UN Office on Drugs and Crime has announced that drug-treatment courts work

http://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6T1B-4W1MMK7-7&_user=861681&_rdoc=1&_fmt=&_orig=search&_sort=d&view=c&_acct=C000046147&_version=1&_urlVersion=0&_userid=861681&md5=3a0a6250b0f6154171a9ec4e79bd40f9
The LancetVolume 373, Issue 967111 April 2009-17 April 2009, Pages 1231-1233
Serial offending: evaluation of drugs courts
Sheila M Bird and Elizabeth LC Merrall
The UN Office on Drugs and Crime has announced that drug-treatment courts work.1 Evaluations from three continents consistently showed that drugs courts effectively reduce recidivism and underlying addiction problems of drug-abusing offenders, but by how much is not revealed. Instead, testimonials to success run along such lines as: "Only 11·6% of those who complete the drug court programme run into trouble again with the law".
The Ministry of Justice wants to evaluate four new drugs courts in England. Four have been established already without formal experiment: in Glasgow and Fife,2 Leeds,3 and west London. Process and cost evaluations, for want of power and a proper comparator, failed to quantify effectiveness at reducing recidivism[4] and [5] or mortality[5] and [6] in drug-dependent offenders.
Drugs courts internationally and their clients vary, but there are common principles. Justices receive training about drug dependency. Clients' progress is reviewed regularly by the sentencing judge. Dialogue between the justice and client results in praise, sanctions (from curfew to incarceration), or additional support. The supervising drugs and probation teams provide prereview reports on the client's progress. Drugs tests are mandatory at a frequency determined by the court and are reported at each review. All legal matters about the client, such as outstanding fines or new charges, should be dealt with by the drugs court. The clients' recovery from drug dependency is supported by encouragement into other activities (family, exercise, education, training, or employment) and practical back-up if relapse threatens.
These are fine principles, but do they deliver enough to warrant extra organisational and judicial costs? A court officer would need to be dedicated to schedule reviews, marshall reports, keep case-files up-to-date, and text reminders. Otherwise continuity of review by the sentencing judge is more honoured in the breach.3 Drugs-rehabilitation orders5 are, of course, made by other courts too. Their judges may also receive training in drug dependency, so that a drug court's distinguishing principles reduce to the second and final two of those listed above.
A systematic review of how drugs courts affect recidivism4 and a half century of randomised experiments in crime and justice with recidivism as an outcome7 had five randomised studies in common. Two studies had fatally flawed randomisation, and other familiar problems were poorly described control groups (justice as usual),8 exclusions from analysis, and lack of statistical power for modest effects on recidivism.[5] and [9] We might avoid serial offending by learning their lessons, and organising a properly randomised, sufficiently powerful, and efficiently followed-up trial of the effectiveness of drugs courts.
England's four new drugs courts could be rigorously evaluated for efficacy in a randomised trial. Eligible participants would be: adults who tested positive for cocaine or opioids on arrest for acquisitive crime; were charged and pleaded guilty before assignment to court; resided within the courts' jurisdiction; and were willing for drugs rehabilitation. They would also need to consent to randomisation between drugs court and conventional court. The randomisation ratio, within each of the four courts' areas, would be determined by the drugs court's capacity, so that the ratio might be 1:2 if the drugs court can only cope with one-third of eligible offenders. Participants cannot be re-randomised within 3 years. Biological or self-report assessments (eg, on injection of illicit heroin or acquisitive crimes) are not essential data, and so participants would be invited to opt-in to their provision. Randomisation is then stratified by the client's consent (or not) for minimum biological or self-report assessments that are common to both groups8 at, say, 4, 13, 26, and 52 weeks after randomisation. Cost and process variables can be collected on a random subsample of clients.
The court officer who usually assigns offenders to court should telephone the randomisation centre5 to register each client's eligibility, consent, and identity. Eligible offenders would then be randomised to a court assignment. Thereafter, follow-up is mainly by database linkage10 to count offences on the Police National Computer, incarceration days, duration of drug treatment (according to the National Treatment Agency), and deaths within 2 years of randomisation. From court files, sentence date11 and disposal (and by whom) could be captured.
The criminal-justice system should register its randomised evaluations of sentencing and offender management, in the manner now required as good practice in health-care evaluations. Besides governmental departments, research councils with health and social science responsibilities should aid and abet the collection of robust evidence on how the courts manage addictions and criminality.
Our suggested randomised design to evaluate drugs courts has limitations. Offenders who enter no or a delayed plea (about 20%) are not eligible. How judges are chosen for the drugs court also matters when extrapolating from the randomised evidence-base: fewer, and perhaps more senior, judges may preside in a drugs court than in conventional courts, where either a judge or magistrates sit.
Medicines of proven efficacy are only provided to patients in the UK's National Health Service if the drugs are cost effective. UK justice needs equivalent appraisal in its use of public funds. Speedier and more effective sentencing in drugs courts-in terms of reduced recidivism-might offset greater organisational and judicial costs. Additionally, recovery from drug dependency may save injectors' lives or reduce claims for welfare benefits. The a-priori case for affordability, and hence for evaluation, would need to be determined and be explicit in the study protocol.
Finding out about effectiveness for recidivism would mean randomising 700-900 offenders to have 80% power to discern a reduction in 2-year recidivism from 70% to 60%, and ten times as many to discern even a one-third reduction in 2-year mortality from 3% to 2%. Assuming only 180 eligible randomised clients in each of the four jurisdictions where drugs courts are planned, one-third of them assigned to the drugs court, should give answers on 2-year recidivism well within 4 years; answers on mortality would take much longer.
Drugs-court evaluations need the discipline of a well-written protocol. Ministers cannot duck the mathematics of numbers needed to neutralise the play of chance. Criminal justice should stop playing at evaluation, and recognise evidential rigour.12
SMB and ELCM are funded by Medical Research Council programme WBS number U.1052.00.002.00001.01. SMB chairs the Surveys, Design and Statistics Subcommittee of the Home Office's Scientific Advisory Committee, chaired the Royal Statistical Society's Working Party on Performance Monitoring in the Public Services (which called for greater attention to formal experiments, including randomisation), and served on the Appraisal Committee of the National Institute for Health and Clinical Excellence (1999-2005). ELCM declares that she has no conflicts of interest.
References
1 United Nations Office on Drugs and Crime, Drug treatment courts work! http://www.unodc.org/pdf/drug_treatment_courts_flyer.pdf (accessed Nov 18, 2008)..
2 G McIvor, L Barnsdale, M Malloch, S Eley and R Yates, The Operation and Effectiveness of the Scottish Drugs Court Pilots http://www.scotland.gov.uk/Resource/Doc/100021/0024203.pdf (2006) (accessed Nov 18, 2008)..
3 Ministry of Justice, Dedicated drug court pilots: a process report http://www.justice.gov.uk/docs/dedicated-drug-courts.pdf (2008) (accessed Nov 18, 2008)..
4 DB Wilson, O Mitchell and DL MacKenzie, A systematic review of drug court effects on recidivism, J Exp Criminol 2 (2007), pp. 459-487. Full Text via CrossRef
5 SM Bird, Prescribing sentence: time for evidence-based justice, Lancet 364 (2004), pp. 1457-1459. Article PDF (131 K)
6 R King, SM Bird, SP Brooks, SJ Hutchinson and G Hay, Prior information in behavioural capture-recapture methods: demographic influences on drug injectors' propensity to be listed in data sources and their drug-related mortality, Am J Epidemiol 162 (2005), pp. 694-703. Full Text via CrossRef View Record in Scopus Cited By in Scopus (1)
7 DP Farrington and BC Welsh, A half century of randomized experiments on crime and justice, Crime Justice 34 (2006), pp. 55-132. View Record in Scopus Cited By in Scopus (9)
8 R Perera, C Heneghan and P Yudkin, Graphical method for depicting randomized trials of complex interventions, BMJ 334 (2007), pp. 127-129. Full Text via CrossRef View Record in Scopus Cited By in Scopus (10)
9 EP Deschenes, S Turner and PW Greenwood, Drug court or probation: an experimental evaluation of Maricopa County Drug Court, Justice System J 18 (1995), pp. 55-73.
10 Surveys, Design and Statistics Subcommittee of Home Office's Scientific Advisory Committee, 21st century drugs and statistical science http://www.homeoffice.gov.uk/documents/science-advisory-committee/21st-century-drugs-stats?view=Binary (Dec 17, 2008) (accessed Dec 17, 2008)..
11 Ministry of Justice, Time intervals for criminal proceedings in magistrates' courts http://www.justice.gov.uk/docs/time-intervals-criminal-proceedings-june-2008.pdf (Aug 21, 2008) (accessed Nov 18, 2008)..
12 J Shapland, A Atkinson and H Atkinson et al., Does restorative justice affect reconviction? The fourth report from the evaluation of three schemes http://www.justice.gov.uk/restorative-justice-report_06-08.pdf (June 2008) (accessed Nov 18, 2008)..