Prescribing ‘open-label’ placebo
There may be some limited circumstances where an actual placebo (eg, a sugar pill) might be suitable as a treatment. These include when placebo and conventional treatment provide similar results and a patient is reluctant to take conventional medicine, or when there is no effective conventional treatment. The deceptive prescription of placebo (providing placebo and calling it a drug) has a long history and was considered ethical—and recommended by medical authorities—until the latter half of the 20th century. This practice was deemed unethical in the 1980s, because it was dishonest and violated patient autonomy. Because it was widely believed that placebos given openly would be ineffective, the end of placebo treatment seemed at hand. An intriguing body of evidence, however, suggests that placebos can be effective even when patients know they are taking a placebo. Patients given an “open-label” placebo are told something along the lines of “the pill being prescribed contains no medicine, but some people improve with it, perhaps because the pill stimulates the body’s self-healing.” Open-label placebo has been evaluated for depression,22 low back pain,23 irritable bowel syndrome,24 neurosis,25 allergic rhinitis,26 and anxiety.27 Most of these studies are small, and some were uncontrolled. Yet they consistently have shown that symptoms improve with a nondeceptive placebo, and improve to a greater extent than with no treatment.
The most recent trial is a promising example of the potential of open-label placebos. In this study, 96 patients with chronic low back pain were randomly assigned to 3 weeks of treatment as usual (TAU) or 3 weeks of TAU plus open-label placebo.23 Patients who received open-label placebo were educated about the placebo effect and shown a film clip describing promising results of a prior open-label placebo study. They were then given placebo pills to be take once daily, and clearly told the pills contained no active medication. After 3 weeks, patients in the TAU plus placebo group reported less pain and less disability than patients who received TAU without a placebo. Some patients even requested a placebo prescription at the end of the study.
The placebo response provides a rational basis for prescribing innocuous alternative therapies with no intrinsic therapeutic value. Patients who prefer and believe in the effectiveness of alternative remedies—herbal compounds, massage, magnets, homeopathic solutions, etc.—can be recommended these treatments to mobilize a placebo response.
Using a conditioning model. Prescribing a placebo to obtain a conditioned drug response has enormous but untapped clinical potential. Both animal and human research indicates that a wide range of drug responses, from immune suppression to motor stimulation, can be conditioned (a neutral stimulus, such as a pill or injection, associated with drug administration can in itself evoke the drug effect). In many conditioning or dose-extending models, a particular response to real medication (such as pain relief after analgesics) first becomes conditioned due to repeated exposure to the drug given in a particular vehicle. Then, the treatment shifts to some doses comprising of real medicine and some doses comprising of placebo. Because the drug response has been conditioned, it is thought that the response to an identically appearing placebo will mirror the drug response. The active drug often is only replaced by placebo for certain doses under a schedule of partial reinforcement, given the ubiquity of extinction (the conditioned response lessens when the conditioned stimulus is presented alone on repeated trials).
In 1 version of a conditioning study, children with ADHD were randomized to 1 of 3 groups.28 One group (full dose) took the standard dose of medication for 2 months, a second group (reduced dose) took a standard dose during 1 month followed by a half dose during the second month, and children in the third group (reduced dose with placebo) took the standard dose plus a visually distinctive placebo during the first month, followed by a half dose plus the visually distinctive placebo during the second month. Not surprisingly, ADHD symptoms were worse among children in the reduced-dose group. However, there was no difference between those in the reduced-dose with placebo group and those in the full-dose group. It appears as though the symptom reduction associated with a 100% dose was an unconditioned response that could be mimicked with the addition of a placebo pill.
In another study, patients with psoriasis were randomly assigned to receive a full dose of active medication (0.1% triamcinolone cream) twice a day, or a full dose of active medication for 25% to 50% of the doses, with a placebo (moisturizing cream) given for the other 50% to 75% of the doses.29 Relapse rates were not statistically different between groups.
