Statistical power in randomized intervention studies with noncompliance

This study examined various factors that affect statistical power in randomized intervention studies with noncompliance. On the basis of Monte Carlo simulations, this study demonstrates how statistical power changes depending on compliance rate, study design, outcome distributions, and covariate information. It also examines how these factors influence power in different methods of estimating intervention effects. Intent-to-treat analysis and complier average causal effect estimation are compared as 2 alternative ways of estimating intervention effects under noncompliance. The results of this investigation provide practical implications in designing and evaluating intervention studies taking into account noncompliance.     

Rational noncompliance with prescribed medical treatment

Despite the attention that patient noncompliance has received from medical researchers, patient noncompliance remains poorly understood and difficult to alter. With a better theory of patient noncompliance, both greater success in achieving compliance and greater respect for patient decision making are likely. The theory presented, which uses a microeconomic approach, bridges a gap in the extant literature that has so far ignored the contributions of this classic perspective on decision making involving the tradeoff of costs and benefits. The model also generates a surprising conclusion: that patients are typically acting rationally when they refuse to comply with certain treatments. However, compliance is predicted to rise with increased benefits and reduced costs. The prediction that noncompliance is rational is especially true in chronic conditions at the point that treatment begins to move closer to the medically ideal treatment level. Although the details of this theory have not been tested empirically, it is well supported by existing prospective and retrospective studies.     

Placebo effect and randomized clinical trials

The achievement of optimal therapeutic results presupposes the use of appropriate treatment combined with maximal utilization of placebo effects. These aims may sometimes be difficult to satisfy in randomized clinical trials (RCTs). The question thus arises whether there is a conflict between the goals of therapy and those of experimental research; and if so, to what extent, and how is it handled in practice by clinicians and researchers. Various ethical problems have been discussed in several reports connected with RCTs. But we have found no discussion concerning the conflict between obtaining informed consent and promoting optimal placebo effects. Information about RCTs can be given in various ways. Sometimes appropriate information about RCTs to patients involves non-optimal utilization of placebo effects. This gives rise to ethical and methodological problems, which are discussed in this article.     

Design of randomized controlled trials: Principles and methods applied to a treatment for early stuttering

The “gold standard” methodology for treatment efficacy research is the randomized controlled trial (RCT), which is used extensively in medical research and in other areas such as psychology. Results from a well designed and conducted RCT, which show a new treatment to be clinically and statistically superior to current standard practice, can lead to a change in clinical practice. This paper presents a tutorial on RCTs, presenting and discussing the following principles and methods: the trial protocol, choice of control treatment, eligibility criteria, random allocation, outcomes and hypotheses, sample size, subject recruitment, analysis by intention to treat, interim analysis, stopping rules, safety data monitoring/trial management, and study documentation. RCTs are a complicated and logistically involved methodology. Hence, prior to the commitment of resources to such a trial in the development of a new treatment, the safety and clinical promise of the new treatment needs to be demonstrated with preliminary studies. Such preliminary studies have been completed for the Lidcombe Program (LP) of early stuttering intervention, and an RCT of that treatment is being conducted at the time of writing. The principles and methods of RCTs are illustrated with reference to that study.

Educational objectives: The reader will understand the design principles and methods of RCTs. The reader will understand the logistics of conducting a RCT of a treatment for early stuttering.     

Lifestyle interventions in the treatment of childhood overweight: a meta-analytic review of randomized controlled trials.

CONTEXT: Evaluating the efficacy of pediatric weight loss treatments is critical. OBJECTIVE: This is the first meta-analysis of the efficacy of RCTs comparing pediatric lifestyle interventions to no-treatment or information/education-only controls. DATA SOURCES: Medline, PsycINFO, and Cochrane Controlled Trials Register. STUDY SELECTION: Fourteen RCTs targetting change in weight status were eligible, yielding 19 effect sizes. DATA EXTRACTION: Standardized coding was used to extract information on design, participant characteristics, interventions, and results. DATA SYNTHESIS: For trials with no-treatment controls, the mean effect size was 0.75 (k = 9, 95% confidence interval [CI] = 0.52-0.98) at end of treatment and 0.60 (k = 4, CI = 0.27-0.94) at follow-up. For trials with information/education-only controls, the mean ES was 0.48 (k = 4, CI = 0.13-0.82) at end of treatment and 0.91 (k = 2, CI = 0.32-1.50) at follow-up. No moderator effects were identified. CONCLUSIONS: Lifestyle interventions for pediatric overweight are efficacious in the short term with some evidence for extended persistence. Future research is required to identify moderators and mediators and to determine the optimal length and intensity of treatment required to produce enduring changes in weight status.     

Statistical power and optimal design for multisite randomized trials

The multisite trial, widely used in mental health research and education, enables experimenters to assess the average impact of a treatment across sites, the variance of treatment impact across sites, and the moderating effect of site characteristics on treatment efficacy. Key design decisions include the sample size per site and the number of sites. To consider power implications, this article proposes a standardized hierarchical linear model and uses rules of thumb similar to those proposed by J. Cohen (1988) for small, medium, and large effect sizes and for small, medium, and large treatment-by-site variance. Optimal allocation of resources within and between sites as a function of variance components and costs at each level are also considered. The approach generalizes to quasiexperiments with a similar structure. These ideas are illustrated with newly developed software.     

Deception by young children following noncompliance

A paradigm devised by M. Lewis, C. Stanger, and M. W. Sullivan (1989) was adapted to study deception and false-belief understanding. In Study 1, 3- and 5-year-olds were asked not to touch a toy in the experimenter's absence. Just over half of the children touched the toy, and of those children, the majority denied having done so. Of control children who were given permission to touch the toy, all touched it and admitted having done so. In Study 2, 3- and 5-year-olds were asked not to look in a box to identify its contents. Almost all children looked, most denied having looked, and a minority consistently feigned ignorance of the contents. False-belief understanding was linked to denial of looking but not to feigning ignorance. Of control children who were given permission to look, all acknowledged looking, and they almost always revealed their knowledge of the contents. The studies confirm that preschoolers deceive in the context of a minor misdemeanor but are less effective at feigning ignorance.     

The efficacy of systemic therapy with adult patients: a meta-content analysis of 38 randomized controlled trials

Systemic therapy is a widely used psychotherapy approach. Yet there exist few systematic reviews on its efficacy. A meta-content analysis was performed to analyze the efficacy of systemic therapy for the treatment of mental disorders in adulthood. All randomized (or matched) controlled trials (RCT) evaluating systemic/systems oriented therapy in various settings (family, couple, individual, group, multifamily group therapy) with adult index patients suffering from mental disorders were identified by database searches and cross-references in other reviews. Inclusion criteria were: index patient diagnosed with a DSM or ICD listed mental disorder, trial published in any language up to the end of 2008. The RCTs were content analyzed according to their research methodology, interventions applied, and results. Thirty-eight trials published in English, German, Spanish, and Chinese were identified, 34 of them showing systemic therapy to be efficacious for the treatment of mood disorders, eating disorders, substance use disorders, mental and social factors related to medical conditions and physical disorders, and schizophrenia. Systemic therapy may also be efficacious for anxiety disorders. Results were stable across follow-up periods of up to 5 years. There is a sound evidence-base for the efficacy of systemic therapy for adult index patients with mental disorders in at least five diagnostic groups.     

A randomized controlled trial of self-directed versus therapist-directed cognitive-behavioral therapy for obsessive-compulsive disorder patients with prior medication trials

Cognitive-behavioral therapy incorporating exposure and response prevention (ERP) is widely considered a first-line psychosocial treatment for patients with obsessive-compulsive disorder (OCD). However, a number of obstacles prevent many patients from receiving this treatment, and self-administered ERP may be a useful alternative or adjunct. Forty-one adult outpatients with a primary diagnosis of OCD, who reported at least 1 current or previous adequate medication trial, were randomly assigned to self-administered or therapist-administered ERP. Patients in both treatment conditions showed statistically and clinically significant symptom reduction. However, patients receiving therapist-administered ERP showed a superior response in terms of OCD symptoms and self-reported functional impairment. We discuss several potential reasons for the superiority of therapist-administered treatment, and propose a stepped-care integration of self-administered and therapist-administered interventions for OCD.     

A note on statistical power in multi‐site randomized trials with multiple treatments at each site

We derive the statistical power functions in multi‐site randomized trials with multiple treatments at each site, using multi‐level modelling. An F statistic is used to test multiple parameters in the multi‐level model instead of the Wald chi square test as suggested in the current literature. The F statistic is shown to be more conservative than the Wald statistic in testing any overall treatment effect among the multiple study conditions. In addition, we improvise an easy way to estimate the non‐centrality parameters for the means comparison t‐tests and the F test, using Helmert contrast coding in the multi‐level model. The variance of treatment means, which is difficult to fathom but necessary for power analysis, is decomposed into intuitive simple effect sizes in the contrast tests. The method is exemplified by a multi‐site evaluation study of the behavioural interventions for cannabis dependence.     

Inefficacy interim monitoring procedures in randomized clinical trials: the need to report

If definitive evidence concerning treatment effectiveness becomes available from an ongoing randomized clinical trial, then the trial could be stopped early, with the public release of results benefiting current and future patients. However, stopping an ongoing trial based on accruing outcome data requires methodological rigor to preserve validity of the trial conclusions. This has led to the use of formal interim monitoring procedures, which include inefficacy monitoring that will stop a trial early when the experimental treatment appears not to be working. For participants, inefficacy monitoring is especially important as it ensures that they are not being treated worse than if they had not enrolled on the trial. We discuss the importance of reporting with trial results the formal interim inefficacy monitoring guidelines that were utilized, and, if none were used, the reasons for their absence. A survey of two leading medical journals suggests that this is not current practice.