Send email for updates


About updates
Effectiveness bank home page. Opens new windowResearch analysis

This entry is our analysis of a study considered particularly relevant to improving outcomes from drug or alcohol interventions in the UK. The original study was not published by Findings; click Title to order a copy. Links to other documents. Hover over for notes. Click to highlight passage referred to. Unfold extra text Unfold supplementary text The Summary conveys the findings and views expressed in the study. Below is a commentary from Drug and Alcohol Findings.

Title and link for copying Comment/query to editor

Adult substance misuse statistics from the National Drug Treatment Monitoring System (NDTMS) 1 April 2016 to 31 March 2017.

Public Health England
Public Health England, 2017.

The annual accounting of the treatment caseload in England registers a continuing fall in total numbers and decreasing success with opiate users, while success with drinkers has increased and has for the last few years remained relatively high and stable. An ageing population of opiate users is the proposed explanation for the former trend – but why hasn’t a similarly ageing alcohol caseload also eroded success rates?

Summary This report combines information on patients receiving specialist treatment for drug problems and/or for problem drinking in England during the financial year 2016/17, other than those treated in prisons. While people who seek treatment for drug versus alcohol problems share many similarities, there are also clear differences, so the caseload is divided into four mutually exclusive groups:
• any mention of opiate use in any treatment episode during the year would result in the client being categorised as an opiate client, irrespective of other substances cited;
• clients who present for treatment related to non-opiate drug use but not opiates or alcohol are classified as non-opiate only;
• clients with non-opiate drug and alcohol use problems (but not opiates) recorded in any treatment episode during the year are classified as non-opiate and alcohol;
• clients who present with problems related to alcohol but no other substances are classified as alcohol-only.

Other substance use categories may not be mutually exclusive.

All in treatment during the year

279,793 patients were treated at drug and alcohol services during 2016/17, a 3% reduction on the previous year, itself 2% down on the year before. At 146,536 or 52% of the total, patients treated for opiate dependence formed the largest proportion, but were nevertheless 2% fewer than last year and 14% fewer since the peak of 170,032 in 2009/10. The decrease in opiate clients is most pronounced in younger age groups; the number aged 18–24 has fallen by 83% from its peak in 2005/06 to 3763, and the number aged 25–34 has fallen by 52% to 36,177 from its peak in 2008/09.

The second largest group in treatment was the 138,606 problem drinkers. For 80,454 their substance use problems were confined to alcohol. Both numbers were down (respectively by 11% and 12%) from their peaks in 2013/14.

Alcohol-only and opiate use patients tended to be much older than those who presented for problems with other substances. The median (middle of the range) age of alcohol-only clients was 46 years and 39 for opiate use clients.

Men comprised 69% of the treatment population in 2016/17 – 73% of the patients using drugs, and 61% presenting with alcohol use problems only. White British people formed 85% of the caseload and a further 5% were recorded as from other white groups. No other ethnic group made up more than 1% of the total treatment population.

Treatment starters and leavers

Of the 279,793 in treatment during 2016/17, nearly half – 131,216 or 47% – had started treatment that year either for the first time or after a break (which may have been during 2016/17 itself); nearly all (98%) had waited three weeks or less to start their treatment; the delay averaged just over two days.

Patients starting treatment for crack cocaine problems (when not used alongside opiates) increased steeply by 23% to 3657, after having increased by 3% the year before. The increase in 2016/17 was seen in nearly all age groups, though numbers remain well below the peak of 5765 in 2008/09. Rising to 21,854, there was also a 12% increase in individuals starting treatment for both crack cocaine and opiate problems, seen primarily in those aged 45 and over, though the total remained below the peak of 25,735 in 2008/09. These increases reflect the rise in the prevalence of crack use probably driven by changes in purity and affordability. In the general population, crack cocaine use has been estimated to have increased by 10% in England between 2010/11 and 2014/15, and in numbers from 166,640 to 182,8281.

11,657 patients aged 18–24 started treatment in 2016/17, most citing problems with cannabis, alcohol or cocaine (respectively: 6322, 54%; 5221, 45%; and 3113, 27%). The number of under-25s (re)starting treatment has fallen by 45% since 2005/06, reflecting shifts in the pattern of drinking and drug use in this age group over the last 11 years. Though numbers remained low, there was a 30% increase in under-25s starting treatment for problems with crack cocaine (not used alongside opiates), an increase from 281 to 364 – the first increase in this figure since it peaked at 1193 in 2007/08.

New psychoactive substances were a presenting problem for 1450 individuals starting treatment in 2016/17, a 29% downturn on the peak of 2042 the previous year, and only 1.1% of all presentations. This fall was mainly driven by a near-halving in those aged under 25, down from 627 in 2015/16 to 321. These patients are more likely than others to be in acute housing need at the start of treatment (18% v. 7%). While relatively small in numbers, when problem use of new psychoactive substances was combined with opiate use, 48% reported housing problems and the same proportion had been referred from prison or probation, compared to 27% and 19% among opiate clients overall.

There was a 23% fall between 2015/16 and 2016/17 in patients starting treatment for ecstasy use, down from 1318 to 1013, and a decrease of 70% in the number starting treatment for mephedrone use, down from 1647 to 502. In both cases, these are the lowest numbers recorded at least since 2005/06.

127,475 individuals left the treatment system in 2016/17, 62,500 or 49% after having successfully completed their treatment free of dependence, down from a peak of 53% in 2012/13. Around 35% to 41% of all patients in treatment during the year whose problems did not include opiate use completed their treatment free of dependence. The corresponding proportion for opiate users was just 7%, reducing the overall proportion for all patients to 22%.


Findings logo commentary Focusing on by far the largest two sectors of the treatment caseload – problem opiate users and problem drinkers – the broad themes introduced in the following paragraphs are explored in greater detail for opiate users under the heading Figures and trends for opiate users, and for drinkers under the heading Figures and trends for drinkers. To reveal trends we largely rely on charts; to see these click the Open chart icons.

The analysis covers the 11 years from 2005/06 to 2016/17, the period for which reliable data has been back-calculated and published. An important caveat is that alcohol treatment providers were not fully incorporated into the monitoring system until 2009/10. That must partly and perhaps entirely account for the steep increase between 2005/06 and 2009/10 in patients recorded as in treatment solely for a drinking problem.

In the 2010s treatment become a choice made by older opiate users and turned away from by younger users

Reflecting earlier trends in users in the population, numbers of opiate users in treatment fell from 2009/10 and their ages increased as younger adults turned away from heroin use. However, changes in the age composition of opiate users in the population was not the whole story. Since at least 2006/07, the gap between the proportion of older and younger opiate/crack users engaged in treatment has widened until treatment seems to have become a choice made by older opiate users but rejected by or unavailable to younger users. Also at least partly reflecting general population trends, numbers and proportions of younger problem drinkers in treatment have been falling while older patients have become more prominent.

The main indicators of treatment success – successful completion and, even more so, successful completion and no return within six months – rose for drinkers, but from 2011/12 fell for opiate users, presumed due to an ageing caseload more entrenched in their addiction and with few resources left to enable them to manage outside treatment. Why according to the same yardsticks, similar trends have not eroded the success of alcohol treatment remains to be explained.

Since records could be relied on in 2005/06, the equivalent of around half or more of the estimated population of opiate users in England in each year have been in treatment during that year, primarily (along with psychosocial support) being prescribed opioid medications like methadone which substitute for heroin or the other illegal opiate-type drugs. A corresponding estimate for dependent drinkers is that in 2014/2015 the equivalent of about a fifth were in treatment, a proportion that has remained the same since 2010/11.

Overall caseload shrinking due to fewer new starters

The treatment caseload has slowly been shrinking since around 2009. By 2016/17, 279,793 individuals were recorded as having been in treatment at drug and alcohol services, the lowest figure since 2009/10 when the total was 311,667. Diminishing influx of patients into treatment is one reason. In 2016/17, 131,216 had started treatment during the year (blue line in chart) either for the first time or after a break, down from the highest figure recorded of 147,578 in 2008/09. Another reason is the increased outflux of patients departing treatment. Open chart

Influx and outflux trends are, however, complicated by the degree to which drinkers have been represented in the figures. Up to 2009/10, rapidly rising numbers of alcohol-only patients as alcohol services joined the recording system ( below) helped create the impression of an increasingly dynamic system, because drinkers are more likely than opiate users to leave treatment, and to do so after recording a successful exit. After the influx of recorded alcohol patients flattened out, proportions of patients leaving or leaving successfully crept up more slowly, largely due ( below) to increasing success with drinkers, but since about 2012 to 2014 these proportions have remained more or less stable.

In greater detail, reduced influx was due to fewer patients starting treatment for the first time (black portion of chart), leaving a caseload composed largely of patients continuing in or returning to treatment. The early years of these figures will have been disproportionately inflated by the 1 April 2005 cut-off (1; personal communication from PHE) for registering prior treatment. Earlier treatment journeys are ignored, meaning a patient who (for example) entered treatment in 2006/07 having previously been treated in 2004/05 would nevertheless be counted as entirely new to treatment. Considering trends from 2009/10 reduces this distortion, and also that due to increased recording of drinkers in the previous years. In 2009/10, 73,240 patients were recorded as entirely new to treatment, 23% of all patients in treatment. By 2016/17, the corresponding figures were 46,396 and 17%. Open chart

Opiate caseload falling, alcohol becomes more prominent

In terms of substances used, the shrinking caseload largely reflected the fall in the number of opiate users in treatment from a peak of 170,032 in 2009/10 to 146,536 in 2016/17, down 14%. Numbers of alcohol patients have remained fairly stable since alcohol services fully joined the system in 2009/10, but fell in the last two years to 138,606, 7% down on 2009/10. Since 2012/13 problem drinkers have numbered almost as many as problem opiate users. However, many combined drinking with problem drug use, which may have included opiate use, and it is unclear how many can be considered primarily dependent on alcohol; in 2016/17, for just 80,454 was drinking their sole substance use problem. Open chart

The previous figures related to patients in treatment at some time during the year, including those continuing in treatment from previous years. More indicative of current and future trends in the ‘demand’ for treatment are patients who started treatment during the year. Even before alcohol treatment services had been fully incorporated, on this score alcohol as a primary or secondary problem substance dominated. In 2016/17, 79,202 patients started treatment citing drinking as a problem, 60% of all treatment starters. At 43,142, far behind were patients with opiate use problems, 33% of all treatment starters, down in numbers from 55,493 in 2009/10 and in proportion from 38%.

The dip in opiate use starters coincided with the ‘heroin drought’ of (roughly) 2010 to 2012, thought via changes in the availability and price of heroin to have reduced the amount consumed by each user as well as the number of users. Since then the reduction in opiate use starters has levelled out, reinforcing the impression of a one-off shock to the system. Because new heroin users can take several years to enter treatment, an earlier influence may ( below) have been the modest reduction in the number of problem opiate users in the population since 2005/06. Open chart

Ageing caseload raises profile of health problems

Since the ageing opiate-using caseload and the ageing set of problem drinkers constitute the great majority of all patients in treatment in England, the total caseload too has also been ageing. From 42% in 2005/06, by 2016/17 patients aged 35 or more constituted 69% of the caseload, while the under 25s fell from 17% to 6%. Up to 2009/10 some of this ageing may have been due to the increasing incorporation of alcohol services into the monitoring system, but since then it has continued, 35-and-overs rising from 54% to 69% of the caseload. To the challenges of treating dependence are increasingly being added those of dealing with the aftermath of many years of substance use and the diseases of age. Open chart

Among all patients starting treatment either for the first time or returning the picture was very similar. From 41% in 2005/06, by 2016/17 those aged 35 or more constituted 64% or nearly two-thirds of the caseload, while the under-25s shrank from 19% to 9%. Open chart

Figures and trends for opiate users

First-time treatment starters plummet

The fall in the opiate-problem caseload and in patients starting/returning to treatment for these problems is overwhelmingly due to the fall in numbers starting treatment for the very first time – or at least, structured treatment of the kind recorded by the monitoring system. The steepness of this fall was remarkable, from 50,383 in 2005/06 to 6117 in 2016/17. For every 100 patients first starting treatment for opiate use problems 11 years before, in 2016/17 just 12 did so. But as explained above, the early years of these figures will have been disproportionately inflated by the 1 April 2005 cut-off for registering prior treatment. Taking 2009/10 as the base reduces this distortion; that year 17,647 patients were recorded as starting treatment for the very first time, meaning that for every 100 patients first starting treatment for opiate use problems in 2009/10, by 2016/17 just 35 did so. Open chart

Public Health England’s analysts have modelled the fall in first-time opiate use treatment starters from January 2011, minimising both the inflation in earlier years and the immediate impact of the ‘heroin drought’ of 2010 to 2011. Based on figures up to November 2016, there remained a steady fall estimated at from nearly 1100 a month down to 860 a month.

The figures for England start in 2005/06. UK-wide figures go back another two years (1 2), but relate to patients whose opiate use was their primary problem, while the English figures concern patients who may combine opiate use with other substance use problems. Dominated as they are by England, these figures suggest that for that country as for the UK, 2005/06 was a turning point in the numbers of new opiate use patients. For at least two years before, the influx of never-before-treated opiate use patients in the UK was rising steeply, but from 2005/06 it fell almost as steeply. Open chart

To a degree, the peak in 2005/06 might have been an artefact of improved recording. The monitoring system which underpins both the English figures and the English component of the UK figures was new and finding its feet, meaning that better recording might partly account for the rises leading up to 2005/06. For example, compared to the year before, the relevant information on treatment history was recorded for 11% more patients overall, missing data having been reduced from 29% to 18% of records.

That does not detract from the remarkable UK-wide drop in new opiate treatment clients from 2005/06, leaving an opiate use treatment population UK-wide and in England overwhelmingly composed of patients continuing in or returning to treatment. By 2016/17, 71% of opiate use patients in England had been continuously in treatment from the year before and another 25% had returned to treatment after a break (which may have been in 2016/17 itself), leaving just 4% starting treatment for the first time in 2016/17 who did not restart it during the same year.

Apart from the one-off shock of the heroin drought, there seem two broad explanations for the shrinkage in new opiate use patients. One is the relative ‘stickiness’ of dependent opiate use and/or of its treatment. As Public Health England has commented, compared to other drugs the treatment history of the opiate use caseload “is indicative of heroin clients being more likely to drop out of treatment and to subsequently re-present, or to relapse after completing a treatment episode and to seek treatment again as a result.” The other factor is the waning in the uptake of heroin use in the general population, resulting in fewer new treatment clients and a progressively older and harder to treat opiate-using caseload, themes explored further below.

Opiate use treatment becoming less ‘successful’

Despite the recent policy emphasis on moving problem drug users through and out of treatment, since at least 2005/06 a stubbornly persistent three-quarters of opiate use patients in treatment at some time during the year were still in treatment at the end and just a quarter had left, most without being recorded as having completed treatment and overcome their dependence. In 2016/17, just 10,439 left having recorded this apparent success, 7% of all those in treatment that year and 26% of all 39,513 who left treatment. Open chart

It seems that nearly all these no-longer-dependent treatment completers notched up for their areas a further contribution to the important public health indicator of not just completing treatment successfully, but also not returning to treatment within six months – presumed indicative of no relapse to dependent opiate use and a lasting treatment success. Forming (in the year up the end of December 2017, latest figure at the time of writing) 6.7% of all opiate using patients in treatment in 2016/17, these estimated 9818 patients were about 621 fewer than all no-longer-dependent treatment completers. This last number can be thought of as referring to apparently successful treatment leavers known soon to have relapsed and re-entered treatment. As with alcohol-only patients (below), it increased from 396 the previous year and as a % of all apparently successful treatment leavers from nearly 4% to 6%, though this last figure is provisional.

On these criteria, considered above all emblematic of successful treatment, things have been getting slightly worse. From a high point of about 9% of all opiate patients in 2011/12, in 2016/17 successful leavers in total and those not returning within six months formed just 7%. From that year too, successful departure also accounted for a diminishing proportion of all opiate users leaving treatment, steadily falling from 37% to 26% in 2016/17, while the ‘drop-out’ proportion increased from 22% to 39%.

As a broader indicator of successful treatment, there is reason (1 2) to include opiate patients retained in treatment at the end of the year. According to this indicator, over the period 2011/12 to 2016/17 the proportion of successful patients (by virtue either of being retained or leaving free of dependence) fell from a peak of 85% to 80%, almost completely reversing the gains since 2005/06.

An ageing and more entrenched heroin-using caseload

For Public Health England, the explanation for diminishing success is an ageing and increasingly treatment-resistant opiate use caseload. A sign of the ageing phenomenon, in England the number of 18–24-year-olds starting treatment for opiate use problems for the first time or after a break fell precipitately from 11,351 in 2005/06 to 2053 in 2016/17, while patients aged 35 or over rose from 19,086 to 27,126. Open chart

It was a similar story among the entire treatment population, including those continuing in treatment from previous years. The number of 18–24-year-olds with opiate use problems plummeted to a sixth of the number 11 years before, down from 22,681 in 2005/06 to 3763 in 2016/17, while corresponding patients aged 35 or over more than doubled from 51,154 to 106,596. Open chart

With older opiate-dependent people staying in or returning to treatment, and fewer young adults using the drugs at all or in a way which results in treatment, in England and the UK in general the opiate-using treatment population is on average getting older. According to Public Health England (1 2), this ageing cohort is often in poor health, with a range of vulnerabilities associated with long-term drug use including social isolation and more entrenched, hard-to-treat addictions. Incidentally, this explanation implies that the persistence of dependent opioid use among patients in treatment should not be taken for granted or attributed to an inherent power of the drug. Persistence and relapse occur in the context of available treatment and social resources devoted to securing lasting remission, and the limited recovery resources of dependent users of the kind driven or forced to seek formal help.

An ageing (and more entrenched?) heroin-using population

Trends in the opiate treatment caseload in England are said to mirror general population trends consistent with an ageing population of opiate (mainly heroin) users who started using in the mid-90s or before. Because they are only partially being replaced by younger recruits, the result is an increase in the average age of problem opiate users in the general population – the pattern thought to account for increases in drug-related deaths. Despite increases in the number of older problem drug users (representing mainly opiate users), the dwindling recruitment of new, younger users, has led to falls in the estimated total in the population between 2004/05 and 2014/15. Open chart

Studies of the prevalence of problem drug use between 2004/05 and 2014/15 which generated these conclusions defined problem drug users as users of opiates and/or crack whose use has brought them into contact with treatment services or the criminal justice system. The estimated number aged under 25 fell by nearly 60% from 72,838 in 2004/05 to 30,190 in 2014/15. Over the same period, estimates of the numbers aged 35 to 64 rose by 56% from 114,459 to 178,785. These shifts cumulated to a total population of opiate/crack users estimated to have fallen from 327,466 in 2004/05 to 300,783 in 2014/15, and the number using opiates in particular from 281,320 to 257,476.

Numbers injecting opiates or crack were also estimated to have fallen steadily from 137,141 in 2004/05 to 87,302 in 2011/12, the latest estimate (1 2).

Though fewer than in the mid-2000s, the numbers of problem opiate/crack users and of problem opiate users were slightly higher in the latest estimates for 2014/15 than in the previous estimates for 2011/12, and the decline in problem drug users aged 15–24 had slowed down, suggesting that the era of diminishing opiate/crack use had come to an end.

Underlying these trends in total numbers were changes in the influx of new (and therefore relatively young) users of opiates or crack – figures which also suggest the era of diminishing opiate/crack use may have come to an end. Depending on the methodology, in England the reduction in numbers starting to use these drugs amounted to 20% or 45% between 2005 and 2013, when analysts could be confident that no more than 10,000 initiated use – a rate per 1000 of the population 11 times lower than some local estimates from the epidemic years of the 1980s and early 1990s. However, from 2011 to 2013 the decrease ended and uptake estimates slightly increased, though the analysts could not be confident this was anything more than a flattening out of what might prove a continuing decline.

To reach these conclusions they calculated the delay in entering treatment after starting opiate/crack use. Treatment initiates in 2004 and 2005 had typically started using six or seven years before starting treatment, at the peak of the spread of heroin use. However, by 2007 the pattern was changing; among that year’s treatment initiates, the peak period for starting treatment was just a year or two after starting to use the drugs. This trend solidified until for 2012’s treatment initiates, the mid- to late-90s hump in year of starting opiate/crack use was merely a small hill, and the dominant pattern was entry into treatment either the year of starting use or within two or three years. These rapid treatment entrants also tended to have started using the drugs beyond the typical 18–22 age range. One result was that the most common age of initiation of heroin/crack use among first-time treatment entrants rose from 18 in 2005 to 25 in 2013. Barring mistakes in the data, it seemed to the analysts that “There is a genuine shift towards new initiates being older, and for them to present to treatment much faster than in previous years.”

In the present context the significance of this finding is that even if the profile of opiate/crack use in the general population was unchanged, all else being equal the speeding up of treatment entry among older initiates into opiate/crack use would tend to increase numbers in treatment and increase the average age of the caseload. In other words, to a degree the ageing of the opiate use treatment caseload will be due not to the ageing of opiate users in the general population, but to older users starting treatment more quickly. This speeding up has obvious limits, so the effects on caseload numbers and age profile should flatten out.

Treatment engages half of all opiate/crack users but engagement of young falling relative to older users

The research analysed in the previous section enables us to calculate how many of the estimated population of opiate users and opiate/crack users in England in a given year are in treatment – the treatment system’s performance at engageing potential patients. In this respect, the opiate use treatment system based on prescribing of substitute drugs like methadone has internationally a creditable record, engageing about 60%.

The latest year for which we have population estimates is 2014/15, when 52% of all estimated problem opiate/crack users were in treatment at some time, down on the peak of 59% in 2010/11 but still up on the 45% estimate for 2005/06. If we assume continuance of trends in the population between 2011/12 and 2014/15, then the ‘guesstimate’ proportions for 2015/16 and 2016/17 would reduce slightly more, reaching 50% in the latest year. Open chart

For opiates in particular (with or without crack), the proportion in treatment in 2014/15 was 59%, slightly down on the peak of around 65% between 2008/09 to 2010/11, but again still an advance on 2005/06. If we assume continuance of trends in the population between 2011/12 and 2014/15, then the ‘guesstimate’ proportions for 2015/16 and 2016/17 reduce slightly more, reaching 57% in the latest year. Since records could be relied on in 2005/06, around half or more of the estimated population of opiate users in England in each year have been in treatment during that year. Open chart

Broken down into age groups, population estimates set against treatment numbers show that part of the reason for the average increase in age of the treatment population has been a widening gap between the ‘capture’ rates of older versus younger opiate/crack users – the proportions engaged in treatment. Up to 2010/11 this took the form of a progressively greater capture of 25s-and-over, and possibly a progressively lesser capture of the under-25s. After this year the capture rates at all ages seems to have fallen, but more steeply among the under-25s, down from 32% in 2010/11 to 19% in 2014/15 (the year of the latest population estimates) – in absolute terms a 13% drop, but in proportional terms, 41%. Over the same period the capture rate among older users fell from 63% to 56%, in absolute terms a 7% drop and in proportional terms about 11%. By 2014/15 an estimated 56% of problem drug users aged 25 and over were in treatment, but just 19% aged under 25, a gap of 37%. In 2005/06 this gap was just 11%, the year after, just 15%. Open chart

One reason ( above) could be that in the more recent years older initiates into opiate/crack use have increasingly been starting treatment more quickly than their younger counterparts. Relative to previous years, in the 2010s treatment seems to be becoming an option for fewer opiate/crack users across the age range, but one most decisively closed to or rejected by younger users.

Figures and trends for drinkers

For alcohol treatment numbers there are at least three variants: drinkers for whom this is their sole recorded substance use problem; those also recorded as problem users of other substances; and up to 2013/14, those treated primarily for drinking problems, who may secondarily be using other substances. The account below draws on all three figures. Before in 2009/10 alcohol treatment services were fully incorporated into the monitoring system, absolute numbers are an unreliable guide to the caseload, but there is no reason to believe that the proportional composition of the alcohol caseloads will have radically altered as more services came on board.

How many drinkers are in treatment?

How many drinkers are in treatment depends on how these are defined. The highest figure for 2016/17 is 138,606, but for some of these patients their drinking was subsidiary to their drug use. In 2013/14 – the last year when alcohol treatment figures were separately reported – of the 155,381 patients whose substance use problems included alcohol, only 114,920 were being treated primarily for drinking problems. Applying the same ratio to 2016/17 suggests that around 102,513 patients were primarily being treated for their drinking. The same type of calculation, but based on numbers of patients with alcohol as their sole problem, results in a reassuringly similar estimate of 100,880 patients primarily being treated for drinking problems in 2016/17. Putting the two together, it seems a fair guess that the figure was about 101,700. Open chart

This can be refined down further to patients whose sole problem substance was alcohol, a number which peaked at 91,651 in 2013/14 and ended at 80,454 in 2016/17. The upshot is three figures for treated problem drinkers in 2015/16, ranging from 80,454 solely with a drinking problem, to an estimated 101,700 with a primary drinking problem but perhaps also using other substances, and finally to 138,606 including many for whom alcohol was not their sole or even their primary substance use problem.

Based on the highest version of the figures, since in 2009/10 alcohol treatment services were fully incorporated into the monitoring system, numbers with drinking problems have fluctuated at around 150,000, peaking at 155,381 in 2013/14 but since then falling each year to end at 138,606 in 2015/16. Over that period the proportion of drinkers recorded as also being treated for other substance use problems has remained steady at 41–42%.

Highlighted above is that since 2012/13 problem drinking patients have numbered almost as many as problem opiate users. The gap closed due to the greater recruitment of drinkers into treatment rather than to their staying longer than opiate users, an issue explored below.

Alcohol treatment more of an in-and-quickly-out proposition

Problem drinkers (including those for whom this is a subsidiary drug) record a radically different treatment history to opiate use clients. New opiate use clients are scarce and most have continued in treatment from the previous year without a break in the focal year. Continuous treatment is much less common for drinkers, and a far higher proportion are starting treatment for the first time. Relative to opiate use treatment dominated by maintenance prescribing, alcohol treatment dominated by psychosocial approaches is more of an in-and-quickly-out proposition. Open chart

Some illustrative figures are that while in 2016/17, 71% of opiate use patients were continuously in treatment from the year before, for problem drinkers it was 43%, a figure which has remained at around this level since 2006/07. The remaining roughly 60% either returned to treatment after a break or were first-time treatment entrants (only for alcohol-only patients do we have this breakdown; see below). Continuous and longer term treatment is related to the dominance of maintenance prescribing for opiate users. Among patients prescribed medications in 2015/16, this treatment had lasted a year or more for 61% of opiate patients but for only 6% of the relatively few alcohol-only patients prescribed medications, for whom this would primarily have been an aid to short-term detoxification rather than a longer term treatment.

Though ‘contaminated’ by the fact that just over 40% of the total alcohol caseload are also experiencing problems with other substances, the figures above show that in England, treatment involving alcohol problems is quite different from treatment involving opiate use problems. A sharper distinction can be made by focusing on the (since 2008/09) just under 60% of the problem-drinking caseload for whom this is their sole recorded substance use problem. Only for the last three years of the figures do we have a breakdown equivalent to that for opiate users. In those years, 29–31% of patients were starting treatment for the first time and 31–35% were continuously in treatment from the year before. The contrast with corresponding figures for opiate user is extreme; in 2016/17 just 4% were starting treatment for the first time and 71% had been continuously in treatment from the year before. Open chart

For alcohol-only patients, numbers of first-time treatment starters have been published back to 2005/06. Among these patients solely with a drinking problem, since 2008/09 the trend in first-time starters has been downwards both in numbers and as a proportion of the alcohol caseload, from 35,170 down to 22,988 and from 45% to 29%. Increasingly, the caseload has been continuing in or returning to treatment.

Alcohol use treatment becoming more ‘successful’

Another contrast with opiate use treatment is that the treatment of patients solely with a drinking problem has on criteria set for public health nationally been more successful, and since 2011/12 yet more successful as the success rate for opiate patients has declined.

Reinforcing the impression that relative to opiate use treatment, alcohol treatment is more an in-and-out process, since at least 2005/06 in all but two years most alcohol-only patients have left treatment during the year rather than still being in treatment at the end, and since 2012/13 the proportion has neared two-thirds. Of these leavers, in 2016/17, 61% were recorded as having overcome their dependence on alcohol, forming 40% of the total caseload compared to 7% of opiate patients. Open chart

Since 2005/06 the proportion of alcohol-only patients leaving treatment free of dependence increased from 18% to 28% in 2009/10, and then increased further until stabilising at 39–40% each year from 2012/13. In numbers the increase was from 24,862 in 2009/10 to a peak of 36,164 in 2013/14, before falling steadily in line with the diminishing overall caseload, reaching 31,982 in 2016/17.

As with opiate patients, we know from other figures that nearly all these no-longer-dependent treatment completers notched up for their areas a further contribution to the important public health indicator of not just successfully completing treatment, but also not returning within six months, presumed indicative of lasting treatment success. Up to the end of December 2017 (latest figures at time of writing), non-returners formed 39% of all alcohol-only patients in treatment in 2016/17 – in numbers, about 31,031 patients, just 951 fewer than all no-longer-dependent treatment completers. This last number can be thought of as referring to apparently successful treatment leavers known soon to have relapsed and re-entered treatment. As with opiate users, it increased from 584 the previous year, and as a % of all apparently successful treatment leavers, provisionally from 1.75% to 3%.

Alcohol caseload also ageing

The typical age of problem drinkers in treatment (including those also using other drugs) has been increasing. From 27% of the caseload in 2005/06 and 32% in 2009/10, the proportion of over-44s increased each year to reach 43% of all drinkers in treatment by 2016/17. In numbers, from 46,996 in 2009/10 the over-44s peaked at 60,815 in 2014/15 before falling to 59,369 in 2016/17 in line with the overall caseload. Over the same period numbers aged below 35 fell steadily from 52,706 to 36,481 and in proportion from 35% to 26%. Open chart

The caseload whose sole problem is drinking are older still. In the three years from 2014/15 – the only years for which figures have been published – just over half (52%, 53% and 54% respectively) were aged 45 or over. Public Health England has warned that many of these people will have been drinking at high-risk levels for some time and are likely to be experiencing illnesses such as liver disease and high blood pressure.

The age breakdown of problem drinkers starting treatment either for the first time or returning was very similar to that of all patients in treatment. As with the total caseload, in 2016/17 43% of patients with a drinking problem (including those also using other drugs) were aged 45 or more and just 7% were below 25 years of age. Also as with the overall caseload, treatment-starters have been increasing in age. In 2005/06, 27% were aged 45 or over and 14% were below 25 years of age; by 2016/17 the proportions were 43% and 7%. From a peak of 9673 in 2008/09, numbers of under-25s fell to 5221 in 2016/17. Open chart

Ageing of the opiate use treatment caseload has been linked to the waning success of that treatment as assessed by successful treatment completions. The explanation put forward by Public Health England is that ageing signifies an increasingly treatment-resistant caseload with more entrenched problems. However, assessed by the same measures, similar age trends among the problem drinking caseload have coincided with increasing success above. Perhaps relevant is that the kind of ill-health associated with many years of heavy drinking, such as liver disease and hypertension, gives drinkers reasons to cut back because alcohol is a direct contributor and cutting back will help. In contrast, heroin does not directly cause serious illness, and cutting back on its use will not necessarily relieve accumulated illnesses associated with a drug using lifestyle. Heroin and its legal substitutes are also highly effective at relieving pain, anxiety and discomfort; illnesses which cause these complaints may give users reasons to continue rather than stop using illegal opiates and/or legal substitutes.

Ageing population of dependent drinkers

Trends in the size and composition of the treatment caseload partially reflect trends in dependent drinking in the general population, revealed by a regular if infrequent national survey of mental health which generates estimates of the prevalence of dependent drinking among the general household population in England. Based on a score of at least 16 on the AUDIT questionnaire (indicative of harmful drinking or mild dependence or more severe drinking), the 2014 report concluded that across the years 2000, 2007 and 2014, at 3.4% to 3.8% of the 18+ population, “Overall, levels of harmful and dependent drinking have remained stable.”

But within the overall stability there were signs that such drinking patterns had become more common in older people, most clearly the 55–64-year-olds, among whom the proportion doubled from 1.4% in 2007 to 2.8% in 2014, but also among 45–54-year-olds, where it rose from 2.0% to 2.4% and finally in 2014, to 2.8%. At the same time such drinking became less common among 16–24-year-olds (falling from 6.2% to 4.2%) and among 25–34-year-olds (falling from 5.2% to 4.4%). These trends parallel those in the overall treatment population and in treatment starters, among whom older age groups also became more prominent and younger patients fewer above.

1 in 5 dependent drinkers in treatment

A key indicator of treatment’s potential to reduce harm is the proportion of drinkers in need of treatment who actually receive it. The top of this fraction – the treatment caseload – is simply counted, but the number of drinkers in need of treatment has to be estimated.

Based on the national survey cited above, it has been estimated that in 2014 there were 595,131 adult dependent drinkers in England potentially in need of specialist treatment, about 1.4% of the same-age population – an estimate subject to considerable uncertainty. The estimate was based not just on suggestive evidence from scores on the AUDIT screening questionnaire, but also confirmatory evidence from a questionnaire specifically intended to assess alcohol dependence. To qualify as dependent respondents had to score:
• 16 or more on AUDIT (harmful drinking and/or mild dependence or worse) and 16 or more on the Severity of Alcohol Dependence Questionnaire (at least moderate dependence); or
• 20 or more on AUDIT (probable dependence) and 4 or more (at least mild dependence) on the dependence questionnaire.

During the corresponding financial year (2014/15), treatment engaged 150,640 people whose substance use problems included alcohol and 89,107 whose sole problem was drinking. Assuming that the ratios of these figures to patients with a primary drinking problem remained the same as the year before generates two estimates for 2014/15 (111,414 and 111,730), of which the mid-point is 111,572 – probably close to the unknown number of patients in 2014/15 treated primarily for a drinking problem. How many of these were dependent is not recorded. Assuming all were, it means that the equivalent of 19% of all dependent drinkers in England had been in treatment at some time during the year. Of this population, survey responses indicated that 57% might consider treatment on the basis that they expressed some desire to cut down, amounting to 341,376 adults. The treatment caseload was equivalent to about a third of this number.

Based on trends in hospital admissions, the same report back-calculated numbers of dependent drinkers to 2010/11 for each local area. Summing these figures gives national estimates. The resulting totals suggest that for those years too, the equivalent of 19% of dependent drinkers in England had been in treatment at some time during the year, implying also that the equivalent of about a third of dependent drinkers who wanted to reduce their drinking had been in treatment.

Another way of estimating treatment-need put forward by the National Institute for Health and Care Excellence excluded low scorers on the dependence questionnaire, requiring scores (16 or more) indicative of at least moderate – not just mild – dependence. The result for 2014 would have been a much reduced figure for the dependent population of 257,626. The estimated treatment caseload with a primary alcohol problem (111,572 patients) is equivalent to 43% of this figure.

This variant in ways of estimating the number of drinkers who might warrant treatment is not the only one. On some grounds the population who might be considered in need of (even if only brief) treatment would be larger still, or those almost certainly in need of structured, care planned treatment, smaller still.

The argument has been made that the method which yielded for 2014 an estimate of 595,131 adult dependent drinkers best embodies NICE guidance on who should be offered treatment. If this is the case, treatment is engageing the equivalent of about 1 in 5 of those whom official guidance suggests are in need of treatment.

Based on what survey respondents living in private households are prepared to admit to, this estimate is almost certainly too optimistic. To the 595,131 adult dependent drinkers found by this method must be added dependent drinkers who are homeless (estimated at 19,328 in 2012) and an allowance for the tendency of survey respondents to underplay their drinking and related problems.

Last revised 05 February 2018. First uploaded 18 January 2018

Comment/query to editor
Give us your feedback on the site (one-minute survey)
Open Effectiveness Bank home page
Add your name to the mailing list to be alerted to new studies and other site updates


Top 10 most closely related documents on this site. For more try a subject or free text search

STUDY 2016 Adult substance misuse statistics from the National Drug Treatment Monitoring System (NDTMS) 1 April 2015 to 31 March 2016

STUDY 2014 Drug treatment in England 2013–14

STUDY 2013 Drug treatment in England 2012–13

STUDY 2015 Adult substance misuse statistics from the National Drug Treatment Monitoring System (NDTMS) 1 April 2014 to 31 March 2015

STUDY 2015 Specialist substance misuse treatment for young people in England 2013–14

DOCUMENT 2017 Young people’s statistics from the National Drug Treatment Monitoring System (NDTMS) 1 April 2015 to 31 March 2016

DOCUMENT 2011 Drug treatment and recovery in 2010–11

REVIEW 2017 An evidence review of the outcomes that can be expected of drug misuse treatment in England

HOT TOPIC 2017 Overdose deaths in the UK: crisis and response

DOCUMENT 2012 Substance misuse among young people 2011–12