CCTMB #002 Your Placebo Will See You Now: Controlling for Placebo Effects in Complex Intervention Trials

By Dr David Hohenschurz-Schmidt, Postdoctoral Researcher in Pain Research and Trial Methods for Complex Interventions

Blinding is a fundamental principle of high-quality randomised controlled trials. It involves withholding key information about treatment allocation from trial participants (and often investigators and outcome assessors). The aim is to balance expectations across study groups so that any observed treatment effects can be more confidently attributed to the intervention itself rather than to participants’ beliefs or expectations.

Closely linked to blinding are the concepts of placebo effects and placebo controls. In drug trials, an active medication is typically compared with a similar-looking and similar-tasting placebo pill that contains no pharmacologically active ingredient. This helps achieve blinding and equalise expectations between groups.

Expectations are a major driver of placebo and nocebo effects, i.e., changes in symptoms or outcomes that arise from receiving a treatment and its surrounding context, rather than from the treatment’s active ingredients themselves. In essence, this is the rationale underpinning the randomised placebo-controlled trial - widely regarded as the gold standard for testing treatment efficacy and often required for regulatory approval of drugs.

Now consider what a placebo-controlled trial might look like for physiotherapy, psychotherapy, surgery, or exercise. Unlike drug trials, these interventions are often evaluated without placebo controls. Such controls are frequently considered infeasible and are generally not required by regulators. This is problematic because placebo effects are likely to be particularly strong in interventions that are interactive, physical, invasive, or expensive. For many widely used interventions, from spinal surgery for back pain to psychotherapies for a range of conditions, we may therefore remain uncertain about the extent to which benefits arise from specific treatment mechanisms versus contextual and expectancy effects.

In some areas where placebo- or sham-controlled trials have been conducted, the findings have been sobering. Sham-controlled trials of arthroscopic procedures for knee pain, for example, have shown little or no benefit over placebo procedures. These studies have fundamentally changed how we think about the efficacy of some common interventions.

Placebo-controlled trials of complex interventions are, admittedly, complex. But they are not impossible. There are now numerous examples demonstrating that such trials are both feasible and essential for understanding the true benefits and harms of interventions. Several methodological guidelines now help researchers design placebo or sham controls for complex interventions. Notable examples include the ASPIRE guideline for surgical trials and the CoPPS Statement for non-invasive therapies, which I led the development of.

These guidelines highlight that designing placebo controls for complex interventions is fundamentally a conceptual exercise. Researchers must decide which component(s) of an intervention are thought to be therapeutically important and therefore should be “switched off” or omitted in the control condition, while keeping everything else as similar as possible. This produces highly similar control interventions. In surgery, for example, participants may receive anaesthesia and surgical incisions without the critical surgical step itself. In manual therapy trials, a therapist may mimic the procedure without applying the hypothesised therapeutic force.

Few would argue that these control interventions are equivalent to swallowing an inert sugar pill. Sham procedures for complex interventions remain complex interventions in their own right. They often still involve substantial contextual elements: time with clinicians, therapeutic interactions, education, touch, rituals, and sometimes anaesthesia. Importantly, these controls may also not be designed to or be unable to remove all potentially therapeutic components, of which, by definition, complex interventions contain multiple.

This has important implications for interpreting placebo-controlled trials of complex interventions, which might differ from interpreting placebo-controlled drug trials. It is conceivable that high-similarity placebo-controlled trials of genuinely effective complex interventions may show relatively small between-group effects because the control intervention itself still contains many active contextual ingredients. Moreover, these trials can typically only isolate the contribution of a limited number of components at a time.

Nonetheless, placebo-controlled trials are indispensable. They cannot simply be replaced by comparative effectiveness trials comparing treatments against waiting lists, minimal care, or “usual care”.

Imagine someone tried to sell you a treatment, whether knee surgery, a physical intervention, or Voodoo*, and told you that it worked better than doing nothing, but only because it generated positive expectations and a supportive experience. Treatments based primarily on placebo effects raise questions around safety, resource allocation, opportunity costs, transparency, and informed consent. Also, a few aspiring practitioners would invest time and money into knowingly training in such interventions.

The relevance of placebo-controlled trials does not mean comparative effectiveness trials are unimportant. On the contrary, they answer highly relevant clinical questions, including which treatment may work best in practice. But they complement rather than replace placebo-controlled efficacy trials. Ideally, complex intervention research programmes should include a range of study designs addressing different questions: mechanisms, efficacy, safety, comparative effectiveness, implementation, and real-world uptake. In practice, researchers rarely have the luxury of funding for such comprehensive programmes. Pragmatic decisions, therefore, need to be made about which questions are most pressing and which study designs are feasible at a given time.

Funders, regulators, and the wider research community nevertheless share a responsibility to ensure that questions about placebo effects and specific efficacy in complex interventions do not remain perpetually unanswered.

Key takeaways:

1. The CoPPS Statement offers essential guidance on how control interventions can be developed, delivered, and reported in efficacy and mechanistic trials across physical, psychological, and self-management therapies.

2. Among its key recommendations is the use of high-similarity control interventions to balance expectation effects between study arms and isolate intervention components of interest.

3. Placebo-controlled trials of complex interventions are irreplaceable for ethical and resource-related reasons; They are complemented by other trial designs for questions of real-world or comparative effectiveness.

   
   

* Apologies to practitioners of Vodou and related religions. Feel free to insert “Catholicism” instead. 

References and resources

1. Abram, S.G.F., Hopewell, S., Monk, A.P. et al. 2020. Arthroscopic partial meniscectomy for meniscal tears of the knee: a systematic review and meta-analysis. Br J Sports Med 54, 652–663. https://doi.org/10.1136/bjsports-2018-100223

2. Beard, D.J., Campbell, M.K., Blazeby, J.M. et al. 2020. Considerations and methods for placebo controls in surgical trials (ASPIRE guidelines). The Lancet 395, 828–838. https://doi.org/10.1016/S0140-6736(19)33137-X

3. Hohenschurz-Schmidt, D., Phalip, J., Chan, J. et al. 2024. Placebo analgesia in physical and psychological interventions: Systematic review and meta-analysis of three-armed trials. European Journal of Pain 28, 513–531. https://doi.org/10.1002/ejp.2205

4. Hohenschurz-Schmidt, D., Vase, L., Scott, W. et al. 2023. Recommendations for the development, implementation, and reporting of control interventions in efficacy and mechanistic trials of physical, psychological, and self-management therapies: the CoPPS Statement. BMJ 381, e072108. https://doi.org/10.1136/bmj-2022-072108

5. Hohenschurz-Schmidt, D.J., Cherkin, D., Rice, A.S.C. et al. Research objectives and general considerations for pragmatic clinical trials of pain treatments: IMMPACT statement. PAIN 164, 1457. https://doi.org/10.1097/j.pain.0000000000002888