Call for randomised criminal-justice trials


NB. Please note that if you are outside North America, the embargo for LANCET press material is 0001 hours UK Time 15 October 2004.

A scientific approach to test the validity of criminal-justice interventions before they become implemented is proposed by the author of a Viewpoint in this week's issue of THE LANCET. The article provides details of how court-based randomisation could be put into practice.

Sheila Bird (UK Medical Research Council/University of Strathclyde, Scotland) focuses on the lack of true scientific evidence that has been used as a basis for implementing UK criminal justice policy on drug-dependent offenders. She outlines the ideal scientific approach for the criminal justice system: "Formal experiments, including the randomisation of sufficient numbers of offenders, are needed to inform judicial approval or "licensing" of new criminal-justice disposals before they are implemented nationally.

'Like doctors, judges should have access to studies comparing salient outcomes in large numbers of eligible drug-dependent offenders who have been randomised between drug treatment and testing order and alternative sentence. Judges have been criticised for unacceptable sentence variation, yet they are denied a rigorous evidence base to inform judicial decisions. Basing the management of offenders on inferior study designs that are non-randomised and underpowered denies them and us evidence about the efficacy and cost-effectiveness of new criminal-justice interventions. For example, punitive random mandatory drugs testing of prisoners has not been demonstrated to reduce inmates' use of heroin but had cost over UK£100 million when the UK's home affairs select committee called for its review in 2000."

Professor Bird adds: "Failure to randomise from the outset, or after pilot studies, means that, more than £100 million and 21 000 drug treatment and testing orders later, UK criminal justice still does not know whether such orders are safe, effective, and cost effective. We know their cost, but not their value. In 2004-05, less costly and intensive orders will be introduced, again without formal experimentation. It is time to end the masquerading of qualitative, implementation studies as outcome evaluations".

Potential solutions to the status quo include the creation of databases for drug-dependent offenders and the randomisation of offenders to a particular programme (eg, drug rehabilitation versus prison sentence) in the court room.

Professor Bird concludes: "We still lack answers to major questions: do such orders prevent drugs-related deaths, do they reduce incarceration days, reconviction rate, or unemployment days? The proposed court-based randomisation, designed to capture prescribed drug treatment and testing order and alternative sentence by court location, injector history, and crime type, would improve vastly on the current indiscipline of data…Because criminal justice deals with damaged as well as damaging young lives, it affects the public and offenders' health. Ignorance also damages wealth: the implementation of criminal-justice initiatives of uncertain efficacy and cost-effectiveness may be at the expense of fewer medicines-of known efficacy and cost-effectiveness-being prescribed by the UK National Health Service."

Source: Eurekalert & others

Last reviewed: By John M. Grohol, Psy.D. on 21 Feb 2009
    Published on All rights reserved.