Goldberg L., Elliot D.L., MacKinnon D.P. et al.
Journal of Adolescent Health: 2007, 41, p. 421–429.
Unable to obtain a copy by clicking title? Try this alternative source.
First randomised follow-up study offers little support for randomly testing US school pupils for drug or alcohol use, adding to a slim evidence base which has so far found little benefit to justify the risks and the costs.
Summary This US study aimed to assess the effects of random drug and alcohol testing among high school athletes. US supreme court decisions explicitly permit random testing of pupils who participate in extracurricular activities but the legality of going beyond this to all school pupils remains in doubt. Methodology was a two-year prospective randomised controlled study of a single cohort across five intervention high schools with a random testing policy compared to six schools with a deferred policy, serially assessed by voluntary, confidential questionnaires. Athletes at schools with random testing policies were liable to be randomly urine-tested throughout the academic year. Positive test results were reported to parents or guardians and counselling of pupils was mandatory. Indices of illicit drug use, with or without alcohol use, were assessed at the beginning and end of each school year for the past month and past year. These showed that student athletes from intervention and control schools did not differ in past-month use of illicit drugs or a combination of drug and alcohol use at any of the four follow-up periods. At the end of the initial school year and after two full school years, student athletes at random testing schools reported less drug use during the past year compared to athletes at deferred policy schools. Combining past-year drug and alcohol use, student athletes at random testing schools also reported less use at the second and third follow-up assessments. Paradoxically, across all assessments athletes at random testing schools reported less athletic competence, less belief that authorities were opposed to drug use, and indicated greater risk-taking. At the final assessment, athletes at random testing schools believed less in the benefits of testing and less that testing was a reason not to use drugs. The authors concluded that no deterrent effects of random testing were evident for past-month use at any of the four follow-up periods, but self-reports indicated that past-year drug use was reduced at two of the four follow-ups, and at two a combination of drug and alcohol use was also reduced. Overall, drug testing was accompanied by an increase in some risk factors for future substance use. More research is needed before random drug and alcohol testing is considered an effective deterrent for school-based athletes.
commentary The featured study is the first to randomly allocate schools to testing versus no testing and then to follow up the pupils to test the results, making it the most important contributor to a very slim evidence base. To participate in the study, schools had to agree to randomly test pupils involved in extracurricular athletics as provided for in US law. They were then randomly allocated to initiate the policy or to wait until after the study was completed. Confidential pupil surveys suggested that the frequency of illegal drug use had been slightly curbed by the testing. However, trends in none of the other three indices of drug/alcohol use were significantly different across the two sets of schools, including indices of more recent (past month) use, the measures most likely to have been affected by testing. At best the results were inconclusive about impacts on substance use and if anything negative in terms of the students' attitudes to risktaking. Possibly too the relatively lenient response to positive tests led pupils in testing schools to weaken in their beliefs that the authorities were opposed to drug use.
For several reasons the study will not settle the issue of testing. Apart from the mixed outcomes, the fact that seven of the original 18 schools could not complete the study weakens confidence Also the test of whether the two sets of schools differed relied on differences in the trends in survey results from the first assessment through to the last. However, the first assessment was not an uncontaminated baseline. It took place after pupils knew whether they were to be tested so could have been influenced by this knowledge, though there is little indication that this was actually the case. in the findings and it was able to test only a non-punitive Parents were informed of positive tests and pupils referred for counselling. As long as this was accepted there were no further sanctions. If counselling was refused, pupils remained in school but could not participate in extracurricular sports. model; a more severe policy might have had greater deterrent effect but (see below) would also have posed greater risks. On the other hand, no account was taken of the cost of implementing testing, in this case paid for by the research project.
Earlier research included a pilot for the featured study involving just two schools which on some measures found reduced substance use in the school with testing but also a deterioration in attitudes to drugs, testing and school. Also, a survey comparing over 700 US schools with and without a testing policy was consistent with there having been no impact on cannabis use. Echoing the featured study's findings on attitudes, another US survey suggested that severe school drug policies may (along with other harsh disciplinary policies) diminish the degree to which pupils feel affiliated with the school, potentially one of the most important safeguards against unhealthy development, including substance use.
In contrast with the USA, Britain has merely flirted with the idea of testing school pupils for drugs. It was tried in at least two schools and was recently being considered by several others, though a planned large scale trial in Kent fell through when schools were unwilling to divert funds from other activities. Police sniffer dogs are an alternative also tried in the UK. An evaluation commissioned by Bedfordshire police concluded that the costs and the risks (among others, of alienating pupils Many felt the aim was to 'catch them out' and that they should not be made to parade past sniffer dogs. and publicly and potentially falsely stigmatising individuals) were balanced by little in the way of benefits. Pupils in the school where the dogs were used actually became less likely to believe that the experience would deter youngsters from having drugs inside school.
Official guidance for England published in 2004 did not explicitly rule out testing or sniffer dogs but did advise "extreme caution" and raised serious concerns such as whether such measures are consistent with a school's pastoral responsibilities. None of the recent UK national policies (for England, Scotland, Wales and Northern Ireland) mentioned these measures, a sign perhaps that initial governmental interest has receded. If so this would be in line with expert opinion from the government's drug advisers which recommended against sniffer dogs and testing in schools. In the virtual absence of research, their concerns were over ethics, practicality, cost, and the potential impact on relationships with pupils. Overall the slim evidence to date and these other concerns give no reason to subject pupils to drug testing or examination by sniffer dogs at random or without cause.
The same research team has developed prevention programmes for male (ATLAS) and female (ATHENA) students involved in sports activities to be implemented by coaches, aiming to provide healthy sports nutrition and strength-training alternatives to alcohol, illicit and performance-enhancing drugs.
Thanks for their comments on this entry in draft to Linn Goldberg of the Oregon Health & Science University. Commentators bear no responsibility for the text including the interpretations and any remaining errors.
Last revised 15 January 2009
DOCUMENT 2012 The government's alcohol strategy
REVIEW 2015 Prevention of addictive behaviours