Does Inclusion of a Placebo Arm Influence Response to Active Antidepressant Treatment in Randomized Controlled Trials? Results From Pooled and Meta-Analyses

Department of Psychiatry, University of Toronto, Ontario, Canada.
The Journal of Clinical Psychiatry (Impact Factor: 5.5). 03/2010; 71(3):270-9. DOI: 10.4088/JCP.08r04516blu
Source: PubMed


To determine if the inclusion of a placebo arm and/or the number of active comparators in antidepressant trials influences the response rates of the active medication and/or placebo.
Searches of MEDLINE, PsycINFO, and pharmaceutical Web sites for published trials or trials conducted but unpublished between January 1996 and October 2007.
2,275 citations were reviewed, 285 studies were retrieved, and 90 were included in the analysis. Trials reporting response and/or remission rates in adult subjects treated with an antidepressant monotherapy for unipolar major depression were included.
The primary investigator recorded the number of responders and/or remitters in the intent-to-treat population of each study arm or computed these numbers using the quoted rates.
Poisson regression analyses demonstrated that mean response rate for the active medication was higher in studies comparing 2 or more active medications without a placebo arm than in studies comparing 2 or more active medications with a placebo arm (65.4% vs 57.7%, P < .0001) or in studies comparing only 1 active medication with placebo (65.4% vs 51.7%, P = .0005). Mean response rate for placebo was significantly lower in studies comparing 1 rather than 2 or more active medications (34.3% vs 44.6%, P = .003). Mean remission rates followed a similar pattern. Meta-analysis confirmed results from the pooled analysis.
These data suggest that antidepressant response rates in randomized control trials may be influenced by the presence of a placebo arm and by the number of treatment arms and that placebo response rates may be influenced by the number of active treatment arms in a study.

24 Reads
  • Source
    • "Observations in clinical trials confirm that the possibility of receiving a placebo decreases the overall efficacy of the drug. Trials with active comparators reveal higher efficacy for a drug than trials with placebo arms [8] [16] [35] [37], and it can be hypothesized that this is due to different expectations generated by these study designs . Second, many drugs induce minor bodily sensations immediately after administration that could induce greater nonspecific effects. "
    [Show abstract] [Hide abstract]
    ABSTRACT: The knowledge of having only a 50% chance of receiving an active drug can result in reduced efficacy in blinded randomized clinical trials (RCTs) compared to clinical practice (reduced external validity). Moreover, minor onset sensations associated with the drug (but not with an inert placebo) can further challenge the attribution of group differences to drug-specific efficacy (internal validity). We used a randomized experimental study with inert placebos (inert substance) vs active placebos (inducing minor sensations), and different instructions about group allocation (probability of receiving drug: 0%, 50%, 100%). One hundred forty-four healthy volunteers were informed that a new application method for a well-known painkiller would be tested. Pain thresholds were assessed before and after receiving nasal spray. Half of the nasal sprays were inert placebos (sesame oil), while the other half were active placebos inducing prickling nasal sensations (sesame oil with 0.014% capsaicin). The major outcome was pain threshold after placebo application. A substantial expectation effect was found for the inert placebo condition, with participants who believed they had received an active drug reporting the highest pain thresholds. Active placebos show substantial differences to passive placebos in the 50% chance group. Therefore, patient expectations are significantly different in placebo-controlled clinical trials (50% chance) vs clinical practice (100% chance). Moreover, minor drug onset sensations can challenge internal validity. Effect sizes for these mechanisms are medium, and can substantially compete with specific drug effects. For clinical trials, new study designs are needed that better control for these effects.
    Pain 10/2012; 153(12). DOI:10.1016/j.pain.2012.09.007 · 5.21 Impact Factor
  • Source
    • "Greater severity and longer duration of illness have been associated with greater drug–placebo separation in some (Fournier et al, 2010; Kirsch et al, 2008; Khan et al, 1991; Kobak et al, 2009), but not all (Khan et al, 2007) studies. Among trial design factors, greater placebo effects have been associated with greater number of treatment arms (Sinyor et al, 2010), more frequent trial visits (Posternak and Zimmerman 2007), and flexible rather than fixed dosing (Khan et al, 2007). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Variability in placebo response greatly complicates the design, conduct, and interpretation of clinical trials of antidepressant medications. To identify factors that impact detection of antidepressant-placebo differences, we conducted a meta-analysis of all relevant phase II-IV clinical trials for major depressive disorder conducted by the manufacturer of venlafaxine and desvenlafaxine completed by March 2011. We examined 15 factors potentially relevant to trial outcomes, using the standardized mean difference on the Hamilton Rating Scale for Depression (HAM-D(17)) score as the primary outcome. Thirty trials comprising 8933 patients were included. In univariate analyses, antidepressant efficacy (ie, drug vs placebo difference) was predicted most strongly (β=3.74, p=0.0002) by the proportion of patients in the trial enrolled from academic sites. Other factors predicting larger drug-placebo differences included lower participant completion rate, fewer post-baseline study visits, earlier year of study, and study drug (venlafaxine>desvenlafaxine). In multivariate meta-regression modeling, only the proportion of patients from academic sites maintained statistical significance as a predictor of drug-placebo separation for both HAM-D(17) continuous score change (β=2.24, p=0.034) and response rate (β=2.26, p=0.035). Including a higher proportion of academic sites may increase the ability to detect differences between active drug and placebo in clinical trials of major depressive disorder.Neuropsychopharmacology advance online publication, 22 August 2012; doi:10.1038/npp.2012.153.
    Neuropsychopharmacology: official publication of the American College of Neuropsychopharmacology 08/2012; 37(13). DOI:10.1038/npp.2012.153 · 7.05 Impact Factor
  • Source
    • "Indeed, inadequate concealment of allocation sequences and inadequate blinding have been shown to increase estimates of treatment effects [19], [20]. Regardless of the success of blinding, patients' beliefs about the likelihood of placebo allocation are associated with the magnitude of placebo response [21]–[24]. Furthermore, merely obtaining informed consent has been shown to alter the magnitude of placebo effects [25] and target treatment effects [26]. There is also preliminary evidence that informing participants that placebos elicit side-effects reduces patient-reports of side-effects [27], and the same adverse events are reported in placebo arms as in the corresponding treatment arm of trials in migraine [28] and depression [29]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Placebo groups are used in randomised clinical trials (RCTs) to control for placebo effects, which can be large. Participants in trials can misunderstand written information particularly regarding technical aspects of trial design such as randomisation; the adequacy of written information about placebos has not been explored. We aimed to identify what participants in major RCTs in the UK are told about placebos and their effects. We conducted a content analysis of 45 Participant Information Leaflets (PILs) using quantitative and qualitative methodologies. PILs were obtained from trials on a major registry of current UK clinical trials (the UKCRN database). Eligible leaflets were received from 44 non-commercial trials but only 1 commercial trial. The main limitation is the low response rate (13.5%), but characteristics of included trials were broadly representative of all non-commercial trials on the database. 84% of PILs were for trials with 50:50 randomisation ratios yet in almost every comparison the target treatments were prioritized over the placebos. Placebos were referred to significantly less frequently than target treatments (7 vs. 27 mentions, p<001) and were significantly less likely than target treatments to be described as triggering either beneficial effects (1 vs. 45, p<001) or adverse effects (4 vs. 39, p<001). 8 PILs (18%) explicitly stated that the placebo treatment was either undesirable or ineffective. PILs from recent high quality clinical trials emphasise the benefits and adverse effects of the target treatment, while largely ignoring the possible effects of the placebo. Thus they provide incomplete and at times inaccurate information about placebos. Trial participants should be more fully informed about the health changes that they might experience from a placebo. To do otherwise jeopardises informed consent and is inconsistent with not only the science of placebos but also the fundamental rationale underpinning placebo controlled trials.
    PLoS ONE 06/2012; 7(6):e39661. DOI:10.1371/journal.pone.0039661 · 3.23 Impact Factor
Show more