Elsevier

The Lancet Psychiatry

Volume 4, Issue 9, September 2017, Pages 725-732
The Lancet Psychiatry

Personal View
Control conditions for randomised trials of behavioural interventions in psychiatry: a decision framework

https://doi.org/10.1016/S2215-0366(17)30153-0Get rights and content

Summary

In psychiatry, comparative analyses of therapeutic options and the aggregation of data from clinical trials across different therapeutic approaches play an important role in clinical decision making, treatment guidelines, and health policy. This approach assumes that trials of pharmacological and behavioural therapies generally produce the same level of evidence when properly designed. However, trial design for behavioural interventions has some unique characteristics and control groups vary widely, which influence the effects observed in any given trial. In this Personal View, we review various control conditions typically used in psychiatry, outline their effect on the internal validity and expected effect size of a trial, and propose a decision framework for choosing a control condition depending on the risk to the patient population and the stage of development of the therapeutic intervention. We argue that the choice of control group and its justification need to be taken into consideration when comparing behavioural and pharmacological therapies.

Introduction

At the core of modern medicine is the notion that all interventions used in clinical practice should be based on a high standard of evidence, a principle often referred to as evidence-based medicine. However, behavioural interventions used in psychiatry do not always have the necessary evidence base.1 Moreover, even if attempted, the implementation of evidence-based medicine standards for behavioural interventions faces specific challenges that are not easily overcome.2 The gold standard for evidence in medicine is generally judged to be the double-blind randomised controlled trial (RCT). Compared with drug trials, the double-blind, placebo-controlled design is not as easily applied to psychotherapy and other behavioural treatment strategies, particularly when it comes to control group design. In this Personal View, we propose a decision framework for the choice of control condition for such trials, taking into consideration the type of intervention, patient population, available resources, purpose of the trial, and trial phase.

Section snippets

Control conditions in RCTs assessing behavioural interventions

The control arm is an essential part of a RCT, intended to account for threats to internal validity resulting from the natural course of the disorder, response biases, and regression to the mean, among others. Several control conditions have been used in trials of behavioural interventions, including no treatment, minimal treatment, treatment as usual (TAU), and waitlist.3 For the purpose of this Personal View, we use definitions of control conditions modified from Mohr and colleagues (panel 1).

Decision framework

The selection of a control condition in a trial for a behavioural intervention will depend on several factors including the disease and its characteristics, the development phase, the purpose of the study, and the resources available. Thus, the choice of comparison condition is always a trade-off between the interests of the participating patients (eg, to have a high probability of receiving an effective intervention), the researchers performing the trial (eg, to protect participants from harm,

Conclusion and outlook

In this Personal View, we aimed to translate previous recommendations for trial design in psychotherapy research3 into a process of control group choice and justification. We derive from these considerations and recommendations the notion of participation risk and integrate this into a decision framework that includes consideration of trial stage and resource availability. Increasing awareness of the effect of control group choice in trials of behavioural interventions will hopefully help to

Search strategy and selection criteria

We searched PubMed using the following search strategy: (“autogenic training”[MeSH Terms] OR (“autogenic”[All Fields] AND “training”[All Fields]) OR “autogenic training”[All Fields]) OR (“behaviour therapy”[All Fields] OR “behavior therapy”[MeSH Terms] OR (“behavior”[All Fields] AND “therapy”[All Fields]) OR “behavior therapy”[All Fields]) OR (“cognitive therapy”[MeSH Terms] OR (“cognitive”[All Fields] AND “therapy”[All Fields]) OR “cognitive therapy”[All Fields]) OR “biofeedback

References (33)

  • Signal to noise: the trouble with psychiatry trials

    Lancet Psychiatry

    (2016)
  • DC Mohr et al.

    The selection and design of control conditions for randomized controlled trials of psychological interventions

    Psychother Psychosom

    (2009)
  • DC Mohr et al.

    Control condition design and implementation features in controlled trials: a meta-analysis of trials evaluating psychotherapy for depression

    Transl Behav Med

    (2014)
  • G Liegl et al.

    Guided self-help interventions for irritable bowel syndrome: a systematic review and meta-analysis

    Eur J Gastroenterol Hepatol

    (2015)
  • C Relton et al.

    Rethinking pragmatic randomised controlled trials: introducing the ‘cohort multiple randomised controlled trial’ design

    BMJ

    (2010)
  • TA Furukawa et al.

    Waiting list may be a nocebo condition in psychotherapy trials: a contribution from network meta-analysis

    Acta Psychiatr Scand

    (2014)
  • Cited by (133)

    View all citing articles on Scopus
    View full text