Danger of Reliance on Placebo Effect

With asthma patients, giving an inhaled bronchodilator medication, like albuterol, can result in a rapid improvement in airflow that can be objectively measured using spirometry testing. This testing can be repeated over a short period of time, thus making asthma an excellent model for assessing a “placebo effect” (benefit resulting from simulated treatment or the experience of care).

In a randomized, double-blind, cross-over study published in the New England Journal of Medicine, Wechsler et al compared 4 intervention arms in asthmatic patients:

  1. Active medication—double-blinded albuterol inhaler
  2. Placebo #1—double-blinded inert inhaler
  3. Placebo #2—single-blinded sham acupuncture
  4. No intervention at all

Only those asthmatics with documented FEV1 bronchodilator responsiveness (of at least 12%) were enrolled and each study subject received each of the four interventions three times.

The authors looked at objective evidence of airflow improvement by measuring spirometry, as well as self-reported subjective improvement, and here is what they found:

  1. Albuterol produced a significant improvement in FEV1 (20.1%), as compared to only about 7% with each of the other three interventions (p<0.001). There was no statistical difference in airflow improvement among either placebo arm or the “nothing” arm.
  2. Patients’ subjective report of improvement did not significantly differ between the albuterol inhaler (50%), placebo inhaler (45%), or sham acupuncture (46%), but all three were significantly better than no intervention (21%)(p<0.001).
  3. In an assessment of treatment credibility (percent of subjects’ subjective belief they had received an active treatment), sham acupuncture (85%) was statistically “more credible”(p<0.005) than either albuterol inhaler (73%) or placebo inhaler (66%).

For the objective testing, there was a powerful medication effect, but no placebo effect. For the subjective outcome, the placebo effects were equivalent to the active medication effect, and all were better than no intervention. Placebos offered no objective bronchodilator effect beyond no intervention of any kind. Incredibly, subjective improvement for placebo was similar to active drug, even though active drug had 3 times the bronchodilator effect and placebo was no better than no intervention.

In this well designed study, with the inclusion of a “no intervention” arm, the authors are able to demonstrate that placebos are primarily detectable in subjective outcomes, and that studies reporting only subjective results should be interpreted with extreme caution.

Don’t trust the “snake oil” study if it relies on subjective results, rather than objective measurements.