Open-label versus double-blind placebo treatment in irritable bowel syndrome: study protocol for a randomized controlled trial

To date, hundreds of thousands of research participants have received DBPs in the context of RCTs. In an RCT, the placebo group serves to control for placebo responses and for other common factors that may result in symptomatic improvement such as detection biases, natural waxing and waning of symptoms, and regression to the mean. If the experimental treatment is found to result in more symptomatic improvement when compared to the placebo treatment, the observed differences are generally considered to be due to the active drug ingredient and not to incidental factors such as placebo effects or spontaneous improvement. In most cases, administration of placebo to research participants occurs after providing informed consent, in which it is clearly explained that there is a chance of receiving either the active study intervention or an inactive placebo intervention. As such, even though research participants are kept blinded to their treatment assignments, there are relatively few ethical concerns regarding placebos in RCTs [5, 6].

The utilization of placebo in clinical practice, however, is traditionally much less transparent and is associated with a long history of ethically questionable practice [7, 8]. After the widespread adoption of the RCT as the “gold standard” of evidence-based research in the 1960s and the acknowledgement of patient rights in medicine, prevailing ethics renounced the deceptive use of placebos in clinical practice [9‐11]. Nonetheless, without much public discussion, deceptive use of placebos continues to be commonly used in clinical settings. For example, a national randomized survey published in 2008 of 679 practicing internists and rheumatologists in the USA found that approximately 50% of respondents regularly prescribed placebos [4] and a survey of 208 randomly selected general practitioners in Germany found that 76% reported prescribing placebos [12]. Systematic reviews of such research have found similar or even higher rates of placebo utilization around the world [13]. This line of research has also consistently found that “impure” placebos – i.e., genuine medications that physicians understood would have no intrinsic pharmacological action on patients’ symptoms – were much more commonly prescribed than “pure” placebos such as sugar pills [14]. In these surveys, physicians clearly indicated that placebos were prescribed for “psychological benefits,” not uncommonly as part of a “watch-and-wait” strategy, and that patients rarely, if ever, were told that they were receiving placebos or pharmaceuticals serving the purpose of placebos. The reason that providers do not inform their patients of their choice to prescribe placebo likely stems from the conventional wisdom that patients must believe that placebo pills are “real” medications and that informing patients that they have been prescribed a placebo would “erase” the placebo effect; in other words, it is widely assumed that honesty and transparency are incompatible with placebo effects [15].

Can placebo effects be ethically harnessed with honestly prescribed OLP? Until recently this question was considered absurd. However, there is significant value in exploring whether an honest placebo strategy might be helpful in treating common symptoms, especially in “watch-and-wait” treatment strategies or before prescribing medications with potentially serious side effects. Recently, there have been several proof-of-concept studies testing whether OLP can improve clinical outcomes compared to a treatment-as-usual (“no-treatment control” (NTC)) group. The first such study involved a sample of patients with IBS (n = 80) randomized to OLP versus NTC [16]. Among OLP patients, 60% reported “adequate relief” of their IBS symptoms versus 37% “adequate relief” in NTC patients (p = 0.03). A more recent study randomly assigned patients with chronic low back pain (cLBP; n = 83) to a similar design and reported that, on a 0-to-10 composite pain scale, patients receiving OLP reported 30% reduction of usual and maximum pain compared to reductions of 9% and 16% in usual and maximum pain, respectively, in the continued NTC (p < 0.001). There was a 29% reduction in pain-related disability in patients receiving OLP compared to 0.02% in the continued NTC [17]. A third, elaborate, within-subject randomized experiment with patients serving as their own control during acute episodic migraine attacks (n = 459 documented attacks in 66 patients) included a nested comparison of OLP versus NTC and found that the OLP reduced pain by 15% compared to 15% worsening in the NTC, so that the total mean difference between conditions was 30 percentage points, p = 0.001) [18]. Other similar, but much smaller, studies in allergic rhinitis [19] and depression [20], as well as other somewhat similar studies using placebo-conditioning models in pediatric attention-deficit hyperactive disorder [21], insomnia [22], psoriasis [23], and healthy participants [24], have reported similar, positive outcomes. Clearly, many questions remain regarding whether OLP can be adopted for clinical use. We have recently begun a National Institutes of Health (NIH)-funded RCT in an attempt to answer some of these unresolved questions.

This paper describes the methodology of an ongoing RCT, titled “Efficacy of open-label vs. double-blind treatment in IBS,” designed specifically to address the OLP research areas outlined below. This project brings together a multidisciplinary team of clinical, behavioral, and basic scientists to explore a series of rarely examined questions in OLP.

The primary aim of this study is to compare symptoms after 6 weeks of OLP, DBP, and NTC in a sample of individuals with IBS (Fig. 1). In order to provide DBP treatment, this study includes a nested trial comparing double-blind peppermint oil and DBP. Peppermint oil was chosen for use in the double-blind arm because of its suggested antispasmodic effects on the gastrointestinal tract [25, 26].

Our secondary areas of interest are largely informed by our previous placebo studies in IBS: