• By Kevin Mount
  • Posted on Thursday 15th May, 2008

A dark art comes to the water-cooler

Randomized controlled trials (RCTs) are nothing if not controversial in the UK. Once a feature of the distant landscape of medical science, they have become the topic of water cooler conversations in children’s services almost overnight. Policy makers once castigated for not sanctioning their use in high-profile evaluations are now sponsors of several UK-wide studies that involve RCTs. Practitioners in youth justice, family support, health and education can escape the dark art no longer.For some, RCTs continue to be anathema. Yet if the more evangelical are sometimes unhelpful in their zeal, so detractors’ concerns are seemingly often the product of ignorance. It’s right-on to be against RCTs, even if you’re not quite sure what an RCT is. Some of the antagonism is undoubtedly due to misinformation, but resistance also stems from experience of trials done badly. Indeed, the historical lack of policy interest in the UK means that there is an urgent need to build research capacity in the field.Edited by Arthur and Maguth Nezu from Drexel University, Philadelphia, this authoritative book, with contributions from successful investigators familiar with RCTs, should be high on any eager learner’s reading list. It is aimed at researchers, new or seasoned, and seeks to provide practical guidance on conducting psychosocial outcome investigations. Chapters are organized in the approximate order in which a researcher would devise and undertake an RCT, moving from concept through sampling and data collection to analysis and writing-up. The editors are at pains to stress that RCTs cannot answer all of the questions that pertain to providing the best services to users, and this tone of critical self-reflection, accompanied by in-depth knowledge and extensive hands-on experience, permeates the volume. Massachusetts research scientist Kimberlee Trudeau and colleagues describe the internationally recognized gold standard for reporting RCTs in medicine known as the CONSORT guidelines . These require that trials be described in terms of 22 items, including eligibility criteria for the service, the randomization procedure, what the service comprised and the primary and secondary outcomes measures used. Trudeau demonstrates the value of these headings using a case study of an RCT of a group anger intervention for patients with hypertension. Significantly, they argue that the guidelines are equally helpful in the design and implementation phases of a study. A later chapter by Fiona Fidler and colleagues from La Trobe University, Australia, offers a nice complement in the form of a worked example of how to analyze and write up the results of an RCT.The perennial problem of research participant recruitment and retention is discussed by Joan Grant and colleagues from the University of Alabama. It notes that RCTs often extend beyond the funding period owing to problems obtaining the required sample sizes, and offers advice on preventing this problem. Thus, research teams should forge good relationships with key people at study sites and develop the skill of writing accessible and visually appealing recruitment materials (guidelines are provided). What makes the advice more than a list of good ideas is the research underpinning it, for example on the value (or otherwise) of offering participants cash and other incentives.A common complaint about RCTs is that they tell you if a service works but not why or in what conditions. Matthew Nock and colleagues from the Harvard School of Psychology concede the point but show how slightly modifying the trial makes it possible to test the “moderators and mechanisms” of clinical change. One of their examples shows how interventions for children from divorced families prevented internalizing and externalizing behavior problems in adolescents but achieved greatest impact among those whose problems were more severe initially. In another, methods to test mediating mechanisms found that MST (Multisystemic Therapy) reduced delinquent behavior by improving family relations and reducing delinquent peer affiliations.Ethical issues are dealt with by Michele Galietta and Barbara Stanley from City University, New York. Much of the resistance to RCTs has its roots in the concern that potentially useful interventions are withheld from needy groups, although, as the authors here point out, experiments in nazi Germany and elsewhere have left a legacy of concern about consent and the risks to participants of testing unproven treatments. As well as offering helpful practical advice on how to practice sound ethical research, their chapter should offer some comfort to unbelievers insofar as "regulatory requirements are increasing, and the burdens on researchers are increasing as well". It is getting harder to conduct unethical trials.Because it has its roots in research on psychotherapy, the book and the examples it offers contain numerous references to “patients” and “treatment”. However, this does not detract from the value of the overall messages. If it is widely read and acted upon, Nezu and Nezu’s collection will unquestionably improve the quality and acceptance of RCTs in children’s services.• Evidence-based Outcome Research: A practical guide to conducting randomized controlled trials for psychosocial interventions, by Nezu, A. M. and Nezu, C. M. (Eds) (2008) is published by Oxford University Press, New York.

Back to Archives