Send email for updates

About updates
Research 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. Free reprints may be available from the authors – click prepared e-mail. 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

The impact of compliance with a compulsory model of drug diversion on treatment engagement and reoffending.

McSweeney T., Hughes C., Ritter A.
Drugs: Education, Prevention and Policy: 2016, early online publication.
Unable to obtain a copy by clicking title? Try asking the author for a reprint by adapting this prepared e-mail or by writing to Dr McSweeney at

Further evidence from England that schemes which force people arrested for certain offences to be tested for heroin or cocaine use and if positive to be assessed for treatment do not pay back in terms of treatment engagement or crime reduction.

Summary In the attempt to facilitate engagement with treatment services, over the 10 years from 2003 over 1.7 million arrests in England and Wales resulted in mandatory testing of the arrestee for recent use of heroin or cocaine. Failure to provide a sample for testing or if the test is positive, to complete an assessment of substance use problems and the need for treatment, can result in further sanctions in the form of fines and imprisonment.

The ‘Tough Choices’ policy

‘Tough Choices’ was the term used to describe pre-sentence options introduced by the 2005 Drugs Act – testing on arrest and required assessments by a suitably qualified drugs worker.

Under the Act, a drug test is required to be administered in custody following the arrest of an adult for a ‘trigger’ offence (mainly linked to theft, drugs and fraud). Senior police officers may also authorise testing after any alleged offence suspected to have been contributed to by heroin or cocaine use. The test uses an oral fluid/saliva sample taken from the suspect to detect recent (up to 48 hours) use of the drugs.

It is an offence to fail to comply with testing without good cause, to fail to attend and remain at an initial assessment within 28 days of a positive test, or (if considered needed) to fail to attend a follow-up assessment. Assessments are intended to provide opportunities for recent heroin/cocaine users to voluntarily engage with treatment and other support.

Since April 2013 ‘Tough Choices’ and the broader drug interventions programme are no longer nationally managed, mandated and funded. Their continuation is now a matter for local discretion and funding through elected police and crime commissioners. There is some evidence of their continuation. For example, in 2012/2013, London’s police service extended test-on-arrest provisions across all of the capital’s boroughs.

Known as ‘Tough Choices’, these provisions ( panel) were integral to a wider drug interventions programme aiming to reduce drug-related crime by at every stage of the criminal justice process identifying criminally involved heroin and cocaine users and engaging and retaining them in appropriate treatment.

The featured study contributed to the scant evidence base for this programme by comparing matched samples of heroin/cocaine users identified through ‘Tough Choices’ testing who did versus did not comply with the requirement to attend for an assessment. At issue was whether in practice treatment records were consistent with assessment having played its key role of engaging drug users in treatment. If it had, among arrestees who tested positive records should reveal an association between completion of a required assessment and engagement with structured treatment for drug problems. If that treatment was effective, records should also reveal reductions in subsequent offending. Also investigated by the study were the factors predictive of subsequent offending. It was the first such study to use a matched comparison group.

In one English police force area over a 12-month period in 2007/2008, the study identified 1630 arrestees who had tested positive for recent use of heroin or cocaine. Of these, 836 complied with ‘Tough Choices’ requirements by attending an assessment within 28 days. The remainder were non-compliant, failing to meet this requirement. Using shared identifiers, arrestees’ records were matched with data on treatment commencement and discharges, and on arrests and cautions/convictions. The featured report assessed cautions/convictions (‘proven offending’) for offences committed in the 12 months after the day of the test (allowing a further six months for the offence to be processed), and compared these with caution/convictions up to five years back from the same day. Included were cautions/convictions for breaching requirements such as those imposed by the ‘Tough Choices’ provisions. Treatment commencement was considered to have been linked to the test if it occurred within 144 days – just under five months.

Key points icon

Key points
From summary and commentary

In England ‘Tough Choices’ schemes require people arrested for certain offences to be tested for heroin or cocaine use and if positive to be assessed for treatment. Failure to comply is an offence.

In one English police force area the most rigorous evaluation to date of such schemes found that attending the assessment neither led to greater engagement with treatment nor reduction in crime.

Together with other evidence from England and Scotland, the implication is that such schemes risk criminalising arrestees who do not comply, without generating countervailing benefits.

An attempt was made to even out differences between compliant and non-compliant groups which might have led them to attend versus not attend the assessment, a process intended to further expose attendance as the likely cause of any downstream differences in treatment engagement or offending. 656 individuals from each group could be acceptably matched based on the four variables most closely related to compliance. The matched samples were also very similar in the time during the follow-up 12 months when they were not in custody and at liberty to offend. Matched arrestees were typically white men in their late 20s and early 30s charged with theft. On average they had a record of nearly a dozen prior convictions. About 47% had tested positive for cocaine only, 39% for both cocaine and heroin, and about 14% for heroin only.

Main findings

Comparing the year before the test with the year after, overall about half the matched samples continued to be convicted or cautioned at the same (19%) or a greater (31%) rate. Neither in respect of proven offending nor treatment engagement was there any substantial or statistically significant indication that attending the required assessment had a positive impact; details below.

Of the compliant arrestees, 26% started treatment within 144 days and 38% within 12 months compared to 21% and 33% of matched arrestees who did not complete the required assessment chart. Of those who started treatment, 72% who completed the required assessment and 74% who did not were retained for 12 weeks, 25% versus 30% completed the treatment they started, and 42% each dropped out.

Indices of treatment engagement

61% of arrestees who completed the required assessment and 60% who did not were cautioned or convicted within 12 months of the test, in each case about three times. Compared to the 12 months before testing, there was a slight increase in the average frequency of proven offending among arrestees who completed the assessment and a slight decrease among the non-compliant group.

Once influences known to the researchers had been taken into account, the likelihood that an arrestee would be cautioned or convicted in the 12 months after testing was greatest if they had tested positive for both heroin and cocaine and least when just cocaine was found, greater if the alleged offence was theft rather than a drugs offence, and 44% greater among arrestees who started treatment within 144 days than those who did not. Whether an individual complied with the ‘Tough Choices’ policy by completing the required assessment was unrelated to the chances of their subsequently being proven to have offended.

The authors’ conclusions

No association was found between compliance with a compulsory system for diverting arrested heroin/cocaine users to assessment for treatment and the chances that they would start and stay in treatment or complete it. Nor was there any link with reductions in reoffending. These findings cast doubt on the effectiveness of this kind of sanctions-based approach to reducing ‘drug-related’ crime, and on the value-for-money of the £3.3 million the police force area in the study received to deliver these kinds of measures during the 12 months of the study.

Other studies in England have also found no significant benefits from referral to treatment by the criminal justice system, and the results are broadly consistent with those to emerge from an earlier national evaluation of the ‘Tough Choices’ policy, in that around half those identified as recent heroin or cocaine users showed on average a 70% reduction in the volume of their proven offending during the follow-up period, while the remainder continued at the same or an increased rate.

The featured study also yielded little evidence of the ‘swift and certain’ sanctions considered prerequisites for effective schemes to divert offenders to treatment or to monitor and sanction further drug use. Only one in seven of the heroin/cocaine users who failed to complete the required assessment were sanctioned as a consequence. On the other hand, there was compelling evidence of ‘net-widening’ – the risk that the attempt to divert offenders into treatment might instead draw them into the criminal justice system; one in six of the identified heroin/cocaine users had no proven offences in the 12 months prior to their identification, and 11% had no proven offences during the same period plus the 12 months after being tested.

Despite little evidence of benefit, efforts in the UK to secure engagement and behaviour change among drug misusers coming into contact with the criminal justice system continue to be based on punishing non-compliance rather than incentivising treatment. Non-compliance with such measures among ex-prisoners in England and Wales accounted for over 2500 recalls to custody during 2015, imposing a considerable burden on the justice system and on the offenders.

More fundamentally, diversion schemes are typically targeted at changing individual actions and behaviours, but have limited effect on broader situational and structural problems linked to relationships, accommodation, education, training and employment opportunities, and other forms of ‘recovery capital’ – yet the main drivers of both drug dependence and criminality are disproportionately associated with poverty and structural disadvantage.

Links found between proven offending and other factors suggest that if such schemes continue to be tried, they might profit from being better targeted at arrestees at heightened risk of offending: younger detainees, those using both heroin and cocaine, those arrested for acquisitive crime rather than drug offences only, and criminally-involved drug users with more extensive offending histories.

These conclusions derived, however, from a study with significant limitations. Despite the matching procedure, differences may have remained between compliant and non-compliant arrestees which affected the likelihood they would engage with treatment or offend. In particular, what proportion of those testing positive for cocaine were using it in the form of crack is unknown. Conducted in just one of 17 English police force areas, the extent to which the results are generalisable to other contexts and settings is clearly limited. Offending was assessed by cautions and convictions, which reflect not just offending but the degree to which those offences come to light and are prosecuted.

Findings logo commentary The ‘no effect’ findings of this study are the more convincing because it could be expected that arrestees compliant with ‘Tough Choices’ would also comply with referral to treatment. The obvious and probable implication is that the legal and enforcement apparatus which required assessment of positive-test arrestees was a waste of resources which risked collateral damage by creating offenders out of non-compliant arrestees who at this stage were simply suspects yet to be charged with an offence. We know from an earlier study that many would not be charged, and inevitably fewer still would be found guilty. The only offence they may be shown to have committed could have been failure to attend the assessment.

However, it is also possible that despite the matching process, arrestees who chose to attend the assessment did so because on average they were in greater need of help due to more severe drug and drug-related problems. In this case the assessment process may have helped prevent worse outcomes in this group than in the non-compliant group. The similarity between the two sets of arrestees and in their subsequent treatment engagement and offending makes this explanation unlikely. Non-impact of attending the required assessment might also have partly been because many arrestees who would have been prompted to start treatment were already being treated. Information gathered during the assessment of the compliant set indicated that about 40% were in some form of treatment at the time. If this was the case, it would still mean that the procedure was ineffective, but that it might have been effective in a context where treatment was less accessible.

A finding which requires comment is that the likelihood that an arrestee would be cautioned or convicted in the 12 months after testing was 44% greater among arrestees who started treatment within 144 days than those who did not. For treatment the most favourable interpretation is that choosing to start treatment was a marker of the severity of the arrestee’s substance use and related problems – that perhaps prompted by their arrest and realising they were in need of help, the more severely affected heroin/cocaine users sought or accepted formal treatment.

An early evaluation of the ‘Tough Choices’ policy in England also questioned whether testing arrestees led to more treatment engagement than would have happened in any event, and found considerable potential for criminalisation due to the process since about 40% of those identified through on-arrest testing had no recent convictions. The evaluation was primarily about comparing the previous testing-on-charge policy with ‘Tough Choices’, which moved testing forward to the pre-charge arrest stage. On-charge testing netted an average 1932 positive tests per month. Testing on arrest raised this to 3672, suggesting that many who never get charged were now being required to undergo test and assessment. 62% of the on-arrest cohort were classified as ‘low crime causing’ compared to 49% who had tested positive after being charged.

In 2009 an evaluation of the corresponding policy in Scotland led the Scottish Executive to terminate funding of mandatory testing schemes, though continuing with voluntary schemes. For two years from June 2007 mandatory testing and assessment similar to the English ‘Tough Choices’ scheme was piloted at three police stations known to have high levels of drug use among arrestees. On average, at each of the schemes just two to three people a month entered treatment who might not otherwise have done so. In terms of the cost per person who started treatment, in two of the three areas the mandatory scheme was less cost-effective by a factor of from six to eight than similar schemes which left it up to the arrestee whether to choose to seek assessment and treatment.

Latest UK evaluation

The lead author of the featured study was also responsible for what seems the latest UK evaluation of a test-on-arrest scheme, comparing it with conventional voluntary arrest referral in the county of Hertfordshire just to the north of London. The results led the county’s police force to decide against extending the ‘Tough Choices’ policy across the force’s area.

Test-on-arrest was piloted in one police station, identifying 219 users of cocaine and opiates like heroin in the 12 months from 1 December 2012. Over the same period, in three other stations 81 arrestees were identified as drug misusers via conventional ‘cell sweeps’ by arrest referral workers. During the follow-up 12 months, indicators of re-offending were roughly the same in both sets of arrestees, though compared to the year before their identification, at 35% versus 28%, the reduction in the number of offences leading to a charge was greater after arrest referral.

Though broadly comparable, one dimension on which the two sets of arrestees differed was their engagement in treatment at the time of identification – 27% of the test-on-arrest sample but none of those identified by arrest referral workers. This quickly evened out; typically within a month, 25% of the arrest referral sample had started treatment compared to just 7% of the test-on-arrest sample. Among those engaged in structured treatment, routinely completed treatment monitoring forms revealed that the reduction in the frequency of illicit drug use was twice as great among the patients identified through arrest referral. However, by the end of the follow-up period roughly the same proportion of arrestees were reportedly using illicit opiates and/or cocaine (67% of the test-on-arrest sample versus 55% of the arrest referral sample), and at 10–11 days a month, their average frequencies of use were virtually identical. As in the featured study, detected re-offending was much more likely if the original offence had been theft rather than solely a drug office.

Though for those identified the results were similar, test-on-arrest did however identify many more drug users than arrest referral. The researchers concluded, “While the results from this review offer little justification for extending test on arrest and required assessment arrangements across the county, it is important to note that testing appeared to be an effective mechanism for identifying Class A drug users ... the number of arrestees tested on arrest during the first year of the pilot exceeded the number of detainees who met existing criteria for targeted testing in the custody suite which trialled these measures. By contrast, conventional cell sweeps ... appeared to identify only seven per cent of eligible arrestees.”

Make the most first of voluntary schemes

Evidence from England and Scotland raises doubts over whether making testing and assessment mandatory is a cost-effective alternative to voluntary arrest referral schemes in which drug workers approach arrestees and if appropriate offer referral to treatment. Of concern are the costs of the mandatory schemes and the potential for them to generate offences and convictions for non-compliance – though related to this, testing on arrest does seems to identify more drug users who might be in need of treatment than either arrest-referral or testing on charge.

Questionable too are the ethics of forcing people into assessment and re-assessment. It may be argued that this amounts not just to compulsory diagnosis but also compulsory treatment, since effective assessment acts not just as a preparation for treatment, but as its start. From this perspective people who refuse assessment risk being criminalised for rejecting the start of a clinical process, not normally considered a criminal act. Set against this is the possibility of drawing people into treatment who will benefit from it with consequential benefits for society, but who would not otherwise have owned up to heroin or cocaine use or accepted an in-depth assessment. Studies to date in Britain suggest that in practice these benefits are largely ethereal. Whatever the balance of benefits and risks, it would seem financially and ethically prudent to maximise the reach of voluntary schemes before resorting to compulsion.

Thanks for their comments on this entry in draft to research author Tim McSweeney of Middlesex University in England. Commentators bear no responsibility for the text including the interpretations and any remaining errors.

Last revised 26 March 2017. First uploaded 16 March 2017

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 2013 Criminal justice responses to drug related crime in Scotland

STUDY 2003 Arrest referral tackles drug-driven crime

STUDY 2008 Testing on arrest scatter gun nets some extra treatment entrants

STUDY 2009 The Drug Treatment Outcomes Research Study (DTORS): final outcomes report

STUDY 2009 Evaluation of the mandatory drug testing of arrestees pilot

STUDY 2012 Adaptive programming improves outcomes in drug court: an experimental trial

STUDY 2012 Summary of findings from two evaluations of Home Office alcohol arrest referral pilot schemes

STUDY 2008 Treatment on bail makes little discernable difference

STUDY 2011 Interim methadone treatment compared to standard methadone treatment: 4-month findings

STUDY 2014 Drugs: international comparators