Nonspecific Medication Side Effects and the Nocebo Phenomenon
Total Page:16
File Type:pdf, Size:1020Kb
SPECIAL COMMUNICATION Nonspecific Medication Side Effects and the Nocebo Phenomenon Arthur J. Barsky, MD Patients taking active medications frequently experience adverse, nonspe- Ralph Saintfort, MD cific side effects that are not a direct result of the specific pharmacological Malcolm P. Rogers, MD action of the drug. Although this phenomenon is common, distressing, and Jonathan F. Borus, MD costly, it is rarely studied and poorly understood. The nocebo phenomenon, in which placebos produce adverse side effects, offers some insight into non- LMOST 3 BILLION PRESCRIP- specific side effect reporting. We performed a focused review of the litera- tions are filled each year in ture, which identified several factors that appear to be associated with the outpatient settings in the nocebo phenomenon and/or reporting of nonspecific side effects while tak- United States, an increase of A50% since 1992.1 Although many side ing active medication: the patient’s expectations of adverse effects at the effects (generally defined as an action outset of treatment; a process of conditioning in which the patient learns of a drug other than the one for which from prior experiences to associate medication-taking with somatic symp- it is being used) result directly from toms; certain psychological characteristics such as anxiety, depression, and these drugs’ pharmacological activity, the tendency to somatize; and situational and contextual factors. Physi- many others cannot be attributed to cians and other health care personnel can attempt to ameliorate nonspecific their specific pharmacological ac- side effects to active medications by identifying in advance those patients tions. These nonspecific side effects dis- tress patients, add to the burden of their most at risk for developing them and by using a collaborative relationship illness, and increase the costs of their with the patient to explain and help the patient to understand and tolerate care. They may lead to nonadherence, these bothersome but nonharmful symptoms. cause physicians to discontinue what JAMA. 2002;287:622-627 www.jama.com is otherwise an appropriate therapy, or prompt attempts to treat these side ef- effects may be positive and beneficial or its beneficial, therapeutic effects,2-4 and fects with additional drugs. negative and adverse. In this article, we in this article it will be used broadly to In this article, we use the nocebo phe- are concerned only with the latter, and refer to all distressing symptoms that nomenon to explore the occurrence of in the interests of brevity will use the accompany placebo administration. adverse, nonspecific side effects in pa- general term “nonspecific side effects” tients taking active medication and sug- to refer only to negative or adverse symp- Methods gest ways in which clinicians can deal toms or physical changes. Similarly, the We conducted a focused review of ar- more effectively with them. Side effects term “side effects” will be used to refer ticles relevant to the nature, inci- occurring in patients taking active medi- to unintended adverse effects. dence, magnitude, and medical man- cation may be divided into 2 types. “Spe- The nocebo phenomenon may help agement of nonspecific medication side cific side effects” are symptoms or physi- us understand (adverse) nonspecific effects. The MEDLINE database was ological changes that result directly from side effects. The nocebo (meaning in searched for English-language articles the specific biological and pharmaco- Latin “I will harm”) phenomenon re- from 1966 through the present, using logical activity of the drug and tend to fers to symptoms and/or physiological the following Medical Subject Head- be dose-dependent and predictable. changes that follow the administra- ings (MeSH) terms: adverse effects, side “Nonspecific side effects” are symp- tion of an inert, chemically inactive sub- effects, symptoms, nocebo, placebo, drug toms or physiological changes that can- stance that the patient believes to be an Author Affiliations: Department of Psychiatry, Brigham not be explained on the basis of the active drug. The term nocebo was origi- and Women’s Hospital and Harvard Medical School, known pharmacology of the drug and nally coined to distinguish the nox- Boston, Mass. Corresponding Author and Reprints: Arthur J. Bar- are idiosyncratic and not dose- ious or distressing effects of a placebo sky, MD, Department of Psychiatry, Brigham and dependent. In theory, nonspecific side (meaning in Latin “I will please”) from Women’s Hospital, 75 Francis St, Boston, MA 02115. 622 JAMA, February 6, 2002—Vol 287, No. 5 (Reprinted) ©2002 American Medical Association. All rights reserved. Downloaded From: http://jama.jamanetwork.com/ by a SCELC - Loma Linda University User on 03/13/2014 MEDICATION SIDE EFFECTS AND THE NOCEBO PHENOMENON reaction, and ambulatory care. The tients receiving a placebo device re- tributable to the drug’s specific, phar- search was extended using the bibliog- ported intensification of their preexist- macological activity. By analogy, we raphies of pertinent recent articles and ing pain.14 Second, the symptoms may suggest here that some fraction of the reviews. Articles were screened for rel- be the somatic concomitants of emo- side effects experienced by the pa- evance based on the title, key words, tion (such as anxiety or depression) or tients taking an active drug can be at- and abstracts. Articles were reviewed, of psychosocial stress. Third, patients tributed to the nocebo effect.18 analyzed, and synthesized, but no for- may mistakenly ascribe the symptoms of Approximately one quarter of pa- mal meta-analysis was conducted for 2 mild infirmities or benign, self-limited ail- tients taking placebo report adverse side reasons: first, this article is not in- ments (such as headaches, cramps, and effects.19,20 (In one striking example, hy- tended to be a systematic or compre- extrasystoles) or of normal physiologi- pervagotonia manifested by an idio- hensive summary of the literature, but cal functioning (eg, orthostatic dizzi- ventricular rhythm occurred with pla- rather a focused review. Second, the re- ness) to the medication. To explore the cebo administration in a double-blind search is too variable in methods and similarity between reported side effects study of a calcium channel blocker.21) quality for any standardized compari- and such endemic symptoms, Reiden- Rosenzweig et al22 found that 19% of son. More weight, however, was given berg and Lowenthal15 ascertained the in- healthy volunteers taking placebos in to empirical studies using more rigor- cidence of 25 commonly reported symp- 109 double-blind, placebo-controlled ous sample selection, comparison toms in healthy persons who were not trials spontaneously reported adverse groups, more sophisticated analytic taking any medicines. Thirty-nine per- side effects. In an earlier survey of 67 methods, and standardized assess- cent reported fatigue, 26% difficulty con- placebo-controlled trials, an average of ment tools. centrating, 23% drowsiness, 14% head- 23% of patients taking placebo spon- ache, and 5% dizziness15; only 19% of the taneously reported at least 1 bother- The Incidence and Nature of respondents reported experiencing no some side effect.23 When subjects are Nonspecific Side Effects symptoms in the previous 3 days. In a actively queried about side effects, a In 1995, drug-related adverse effects more recent study, Khosla et al16 found substantially higher incidence (be- and illnesses were estimated to ac- that 73% of 236 healthy volunteers tween 27% and 71%) is found.24-27 count for $76.6 billion in hospital costs who were not taking any medications In placebo-controlled trials for dis- and 17 million emergency depart- reported symptoms in the preceding eases that are largely asymptomatic, the ment visits in the United States.5 Most 3 days. The most common were fa- incidence of nocebo side effects may studies have focused on the incidence tigue, headache, difficulty concentrat- equal or even exceed the incidence of side of serious side effects among hospital- ing, and somnolence. effects reported by patients taking the ac- ized patients and little attention has Thus, when a patient starts taking a tive drug. Thus, in trials of antihyper- been devoted to nonspecific side ef- new medication, there is already a large tensive medications and agents to treat fects in ambulatory settings.6-9 In gen- reservoir of bodily symptoms avail- cerebrovascular insufficiency, side- eral, only a small fraction of such side able for misattribution by the patient effect rates among those taking placebo effects are reported,10,11 due in part to to the medication. This misattribution are comparable to those reported tak- uncertainty as to whether the symp- is more likely to occur in: (1) patients ing an active drug,17,28-30 and headache toms were definitely caused by the who expect to experience side effects; in particular is more common among medication. In one study of drugs com- (2) patients who have been previously those taking placebo.17,28 Many com- monly prescribed in primary care prac- conditioned to experience side ef- monly reported nocebo symptoms are tice, 10.9% of reported adverse effects fects; (3) patients who have particular generalized and diffuse such as drowsi- were clearly attributable to the medi- psychological characteristics; and (4) ness, nausea, fatigue, and insomnia.13 In cation, 68.7% were judged to be prob- certain circumstances and conditions. summarizing a large number of studies, ably