go? It seems to make sense to pay services to achieve desired outcomes like abstinence, rather than to do things which it is hoped will lead to those outcomes, but may not. Certainly it made sense to the UK government, which has advocated this mechanism and set up ‘payment-by-results’ pilot schemes in eight areas to test it, and which remains enthusiastic, at least in respect of probation service providers. Also not to be lightly dismissed is the hunch of two of the world’s most respected addiction experts (one of whom was appointed to advise on addiction treatment in England) that payment-by-results arrangements are among the most promising strategies for improving treatment outcomes. Are they are on the right track?

Look at the discussion of the schemes in our commissioning hot topic. It points out that such schemes tend to be uniform, concrete and prescriptive about what they expect from different treatment services and what counts as success for different patients. Is that desirable, preventing services glossing over their shortcomings or claiming unverifiable ‘soft‘; outcomes? Or does it stifle patient-centred practice, preventing treatment objectives being based on the patient’s priorities? Maybe both?

A US study listed above seems to confirm concerns that (like contingency management incentives for patients) payment-for-performance systems engender a mentality of doing just enough to get the money, but no more. Surely a charity or health service should strive to do the best for its patients, without external incentives? Yet without these, services may stay unstretched within acceptable-quality comfort zones.

To these in-principle issues can be added the particular criteria prioritised by the schemes. Though introduced in the name of recovery, UK schemes place a premium not on the long-term contact presupposed by the recovery vision and associated understandings of addiction, but on discharging patients who then are not seen again for at least a year. Does that incentivise the achievement of lasting recovery, or tempt services to counterproductively place hurdles in the way of treatment re-entry? If addiction of the type seen at specialist services at least behaves like a chronic, relapsing condition, is it appropriate to punish services for post-treatment relapse? Should (as explored in cell D2’s bite) we instead incentivise long-term support?

Research to help answer these questions is almost entirely lacking; in evidence terms, payment-by-results in health and social care of any kind is a leap in the dark. A review of reviews could find no evaluations which reported on patient outcomes, and a review specific to drug and alcohol treatment could find “no peer-reviewed evidence that [payment for performance] ... improves client outcomes post-treatment”. When a leading UK commentator on such schemes reviewed the literature, he found “consensus” about the evidence base – consensus that is “not able to give a clear indication as to whether payment by results works”, and also that “unexpected, often perverse, consequences are commonplace”.

% of patients recorded as not drinking and leaving treatment no longer dependent during English payment-by-results pilots versus comparators. Shows that on all four substance use comparisons things were worse in the pilot areas once the schemes had started

On all four substance use comparisons things were worse in the pilot areas once the schemes had started

Evidential uncertainty and the risk of counterproductive effects are presumably among the reasons why the English schemes were evaluated pilots. Analysed by the Department of Health, initial results were not encouraging. The report used routinely collected treatment monitoring data to compare performance in the pilot areas in the 11 months from April 2012 to February 2013 with the same months a year before, when the pilot schemes had yet to start. This before-after comparison within the pilot areas was supplemented by a contemporaneous comparison between the pilot areas and the rest of England, using data generated from April 2012 to February 2013.

During this time, in the pilot areas records were available for 3081 patients whose problems were primarily with alcohol. Against both comparators, the proportion recorded as no longer drinking (just over 40%) had worsened by 2% chart. The proportion recorded as leaving treatment free of their dependence on alcohol or other drugs – for the government, the key indicator of recovery – was 15% lower than in the rest of England and 11% down on previous performance in the same areas, statistically significant differences. In other words, on all four substance use comparisons things were worse in the pilot areas once the schemes had started.

Beyond substance use, there was some indication that more patients with housing problems had resolved these, but none that patients’ quality of life had been improved by introducing the pilot schemes. “Mixed” was the document’s characterisation of the results; ‘disappointing’ might also have been justified. These were, however, early days, and the intention was that “This report will now be updated with the latest information every three months.” If that happened, it has been kept very quiet.

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