Subscribe to eTOC

Can You Avoid the Nocebo Response But Still Fully Inform Patients About Potential Medication Side Effects?

A 35-year-old thin woman with a history of insomnia arrives in your office and you diagnose her with chronic tension-type headaches and episodic migraines without aura. Her headaches have become unbearable and are interfering with her life, and she has never been on a daily prophylactic medication for treatment. You consider amitriptyline, as it may treat both her tension-type headaches and migraines, and likely will also help her insomnia, but you recognize that there can be many side effects associated with this medicine. You wonder if telling her about the side effects may increase the likelihood that they will occur — creating a nocebo response — but fear that not telling her about them limits her right to full informed consent. How would you decide what to tell her about potential side effects?


Research studies show that what we tell our patients about potential adverse events can directly affect — or even influence — the side effects they experience. The very act of describing side effects may result in a self-fulfilling prophecy; on learning about these symptoms, patients may experience negative placebo — or nocebo — responses that may actually result in more harm. [See “Evidence for the Nocebo Response.”]

There are practical issues that play a role in this process as well. Some physicians feel compelled to describe all possible side effects, or provide handouts listing all such events, in order to protect themselves from the risk of future lawsuits if a patient experiences an untoward side effect without being fully informed. Furthermore, patients have access to a wide range of resources that describe side effects, from Internet websites to pharma medication inserts.

How then do we fully inform our patients about potential side effects and still comply with the principle of nonmaleficence — to “do no harm”? This is an ethical challenge.

Should we avoid a discussion of possible side effects altogether? The obvious answer is no. Patients have to be a part of the decision-making process regarding their care and treatment; withholding this information would take us back to the days where paternalism prevailed. In fact, studies have shown that patients feel empowered — and compliance and outcomes improve — when they participate in decisions about their treatment. Furthermore, truthful and meaningful communication is a critical part of the patient-physician relationship.

I believe that we can both inform patients about side effects and limit the potential for the nocebo response through a process of “contextualized informed consent.” How? By tailoring our conversations to account for the potential side effects, the patient being treated, and his or her particular diagnosis.

Research suggests, for example, that patients who learn about nonspecific symptoms — such as dizziness and fatigue — may be more susceptible to experiencing those effects than the adverse events associated with certain drugs, such as glaucoma associated with topiramate use. Moreover, patients with a history of adverse events, nonspecific symptoms at baseline, and a tendency toward somatization, anxiety, and depression are more likely to experience nocebo responses. The underlying diagnosis being treated and its severity and life-disabling state also should be balanced against the nature of the side effect.

In the Feb. 8 Journal of the American Medical Association, Luana Colloca, MD, and Damien Finniss, MSc Med, suggest that negative information about medications can be minimized by framing the information in a positive way. Further, they argue that the information should be tailored to account for what patients already know and want to know.

Getting back to our 35-year-old patient, we know that amitriptyline could significantly improve her headaches and quality of life; it is one of the few standard treatments effective for both tension-type headaches and migraines. We also know that the therapy could lead to weight gain and constipation and should be avoided in patients with these problems; the medication should also be avoided in patients who are psychiatrically unstable or have a history of cardiac disease or glaucoma. This patient has no such contraindications.

I would, therefore, start the discussion about this medication by describing the potential benefits for her headaches and framing the side effect of drowsiness as a way to treat her insomnia. For example, you might say: “I am very optimistic about your response and believe that this medication will not only help your headaches, but also your insomnia. I recommend you take this medication before bedtime as the drowsiness side effect will actually aid your sleep.”

I would also emphasize the importance of eating a healthy diet as weight gain and/or constipation can be a problem for some patients on this medicine. I would explain that initially when starting this medication or when taking it at high doses, other side effects may occur but that these problems will be monitored. I would tell her that we will start off at a low dose, slowly increase the dose, and that she can contact me if she experiences other problems.

Framing the information in this way helps accomplish multiple goals: it informs the patient about potential side effects, yet minimizes the potential for the nocebo effect by focusing on the positive aspects of the medicine without emphasizing the numerous nonspecific side effects; at the same time it opens up opportunities to discuss freely any side effects that may emerge later.

Having discussions about side effects with our patients is critical for supporting patient adherence, enhancing strong patient-doctor communication, and keeping patients informed of risks/benefits in the decision-making process. Yet it is also important for physicians to be aware of the issues that surround the appropriate disclosure of medication side effects, to think critically about what information they discuss with patients, and to realize that they can shape discussions to optimize outcomes while maintaining patient autonomy.

Clearly, we need more evidence-based research on how best to fully inform our patients about their therapies and promote their healing — without risking the potential for nocebo responses — and potential harm. An open discussion among physicians on the question of limiting non-specific adverse events that are likely due to nocebo effects would benefit patients and the profession.

Dr. Wells is an assistant professor of neurology and the director of clinical services at the Center for Integrative Medicine at Wake Forest Baptist Health in Winston-Salem, NC. She completed a three-year research fellowship in integrative medicine at Harvard Medical School and Beth Israel Deaconess Medical Center and a one-year headache fellowship at Brigham and Women's Faulkner Hospital.


There is a body of evidence that demonstrates that the specific information provided to patients about side effects can directly affect the side effects patients experience. For example, a trial assessing the impact of knowledge of sexual side effects of beta-blockers showed that those blinded to drug treatment reported significantly fewer sexual side effects (3.1 percent) than those told the drug name (15.6 percent), and both reported fewer adverse events than those told the drug name and the possibility of sexual side effects (31.2 percent, p<0.01).

A systematic review of the adverse events patients experience in the placebo arms of anti-migraine clinical trials showed that the adverse events experienced by those with placebo corresponded to the side effects that were expected for the active anti-migraine drug against which the placebo was compared.

The way information is framed affects outcomes as well: patients told that a local anesthetic injection was going to “feel like a bee sting” experienced more pain than those told that they would receive a local anesthetic “to numb the area so you will be comfortable during the procedure.”

While patients on placebo in clinical trials may drop out because of the side effects they experience and later feel frustrated or “duped” on discovery that they were “just on placebo,” neurobiological evidence suggests there is an objective neurophysiologic correlate to the nocebo phenomenon.

Verbally-induced nocebo hyperalgesia has been shown to result in hyperactivity of the hypothalamic-pituitary-adrenal axis that can be antagonized by a benzodiazepine. The nocebo response may be involved in deactivation of dopamine and opioid release, the opposite neurophysiologic phenomenon to the placebo response. And a recent study even demonstrated that these placebo and nocebo responses may occur without conscious awareness of suggestive environmental cues. See the citations for these studies related to nocebo response in the box, “For Further Reading.”

—Rebecca E. Wells, MD


• Epstein RM, Alper BS, Quill TE. Communicating evidence for participatory decision making. JAMA 2004;291(19):2359–2366.
    • Wells RE, Kaptchuk TJ. To tell the truth, the whole truth, may do patients harm: the problem of the nocebo effect for informed consent. Am J Bioeth 2012;12(3):22–29.
      • Colloca L, Finniss D. Nocebo effects, patient-clinician communication, and therapeutic outcomes. JAMA 2012;307(6):567–568.
        • Silvestri A, Galetta P, Rosano GM, et al. Report of erectile dysfunction after therapy with beta-blockers is related to patient knowledge of side effects and is reversed by placebo. Eur Heart J 2003;24(21):1928–1932.
          • Amanzio M, Corazzini LL, Benedetti F, et al. A systematic review of adverse events in placebo groups of anti-migraine clinical trials. Pain 2009;146(3):261–269.
            • Varelmann D, Pancaro C, Camann WR, et al. Nocebo-induced hyperalgesia during local anesthetic injection. Anesth Analg 2010;110(3):868–870.
              • Benedetti F, Amanzio M, Vighetti S, Asteggiano G. The biochemical and neuroendocrine bases of the hyperalgesic nocebo effect. J Neurosci 2006 Nov 15;26(46):12014–12022.
                • Scott DJ, Stohler CS, Zubieta JK, et al. Placebo and nocebo effects are defined by opposite opioid and dopaminergic responses. Arch Gen Psychiatry 2008;65(2):220–231.
                  • Jensen KB, Kaptchuk TJ, Kong J, et al. Nonconscious activation of placebo and nocebo pain responses. Proc Natl Acad Sci U S A 2012;109(39):15959–15964.