EFTA00603108.pdf

OPEN a ACCESS Freely available online PI oS one Clinical Trial Placebos without Deception: A Randomized Controlled Trial in Irritable Bowel Syndrome Ted J. Kaptchuk''2•, Elizabeth Friedlander', John M. Kelley3'4, M. Norma Sanchez', Efi Kokkotou', Joyce P. Singer2, Magda Kowalczykowskil , Franklin G. Millers, Irving Kirsch's, Anthony J. Lembo' 1Beth Israel Deaconess Medical Center. Harvard Medical School, Boston, Massachusetts. United States of America, 2 Osher Research Center, Harvard Medical School, Boston, Massachusetts. United States of America, 3 Psychology Department, Endicott College, Beverly, Massachusetts, United States of America, 4 Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, United States of America, 5 Department of Bioethks, National Institutes of Health, Bethesda, Maryland, United States of America, 6 Department of Psychology, UnNersity of Hull, Hull, United Kingdom able physiological changes in response to placebo treatment that Abstract could explain how placebos alter symptoms. [2] A critical question is establishing how physicians mid other providers can take Background: Placebo treatment can significantly influ- optimal advantage of placebo effects consistent with their ence subjective symptoms. However, it is widely believed responsibility to faster patient trust and obtain informed consent. that response to placebo requires concealment or Directly harnessing placebo effects in a clinical setting has been deception. We tested whether open-label placebo (non- problematic because of a widespread belief that beneficial deceptive and non-concealed administration) is superior to a no-treatment control with matched patient-provider responses to placebo treatment require concealment or deception. interactions in the treatment of irritable bowel syndrome [3] This belief creates an ethical conundrum: to be beneficial in (IBS). clinical practice placebos require deception but this violates the ethical principles of respect for patient autonomy and informed Methods: Two-group, randomized, controlled three week consent. In the clinical setting, prevalent ethical norms emphasize trial (August 2009-April 2010) conducted at a single that "the use of a placebo without the patient's knowledge may academic center, involving 80 primarily female (70%) undermine trust, compromise the patient-physician relationship, patients, mean age 47:18 with IBS diagnosed by Rome III and result in medical harm to the patient." [4] Nevertheless, a criteria and with a score -.-2150 on the IBS Symptom recent national survey of internists and rheumatologists in the US Severity Scale (IBS-SSS). Patients were randomized to found that while only small numbers of US physicians surrepti- either open-label placebo pills presented as "placebo pills tiously use inert placebo pills and injections, approximately 50% made of an inert substance, like sugar pills, that have prescribe medications that they consider to have no specific effect been shown in clinical studies to produce significant on patients' conditions and arc used solely as placebos (sometimes improvement in IBS symptoms through mind-body self- called "impure placebos.") [5] Many other studies confirm this healing processes" or no-treatment controls with the same quality of interaction with providers. The primary finding. [61 Given this situation, finding effective means of outcome was IBS Global Improvement Scale (IBS-GIS). harnessing placebo responses in clinical practice without deception Secondary measures were IBS Symptom Severity Scale is a high priority. (IBS-555), IBS Adequate Relief (IBS-AR) and IBS Quality of Irritable bowel syndrome (IBS) is one of the top 10 reasons for Life (IBS-QoL). seeking primary• care and with a world-wide prevalence of Findings: Open-label placebo produced significantly higher mean (=SD) global improvement scores (IBS-GIS) Citation: Kaptchuk Ti. Friedlander E, Kelley RA Sanchez MN, Kokkotou E, et al. at both 11-day midpoint (5.2±1.0 vs. 4.0:1.1, p-(.001) (2010) Placebos without Deception: A Randomized Controlled Trial in Irritable Bowel Syndrome. PLoS ONE 5(12): e15591. am:D.137i tioumai.pone.00issei and at 21-day endpoint (5.0:1.5 vs. 3.9±13, p=.002). Significant results were also observed at both time points Editor: Isabelle Borman, University Paris Descartes, France for reduced symptom severity (IBS-SSS, p=.008 and Received August 24, 2010; Accepted November 13, 2010: Published p=.03) and adequate relief (IBS-AR, p=.02 and p = .03); December 22, 2010 and a trend favoring open-label placebo was observed for This is an open-access ankle distributed under the terms of the Creative quality of life (IBS-QoL) at the 21-day endpoint (p= .08). Commons Public Domain declaration which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful pianos.. Conclusion: Placebos administered without deception may be an effective treatment for IBS. Further research is Funding: This study was partially supported by grant K24 AT004095, R01 warranted in IBS, and perhaps other conditions, to ATCO402-01 and RCHAT0044562 from National Center for Complementary and Alternative Medione.NIH and in pan from a gift from The Bernard Osher elucidate whether physicians can benefit patients using Foundation. The opinions expressed by the authors are the:, views alone and do placebos consistent with informed consent. not reflect the official views or policy of the National Center for Complementary and Alternative Medicine, National Institutes of Health, Public Health Service or Trial Registration: ClinicalTrials.gov NCT01010191 the US. Department of Health and Human Services. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: AJL has worked as a consultant for Ironwood, GSK. Salix, Introduction Alkermes, and Ardelyx. These companies have had no relationship to this study. Al ocher authors report no competing interest or appearance of competing Placebo treatment can have a significant impact on subjective interest. complaints. [I] Furthermore, recent studies have shown measur- • E-mail ted_kaptchuk0hmsturvard.edu re, PLoS ONE I www.pbsone.org 1 December 2010 I Volume 5 I Issue 12 I e15591 EFTA00603108  Placebos without Deception approximately 10 to 15%. [7,8] It is a chronic functional temlogist (AJL) or a nurse practitioner (BF) experienced hi gastrointestinal disorder characterized by abdominal pain and functional bowel disorders. Patients were excluded if they had discomfort associated with altered bowel habits. [9] The symptoms any unexplained alarm features (i.e. weight loss >10% body of IBS not only adversely affect a person's health-related quality of weight, fevers, or blood in stools, or had family history of colon life (QOL), [10,11] but arc associated with a substantial financial cancer, or inflammatory bowel disease). Patients with a history of burden of reduced work productivity and an over 50% increase in pelvic floor dyssynergia, the need to use manual maneuvers hi the use of health-related resources. [11,12] While many therapies order to achieve a bowel movement, surgery of the colon at any arc commonly used to treat individual IBS symptoms such as time, abdominal surgery within 60 days prior to entry into the constipation or diarrhea, few therapies have been shown to be study, or laxative abuse were excluded from the study. Patients effective and safe in relieving the global symptoms of IBS. [11,13] with other medical conditions (e.g., neurological disorders, Previous research has demonstrated that placebo responses in IBS metabolic disorders, or other significant disease), or pretreatment are substantial and clinically significant. [14,15] Furthermore, data laboratory or ECG findings believed to impair their ability to from our previous qualitative study of IBS patients being treated participate in the study were also excluded. Any surgery within the single-blind with placebos indicated that patients can tolerate a past 30 days, pregnancy, breast-feeding, or participation in high degree of ambiguity and uncertainty about placebo treatment another clinical study within 30 days prior to the start of the and still benefit. [16[ In view of these considerations, we selected study were also disqualifying factors. IBS as a suitable condition to test the widespread belief that Patients were allowed t0 continue IBS medications (e.g., fiber, placebo responses are neutralized by awareness or knowledge that anti-spasmodics, lopemmide, etc.) as long as they had been on the treatment is a placebo. stable doses for at least 30 days prior to entering the study and The objectives of this study were to assess the feasibility of agreed not to change medications or dosages during the trial. recruiting IBS patients to participate in a trial of open-label Patients were asked to refrain from making any major life-style placebo and to assess whether an open-label placebo pill with a changes (e.g., starting a new diet or changing their exercise persuasive rationale was more effective than no-treatment in pattern) during the study. relieving symptoms of IBS ill the setting of matched patient- provider interactions. Interventions Patients were randomly assigned either to open-label placebo Methods treatment or to the no-treatment control. Prior to randomization, patients from both groups met either a physician (AJI..) or nurse- Design practitioner (BF) and were asked whether they had heard of the A three week randomized controlled trial (RCT) comparing "placebo effect." Assignment was determined by practitioner open-label placebo to no-treatment controls was conducted availability. The provider clearly explained that the placebo pill between August 2009 and April 2010 in a single academic was an inactive (.e., "inert") substance like a sugar pill that medical center. Written informed consent was obtained from each contained no medication and then explained in an approximately patient prior to participation on the study. Beth Israel fifteen tninute a priori script the following "four discussion points:" Deaconess Medical Center Institutional Review Board approved 1) the placebo effect is powerful, 2) the body can automatically the design and informed consent. respond to taking placebo pills like Pavlov's dogs who salivated Patients who gave informed consent and fulfilled the inclusion when they heard a bell, 3) a positive attitude helps but is not and exclusion criteria were randomized into two groups: 1) necessary, and 4) taking the pills faithfully is critical. Patients were placebo pill twice daily or 2) no-treatment. Before randomization told that half would be assigned to an open-label placebo group and during the screening, the placebo pills were truthfully and the other half to a no-treatment control group. Our rationale described as inert or inactive pills, like sugar pills, without any had a positive framing with the aim of optimizing placebo medication in it. Additionally, patients were told that "placebo response. It was emphasized that each group was critical for the pills, something like sugar pills, have been shown in rigorous trial. All patients were told that they would receive educational clinical testing to produce significant mind-body self-healing recommendations for their IBS at the end of the study. After processes." The patient-provider relationship and contact time completion of the physical examination and assessments, patients was similar in both groups. Study visits occurred at baseline (Day were then randomized using a sequentially numbered opaque 1), midpoint (Day II) and completion (Day 21). Assessment sealed envelopes that contained treatment assigmnents drawn questionnaires were completed by patients with the assistance of a from a computer-generated random number sequence. Until this blinded assessor at study visits. (The protocol for this trial and point, the patient-provider interaction — including delivering the supporting CONSORT checklist are available as supporting persuasive rationale and the explanation of the importance of both information; see Checklist SI and Protocol SI.) groups was similar for all participants. At this point, during the last moments of the interview, they were told their assignments. Patients Patients randomized to the open-label placebo group were given a Participants were recruited from advertisements for "a novel typical prescription medicine bottle of placebo pills with a label mind-body management study of IBS" in newspapers and fliers clearly marked "placebo pills" "take 2 pills twice daily." and from referrals from healthcare professionals. During the placebo pills were blue and maroon gelatin capsules filled with telephone screening, potential enrollees were told that participants avicel, a common inert excipient for pharmaceuticals (Bird's Hill would receive "either placebo (inert) pills, which were like sugar Pharmacy, Needham, MA). Patients in the no treatment awn were pills which had been shown to have self-healing properties" or no- reminded of the hnportance of the control ann. All visits were hi treatment. Participants were adults (a 18 years old) meeting the the context of a warm supportive patient-practitioner relationship. Rome III criteria for IBS [17] with a score of ≥150 on the IBS The midpoint II day visit was brief (approximately 15 minutes) Symptom Severity Scale (IBS-SSS). [18] The diagnosis of IBS was and included an opened question regarding adverse events, based on typical symptoms and exclusion of patients with alarm concomitant medications and a brief physical examination. After symptoms. [19,20] was confirmed by a board certified ga.stmen- the examination, a treatment-blind researcher administered rc., PLoS ONE I www.pbsone.org 2 December 2010 I Volume 5 I Issue 12 I e15591 EFTA00603109  Placebos without Deception questionnaires. Patients receiving placebos received a short A pill count was taken at visits two and three. Given the reminder regarding the "four discussion points." In the no unprecedented nature of the study, at the completion of the trial treatment arm, patients were encouraged and thanked for helping patients were given a short qualitative open-ended check-out make a successful study. questionnaire and asked for written responses. The questions were Before the study began the providers practiced the trial different for each group. Thaw in the placebo treatment ann were procedures on simulated and real patients. Once a month, the asked four questions: What do you think of about the idea of two providers (Ail., Eli and a third researcher (13K) met to taking placebo? Did you expect it to work or were you skeptical? discuss fidelity to the protocol and any other problems. AJL and What did you think was in the placebo pills? Any further El.- consistently reported that they had no problem holding the comments? Those in the no-treatment were asked three questions: entire initial interview process to approximately 30 minutes and Were you disappointed to be in the treatment as usual ann? What the mid-point to 15 minutes. did you like mast and least about the trial? Any further comments? All assessments were performed by a researcher who was blind to treatment assignment. Assessment Our primary outcome measure was the IBS Global Improve- Intim Scale (IBS-GIS) which asks participants: "Compared to the Statistical Analysis way you felt before you entered the study, have your IBS All tests were two-tailed with alpha set at .05. All results are symptoms over the past 7 days been: 1) Substantially Worse, 2) reported as mean ±SD unless otherwise noted. All analyses were Moderately Worse, 3) Slightly Worse, 4) No Change, 5) Slightly intent-to-treat, and missing data were replaced using the last Improved, 6) Moderately Improved or 7) Substantially Im- observation carried forward method. Since IBS-GIS and IBS-AR are change scores and arc not assessed at baseline, we carried proved.[21,22] Other measures included: the IBS-SSS measure, forward scores for patients who had at least one follow-up visit. For which contains foe 100-point scales, that assess the severity of our main outcome measure (IBS-GIS at 21-day endpoint), we abdominal pain, the frequency of abdominal pain, the severity of planned an independent samples t-test. We estimated a pinto that a abdominal distention, dissatisfaction with bowel habits, and total sample size of 80 would provide 94% power to detect a large interference with quality of life, [18] All 5 components contribute effect (d = .8) and 60% power to detect a medium effect (d = .5). For to the score equally yielding a theoretical range of 0-500, with a IBS-SSS and IBS-QO1., we computed change scores from baseline higher score indicating greater symptom severity. The IBS- and then conducted independent samples t-tests. We used chi Adequate Relief (IBS-AR) is a single dichotomous (yes or no) square tests of independence for IBSAR. Per protocol analyses item that asks participants "Over the past week have you had were also conducted, but they produced no substantive differences adequate relief of your IBS symptoms?" [23] The IBS-QoL is a from our planned intent-to-treat analyses and are not reported here. 34-item measure assessing the degree to which IBS interferes with patient quality of life. Each item is rated on a 5-point Liken scale Results and a linear transformation yields a summed score with a theoretical range of 0 to 100, with a higher score indicating better As shown in Figure 1,92 patients were screened, and 80 eligible quality of life. [24] Side effects were recorded at each assessment. patients were randomized into the two arms (43 into no-treatment Indmduals Assessed for (N-92) Did Not Meet Inclusion Criteria (N-I2) Randomized (h-110) No Treatment Control OpemLabel Placebo (N-43) (N-37) Did Not Complete (N-I) Did Not Complete (N-6) Scheduling Conflict (N-I) Scheduling Conflict (N—I) Drop Out (N-3) Drop Out (Nat) Advase Event (N—I) No Treatment Control Open-Label Placebo Completers Completer' (N-30) (N-31) Figure 1. Enrollment Flowchart. doi:10.1371/joumalpone.0015591.9001 ICs PLoS ONE I www.plosone.org 3 December 2010 I Volume 5 I Issue 12 I e15591 EFTA00603110  Placebos without Deception and 37 into open-label placebo). There were missing outcome data as usual arm?") To answer these questions two researchers (UK, for 13 patients at midpoint (16%; 6 no-treatment control, 7 open- MK) independently extracted the responses to these questions. A label placebo), and for 10 patients at endpoint (13%; 4 no- third researcher gps) compared these extracted responses and a treatment control, 6 open-label placebo). As noted above, missing discussion settled two occasions where handwriting that was data was replaced using the last observation carried forward difficult to interpret. TJK categorized the data using the iterative method. Table 1 shows baseline data. and emergent methodology of grounded theory. [25,26] When As shown in Figure 2 and Table 2, patients treated with open- participants in the placebo arm were asked: " What did you think label placebo had significantly greater scores than the no- was in the placebo pills?" of the 29 who responded, 16 wrote treatment control on the main outcome measure, Global "sugar" (12), "flour" (3) or "calcium" (I)," 6 responded Improvement Scale (IBS-GIS), at both the II -day midpoint "nothing," 5 responded "did not know," 1 responded "symbolic (5.2±1.0 vs. 4.0±1.I, /K.001, d= 1.14) and the 21-day endpoint reminder," and I responded "possible test medication." When (5.0±1.5 vs. 3.9±1.3, y=.002, d= 0.79). In addition, there were participants in the no-treatment ann were asked: "Were you statistically significant differences at both time points on reduction disappointed to be in the treatment as usual arm?" of the 38 who on in symptom severity (IBS-SSS) and adequate relief (IBS-AR), responded, 29 said "no" and only 9 said "yes" or "a little". We and a trend toward significance at the 2I-day endpoint on then looked at the responses of the nine who expressed improvement in quality of life (IBS-QO1.). disappointment, to see how they responded to: "What did you Forty-three patients saw the male physician for all three visits, like most and least about the trial?" All gave uniformly positive 20 patients saw the female nurse-practitioner for all three visits, answers such as "I liked that my feeling about the intensity of the and 17 patients saw a combination of the two or missed a problem was validated and was taken seriously...and was able to treatment session. Given that the two treatment providers differed discuss my IBS," "the doctor and the nurse were wonderful and by gender and discipline (MD vs. NP), we tested for differences in accommodating," "I liked the one-on-one attention with the MD, treatment outcomes. No significant differences were found able to ask questions about IBS with a person trained in the illness; between providers on the primary outcome measure, IBS-GIS this MI) is yen. kind" (underling in the original). This qualitative (p = .57 at midpoint, and p = .51 at endpoint). Similarly, there data seemed to indicate that, in general, patients understood they were no significant differences between providers on any of the were taking placebo and were not overly disappointed in being in secondary outcome measures. the no-treatment arm. Adverse events were reported by only three placebo-treated patients (8%) at midpoint and five patients (14%) at endpoint. The Discussion most common adverse events that patients reported were upper respiratory infection (N = 3) and pain (N = 2); other events We found that patients given open-label placebo in the context included rash, runny stools, and a sty on the eye. of a supportive patient-practitioner relationship and a persuasive The detailed results of the qualitative check-out questionnaire rationale had clinically meaningful symptom improvement that will be reported elsewhere. However, responses to two questions was significantly better than a no-treatment control group with seemed especially relevant to the interpretation of this quantita- matched patient-provider interaction. To our knowledge, this is tive report. Specifically, I) did patients in the open-label arm the first RCP comparing open-label placebo to a no-treatment understand that they were taking a placebo ("What did you think control. Previous studies of the effects of open-label placebo was in the placebo pills?") and 2) were patients in the no treatment treatment either failed to include no-treatment controls [27] or arm disappointed ("Were you disappointed to be in the treatment combined it with active drug treatment. [28] Our study suggests Table 1. Demographics and Baseline Characteristics. No Treatment Open Placebo Demographics end Baseline Characteristics (N=43) (N=37) Age 46118 47-118 Female - no. (r...6) 32 (74) 24 (65) White - no. (16) 36 (84) 26 (70) IBS Type - no. CM Diarrhea Predominant 16 (37) 10 (27) Constipation Predominant 14 (33) 16 (43) Mixed 13 (30) 11 (30) Ns Duration in Years 13111 16112 Symptom Severity (IBS-SSS) 297-1-58 310-s82 Quality of Life 0BSCIOL) 59121 55121 Upper GI Symptoms (GERD & Dyspepsia) - no. (96) 18 (42) 11 (30) Taking Medications for IBS - no. (%) IS (35) 20 (M) Taking Antidepressants - no. (96) 7 (16) 9 (24) Note: All values are means L-SD. rats otherwise noted. Group differences were examined using Independent t-tests for continuous measures and chi square test for categorical measures_ IBS = Irritable bowel syndrome; IBS-5S5 = IBS Symptom Severity Scale; IBSQ0L = IBS Quality of Life Scale; GI = Gastrointestinal; GERD = Gastroesophageal Reflux Disease. dok10.137ujournalpone.00t5S91.t00l f6, PLoS ONE I www.plosone.org 4 December 2010 I Volume 5 I Issue 12 I e15591 EFTA00603111  Placebos without Deception II 6 120 L'2 in ..7-: g 100 ..? za __ le, 80 a p. 0t .0 C ,./‘ ,.., . 4, w T k g 4 4 c 20 E cr. 0 No Treatment Open Placebo No Treairnent Ora. n Pla.eho C la ;‘, 80 15 1 rA 70 4 60 I ! —I 13 it io it50 x ,..0 i 3° - Z 5 . I 20 . c .t e 1 aP 10 -, 01 0 I a- 0 0 NoTatatat Opo•Placebo Noncom:al Diu Play he Figure 2. Outcomes at the 21-Day Endpoint by Treatment Group. doi:l 0.1371/journal.pone.00 I S591.9032 that openly described inert interventions when delivered with a harnessing placebo effects without deception is possible in the plausible rationale can produce placebo responses reflecting context of I) an accurate description of what is known about symptomatic improvements without deception or concealment. placebo effects, 2) encouragement to suspend disbelief, 3) Our results challenge "the conventional wisdom" that placebo instructions that foster a positive but realistic expectancy, and 4) effects require "intentional ignorance." [29] Our data suggest that directions to adhere to the medical ritual of pill taking. It is likely Table 2. Treatment Outcomes. No Treabnent Open Placebo (N=431 (N=37) p-value Midpoint 411 Days) Global Improvement 1185-6151 4.0-11.1 52±1.0 ‹.001 Adequate Relief 085-AR) - no. PH 104231 18 (49) .02 Symptom Severity Reduction (185-555) 28166 75187 .008 Quality of Life Improvement (1135-Q0Q 4A±8.9 8.3-111.6 .10 Endpoint (3 Weeks) Global Improvement (I85-051 3.9-113 513±1.5 .002 Adequate Relief 0B5-AR1 - no. I%l IS 1351 22 1591 .03 Symptom Severity Reduction (1B5-555) 46174 92±99 .03 Quality of Life Improvement (1B5-Q01-.) 54±13.8 11.4-116.6 .08 Nott All values are means ±SD except where noted. IBS = Irritable bowel syndrome; 18S-GIS = 185 Global Improvement Scale; IBS-AR = IBS Adequate Relief; 185-555 = NS Symptom Severity Scale; I8S-QoL = IBS Quality of Life Scale. dot10.137Moumalecae.00l5591.t002 I PLoS ONE I www.plosone.org 5 December 2010 I Volume 5 I Issue 12 I e15591 EFTA00603112  Placebos without Deception our study also benefited from ongoing media attention giving There are intimations in the placebo literature that providers credence to powerful placebo effects. with greater perceived expertise or authority (e.g., physician versus Both treatment arms were given in a context of a warm patient- nurse, dentist versus technician) will elicit greater placebo provider relationship. It is possible that this relationship had a responses. [36,37] In our study, we found no evidence for positive benefit for the patients, and indeed, the no-treatment arm significant differences between male physician and female nurse- showed improvement. Given that patients in both treatment arms practitioner. experienced the same frequency and duration of contact time and In addition to its clinical significance, our study has important the content of the interaction was way similar, we believe that the ethical implications. As mentioned above, evidence indicates that incremental improvement in our open-label arm was due to the physicians continue to use placebo treatment without transparent addition of open-label placebo treatment. The magnitude of disclosure to patients [5,6] Our results suggest that the placebo improvement reported by those on open-label placebo treatment response is not necessarily neutralized when placebos arc was not only statistically significant but also clinically meaningful. administered openly. 'Thus our study points to a potential novel The effect size for the primary outcome, calculated as the strategy that might allow the ethical use of placebos consistent with standardized mean difference (d) between the open-label-placebo evidence-based medicine. Minimally, open-label placelm may and no-treatment groups, was 0.79 at endpoint, which is have potential as a "wait and watch" strategy before prescriptions conventionally interpreted as a large effect. [30] At endpoint, drugs are prescribed. Further studies of open placebo arc merited we also observed medium sized effects for the differences between not only for IBS but for illnesses primarily diagnosed by subjective placebo and control groups on symptom severity (d=0.53) and symptoms and introspective self-appraisal. In sum, our study quality of life (d =0.40). An improvement from baseline of 50 suggests that for sonic disorders it may be appropriate for points on the IBS-SSS reliably indicates meaningful symptomatic clinicians to recommend that patients try an inexpensive and safe improvement. [18] The open-label group improved by 92 points placebo accompanied by careful monitoring before and after on this measure; in addition, the improvement shown by the prescribing medication. Clearly replication and further research is open-label placebo group exceeded that shown by the no- essential before such a practice could be implemented. treatment group by 46 points. Similarly, an increase of 10 points on the IBS-Qol, indicates a clinically meaningful improvement, Limitations and we observed an increase of 11 points on this measure for the This RCT has several limitations. Most importantly, our sample open-label group. [24] Finally, the percentage of patients size was relatively small and the trial duration was too short to reporting adequate relief during the preceding 7 days at the obtain estimates of long-term effects. Therefore, the trial could br 21-day endpoint (59%) is comparable with the responder rates in described as a "proof-of-principle" pilot study. Obviously, clinical trials of drugs currently used in IBS. [31,32] A recent replication with a larger sample size and a longer follow-up is meta-analysis of double-blind, placebo-controlled trials of needed before clear clinical decisions could be made based on our alosetron in IBS estimated that 51% of patients treated with data. alosetron had adequate relief as compared to 38% of patients Other potential limitations of our study may be the issue of treated with placebo. p31 Our results were remarkably similar report bias (e.g., "wishing to please the experiments?). However, (59% for open-placebo; 35% for no-treatment control), suggest- given the impossibility of double-blind assessment of open placebo ing that open-label placebo in the context of a persuasive versus no-treatment control, the effects of report bias cannot be rationale may show comparable efficacy to established IBS eliminated. Another related limitation is that patients assigned to treatments. no-treatment may have been disappointed, thus inflating the The placebo response in this trial (59% on IBS-AR) was differences between open-label placebo and no-treatment control substantially higher than typical reported placebo responses of 30 groups. Importantly, our qualitative check-out data found the no- 40% in double-blind IBS pharmaceutical studies. [15] This finding treatment group experiencing positive support, with 76% of them seems counterintuitive. We speculate that it is an indication of the reporting that they were not disappointed with their assignment. credibility of our open-label rationale. Patients in our study This argues against disappointment being a significant factor. A accepted that they were receiving an active treatment, albeit not a further possible limitation is that our results are not generalizable pharmacological one, whereas patients in double-blind trials because our trial may have selectively attracted IBS patients who understand that they have only a 50% chance of receiving active were attracted by an advertisement for "a novel mind-body" treatment. It may be that one hundred percent certainty that one intervention. Obviously, we cannot rule out this possibility. is receiving the "treatment of interest" (in this case open-label However, selective attraction to the advertised treatment is a placebo) is more placebogenic than a fifty percent probability of possibility in virtually all clinical trials. In any case, patients in receiving an inactive control. clinical practice arc ultimately given choices and it may turn out It may be worthwhile to interpreted our study in light of the that open-label placebo will be helpful only for those who elect to 2001 landmark meta-analysis of placebo effects and its 2010 try this option. Finally, it could be argued that IBS is a poor illness expanded and updated version. [34,35] In the recent analysis, the to study placebo effects because it lacks objective measures. authors found 202 randomized trials in 60 medical conditions that However, there arc many serious conditions for which primary included placebo and no-treatment groups. When meta-analyti- outcomes arc primarily subjective (e.g. depression, anxiety and cally combined, in general, little evidence of clinically meaningful chronic pain), and the preponderance of evidence indicates that effects of placebo beyond no treatment was found. 'Ilse meta- placebo treatments arc most effective for such patient-centered analysis, however demonstrated a significantly larger placebo complaints. [1] effect for a subset of 28 studies with a specific aim of investigating In summary, our study suggests that patients are willing to take the placebo effect. Perhaps this subset is most relevant to our study open-label placebos and that such a treatment may have which was also specifically examined placebo effects. Further salubrious effects. Further research is warranted in IBS and prospective research will be necessary to clarify under what perhaps other illnesses to confirm that placebo treatments can be circumstances and in what conditions one can expect or not expect beneficial when provided openly and to determine the best to find robust placebo responses. methods for administering such treatments. PLoS ONE I www.plosone.org 6 December 2010 I Volume 5 I Issue 12 I e15591 EFTA00603113  Placebos without Deception Supporting Information Author Contributions Checklist SI Conceived and designed the experiments: TJK JMK EK DC;NI IK (DOC) Performed the experiments: AIL. El, JNIK MNS JPS MK. Analyzed the data: JNIK AJL IK MK JPS. Wrote the paper: TJK AJLJMK KAI Protocol SI 1K EK. (DOC) References I. Miller EG, (lollesca I.. Kaptchuk 13 (1009) llw placebo effect illness and 19. Hammer J. Eslick Cl), Howell SC. Ahiparmak E. Talky NJ (20041 Diagnostic interpersonal healing. Perspectives at Biology and Medicine 52:518 39. ykki of alann features in irritable fumy) syndrome and functional dyspepsia. Cut 2. Danis, Kaptchuk '13. Miller F, Benrikiti I' 10103 Plant* effects: 53: 666 71. hiological, clinical and ethical advances. lancet 375: 686 95. 20. Vanner SJ. Ikpew WP. Paterson WG (21MHE Predictive value of the Rome 3. Miller EG, 4:ollessa 1. (1009)'11w legitimacy of placebo treatments in clinical criteria liar diagnosing the imitable bowel syndrome and functional dyspepsia. pracrice: evidence and ethics. Amtritan.jmunal M Binethicx Ik 39 47.