Original article
What are the main methodological problems in the estimation of placebo effects?

https://doi.org/10.1016/S0895-4356(01)00496-6Get rights and content

Abstract

Researchers disagree whether placebo effects are clinically important. The controversy is partly due to conceptual and methodological problems. The objective of this paper is to analyze the main methodological problems in the estimation of placebo effects. Variations in the meaning of the concept of placebo effect are described, and the main methodological problems are analyzed. The notion of placebo effect has at least two main meanings: effect of placebo intervention, and effect of patient-provider interaction. When the terms are defined pragmatically, effects of placebo can be estimated as the difference between placebo and no-treatment in randomized trials. The effect of patient-provider interaction can similarly be assessed by comparing manipulation of the patient-provider interaction with no manipulation. In both cases bias due to lack of double-blinding is a potential problem.

Introduction

Investigations of the placebo effect intensified in the 1950s, when several researchers, e.g., Beecher [1], Wolf [2], and Lasagna [3], reported dramatic placebo effects. The standard view has since been that placebo effects have repeatedly been demonstrated on both objective and subjective outcomes in a broad spectrum of clinical conditions. For example, a narrative review reported clinically important placebo effects in asthma, hypertension, and survival after myocardial infarction [4].

However, the position has met severe criticism. Detailed methodological analysis of the primary trials in Beecher's frequently quoted work showed that, to a large extent, his conclusion was based on uncontrolled evidence [5]. Shapiro's and Shapiro's overview of the placebo literature furthermore found that “… there is no systematic approach in published studies of the placebo effect. All … are anecdotal reports, clinical impressions, theoretical formulations and post hoc extrapolations of chance findings.” [6].

The scientific controversy over placebo effects is partly caused by misinterpretations of results from clinical research 5, 6, and partly by conceptual inconsistencies as the term is used differently by various authors. The primary objective of this article is to analyze the main methodological problems in the estimation of the placebo effect. However, the methodological discussion of how to measure such an effect rests on a clarification of what is meant by the placebo effect, so a preliminary analysis will focus on what authors actually do imply with the notion.

In the following article, I describe three main meanings of the placebo effect: change after a placebo intervention, the effect of a placebo intervention, and the effect of the patient-provider interaction. I will argue that despite common use, the three meanings are either not interesting to placebo research or so vague that no clearly delineated group of interventions can be defined. However, a possible pragmatic approach to the estimation of the effect of placebo is to compare interventions used as placebos in clinical trials with no-treatment. Similarly, the effect of defined manipulations of the patient-provider interaction can be estimated by comparing manipulations with no manipulation.

Section snippets

Change after placebo medication

Numerous articles have been published that claim to demonstrate a placebo effect on the basis of the before-after difference in a placebo group 7, 8, 9. In many cases this reflects the classic methodological error of the placebo research: Effects are claimed without a proper control group. There is no way effects of placebo intervention can be distinguished from the natural course of the disease, regression to the mean, or other factors that change the reported intensity of signs and symptoms

The concept of placebo

I will discuss the notion of placebo as it applies to the estimation of placebo effects. The concept of placebo is difficult to define satisfyingly. Many influential definitions link placebo with the terms specific and unspecific. One example is provided by Shapiro who in short defines a placebo treatment as an intervention without any specific effect on a given disease (but with a possible unspecific effect) [26]. This is highly problematic, as the concept of specificity is vague, and

The concept of patient-provider interaction

The patient-provider interaction describes an interpersonal process between a patient and health providers. Often this process incorporates the totality of influences from both the patient, the practitioner, and the setting [22]. In some situations these factors are split up, and the patient-provider interaction is a narrower concept focused on the dynamics between patient and provider only. Effects of such interactions have been called context effects [33]. However, in addition to an unclear

Conclusion

When trials and reviews of placebo effects are appraised, several conceptual and methodological problems must be addressed. The first main problem is to assure that the term placebo effect meant a causal relation between some kind of intervention and a therapeutic effect. Second, among several variations in meaning, placebo effect can imply both effect of placebo and effect of patient-provider interaction. Third, neither placebo nor patient-provider interaction is clearly defined as a notion

References (41)

  • W.A. Brown

    The placebo effect

    Scientific American

    (1998)
  • A.K. Shapiro et al.

    The placebois it much to do about nothing?

  • L.E. Hollister

    Placebologysense and nonsense

    Curr Ther Res

    (1960)
  • S. Ross et al.

    Placebo agentryassessment of drug and placebo effects

  • O. Lindahl et al.

    Is all therapy just a placebo effect?

    Metamedicine

    (1982)
  • E. Ernst et al.

    Concept of true and perceived placebo effects

    BMJ

    (1995)
  • R. Rosenthal

    Designing, analysing, interpreting and summarizing placebo studies

  • G. Honigfeld

    Non-specific factors in treatment IIreview of social-psychological factors

    Diseases of the Nervous System

    (1964)
  • K.L. Kwekkeboom

    The placebo effect in symptom management

    Oncology Nursing Forum

    (1998)
  • A.J.M. deCraen et al.

    Placebos and placebo effectsA historical overview

  • Cited by (144)

    • A pre-trial evaluation of blinding for a Chinese herbal medicine trial

      2020, Contemporary Clinical Trials Communications
    View all citing articles on Scopus
    View full text