|Home | About | Journals | Submit | Contact Us | Français|
Previous studies have shown that acupressure bands can reduce chemotherapy-related nausea. Patients' expectations of efficacy account for part of this outcome. We conducted a three-arm randomized clinical trial to investigate the effectiveness of acupressure bands in controlling radiation therapy-induced nausea and to test whether an informational manipulation designed to increase expectation of efficacy would enhance the effectiveness of the acupressure bands. Patients who experienced nausea at prior treatments were randomized to either standard care (Arm 1, n = 29) or standard care plus acupressure bands with either neutral (Arm 2, n = 30) or positive (Arm 3, n = 29) information regarding the efficacy of the bands. Patients reported nausea for two days prior to randomization (baseline) and for five days following using a 7-point semantic rating scale (1 = not nauseated to 7 = extremely nauseated). Patients in Arms 2 and 3 combined reported greater reduction in average nausea than patients in Arm 1 (P = 0.01; meanbands = 0.70, meanno bands = 0.10). This equates to a 23.8% decrease in nausea in the band groups compared to a 4.8% decrease in the control group, a 19% difference. The informational manipulation failed to alter efficacy expectations and there was no statistically significant difference in nausea between patients in Arms 2 and 3. Acupressure bands are an effective, low- cost, non-intrusive, well-accepted, and safe adjunct to standard antiemetic medication. An attempt to boost the efficacy of the acupressure bands by providing positive information was not successful.
Nausea remains one of the most troublesome and frequent side effects associated with cancer treatment.1-6 Researchers and clinicians are increasingly aware of the role of expectation as a critical psychological factor that influences symptom development, and many studies have reported that patients' expectancies for the development of nausea following chemotherapy are a significant predictor of its subsequent occurrence.7-15
While the bulk of the research studies examining the relationship between nausea expectancies and subsequent nausea development have been observational, some have used experimental designs and provide limited evidence that such expectancies, may, in fact, be causal.11,16-18 Our research group has conducted two large multicenter intervention studies designed to reduce nausea by reducing nausea expectancies. The first, conducted in 355 patients about to begin chemotherapy, examined whether a modest educational intervention designed to reduce patients' nausea expectancies by dispelling misconceptions about chemotherapy-related nausea and building confidence in the efficacy of their antiemetic drug regimen would result in less nausea.19 While the expectancy manipulation significantly reduced the expectation for nausea in the intervention group compared to the control group, no changes were observed between the intervention and control groups in average nausea following treatment.
The second study (n =739) used acupressure and acustimulation wristbands as a means of changing patients' nausea expectancies.18 Prior to chemotherapy patients were randomly assigned to: 1) acupressure bands (Sea-Band®, Sea-Band International, England), 2) an acustimulation band (Reliefband®), or 3) a no band control condition. Both types of bands were designed to stimulate the P6 acupuncture point, which is on the inside of the wrist and is the most commonly used point for nausea relief. In that study, patients who received the acupressure wrist bands and expected them to be effective in nausea control experienced significantly less nausea than all other patients in the study, including those patients who received acustimulation bands and those patients who received the acupressure bands but did not expect them to be effective. Interestingly, patients receiving the acupressure bands who did not expect them to be effective did not differ from the control group or the acustimulation group in reported nausea. These findings suggest that expectation of efficacy for the acupressure bands was a significant determinant of actual efficacy.
While it is not clear how stimulation of the P6 acupuncture point reduces nausea, several reviews have concluded that the practice does appear to provide relief for a significant proportion of patients, particularly in reducing nausea associated with anesthesia and pregnancy.20-24 The most recent of these was a systematic review and meta-analysis of acu-point stimulation for the treatment of chemotherapy-induced nausea and vomiting.24 In that review, the three acupressure studies (including our study discussed above) were analyzed as a group, and the findings were inconclusive. The authors noted a significant decrease in acute, i.e., on the day of treatment, but not delayed nausea in the acupressure band condition compared to the no band controls, with no effect on vomiting. In addition, the authors cautioned that they could not determine whether the reduction of acute nausea severity was a true finding, a function of expectancy, or an artifact related to the well-documented desire of subjects in unblinded studies to give answers they believe experimenters want.
Physiologic effects of acupressure are poorly understood, making it difficult to distinguish the relative contribution of the acupressure manipulation and the role of expectation in changing the patient's perception of nausea. This difficulty is exemplified by the results of an experiment assessing the efficacy of acupressure in reducing nausea and vomiting (NV) associated with outpatient surgery.25 Ninety participants were randomly assigned to receive standard treatment, standard treatment plus acupressure, or standard treatment plus sham acupressure (acupressure applied to a point on the body not predicted by acupuncture models to affect the sensation of nausea). Both the true acupressure and sham acupressure groups reported significantly less nausea than the control group, Ps < 0.01 (10% of the treatment group, 20% of the sham acupressure group, and 50% of the control group reporting symptoms, overall P < 0.001). A similarly designed 3-arm acupressure band study of 60 patients undergoing minor outpatient gynecological surgery also reported positive results for placebo bands. Patients randomized to both the active acupressure band condition and to the placebo band condition had significantly less postoperative nausea than patients randomized to the no band control group with no differences between the two acupressure band groups.26
The accumulated prior research provides evidence that acupressure bands are effective in reducing nausea and that their effectiveness, at least in part, can be accounted for by patients' expectations of efficacy (i.e., a placebo effect). The present study extends these earlier findings and involves a three-arm randomized clinical trial examining whether an informational manipulation designed to increase patients' expectations of efficacy for the acupressure bands results in decreased nausea from radiation treatments for cancer. We choose radiation patients for this study, even though our prior work with nausea expectancies was entirely with chemotherapy patients, as a means of examining the relevance of such expectancies in a different patient population. To our knowledge, this is the first study of acupressure bands for the control of nausea during radiation therapy for cancer. Our primary hypothesis was that patients receiving information about the acupressure wrist bands that was designed to increase expectations the bands would be helpful would expect less treatment-related nausea, which, in turn, would result in less treatment-related nausea. Additional hypotheses were that acupressure wrist bands, regardless of type of accompanying information, would reduce nausea and vomiting and improve patient quality of life (QOL).
Patients with any cancer diagnosis were eligible for this study if they had received at least two radiation treatments prior to study entry, had experienced nausea or vomiting subsequent to at least one of those prior treatments, and were scheduled to receive at least seven additional daily radiation treatments. Upon providing signed consent, patients were asked to familiarize themselves with the acupressure bands by wearing them for one minute. Patients were told that we were studying the efficacy of acupressure bands and how information was delivered. At no point did we inform patients we were examining the role of expectancy and that we were testing an expectancy manipulation. The acupressure wrist bands used were Sea-Bands® (Figure 1), a commercially available elastic wrist band with a small plastic button used to apply pressure. After wearing the bands for one minute and returning them to the research coordinator, study participants completed a QOL measure and an on-study questionnaire that included items related to expected efficacy of the acupressure bands. Patients were also provided with a take-home diary to record the severity of their NV for the next two treatments as a baseline assessment. They were not provided with acupressure bands at this time.
After returning the two-day baseline NV diary, patients were randomized by means of previously randomized sealed envelopes to either standard care (Arm 1), standard care plus acupressure bands plus a handout conveying information not intended to enhance efficacy expectations (Arm 2), or standard care plus acupressure bands plus information designed to enhance efficacy expectations (Arm 3). Patients in Arms 2 and 3 reviewed the handouts and completed a second set of expectancy questions. The post-randomization phase of the study lasted for five treatments. Patients recorded NV in a five-day diary and completed a QOL measure before going to bed after the fifth treatment. Patients randomized to Arms 2 and 3 were told that the bands may be worn or removed throughout the five days at their discretion and that they may choose to wear only one band if they so desired. All ancillary treatments, including antiemetic medications, as appropriate for control of symptoms caused by the cancer or its treatment, were allowed for all patients and were not standardized. The protocol was approved by the Institutional Review Board of our medical center.
The information given to patients in trial Arm 3 to enhance expected acupressure band efficacy consisted of data presented as full color graphs from the two positive acupressure band studies with supporting captions and an explanatory statement. We provided a very positive and very limited (although quite truthful) interpretation of the data to these patients. For example, the caption under the first graph was: “Only 3 patients wearing Seabands had nausea compared to 15 patients not wearing Seabands.” Our description of the second study included the phrase: “Patients wearing Seabands had less severe nausea on the day they received their chemotherapy than patients not wearing Seabands.” Arm 2 patients, by contrast, received information on only the second of these two studies in a grey-scale graph. The information we gave these patients was also truthful, but we gave the data a less positive interpretation. The corresponding description given the Arm 2 patients concerning this study was: “Although patients wearing Seabands in a prior study appeared to have less severe nausea compared to patients not wearing the Seabands, our analysis suggested that this was probably due only to a placebo effect.”
Nausea and emesis were measured by a patient report diary developed by Burish and Carey.27,28 Each day of radiation treatments was divided into 4 sections: morning, afternoon, evening, and night. Patients reported severity of nausea and number of vomiting episodes for each period daily. Severity of nausea was assessed on a 7-point rating scale, anchored at one end by 1 = “Not at all nauseated” and at the other end by 7 = “Extremely nauseated.” The description “Moderately nauseated” was centered on the scale below the 4. Patients were given the questionnaires to complete at home. The eight nausea reports in the baseline period were averaged to create a single baseline score and the 20 nausea reports following the intervention were averaged for the primary outcome variable. We also assessed the amount of antiemetic medication taken and the number of vomiting episodes.
The measure of patients' expectation for nausea is based on a questionnaire used previously by our group.10,18,19 Nausea expectancy was assessed on separate 5-point Likert-scales that are anchored at one end by 1 = “I am certain I WILL NOT have this,” and at the other end by 5 = “I am certain I WILL have this.” Expected efficacy of the wrist bands was assessed on a 5-point scale anchored at one end by 1 = “Not at all effective” and at the other end by 5 = “Very effective.” A question assessing whether this expectation was changed by the information materials provided was on similar scale anchored at one end by 1 = “More likely to be effective” and at the other end by 5 = “Less likely to be effective.” The phrase “Did not change my thoughts” was centered under the three.
Quality of life was assessed using the Functional Assessment of Cancer Therapy Scale – General (FACT–G). The FACT–G is a 28-item scale developed specifically for use in cancer clinical trials.29 This measure was developed through extensive interviews with patients experiencing symptoms of cancer and oncology professionals. The FACT-G has been validated with cancer patients and has shown very good test-retest reliability as well as validity.30,31
Demographic data and detailed information about other medications administered during the seven-day study period, i.e., two baseline days and five intervention days, were abstracted from medical records by study personnel. A feedback questionnaire was completed by patients at the conclusion of the study period concerning use of and recommendations for the bands.
The primary outcome variable for this study was change in average nausea from baseline to post-treatment. Analysis of variance (ANOVA) with a significance level of P ≤ 0.05 was used to compare the change in nausea severity among the three groups. Secondary analyses involved investigating the efficacy of Seabands in reducing nausea, and if they were effective, to determine if that effectiveness was related to expected efficacy. T-tests, ANOVA, and correlational analyses, as appropriate, were used for these analyses. Methods similar to those used for the primary analyses were employed in additional analyses examining the intervention effects on changes in three additional study outcomes, i.e., QOL, the number of antiemetic pills taken, and occurrence of vomiting. Antiemetic dose and type, other than a simple pill count, was not assessed or controlled for in the analyses.
Eighty-eight (89%) of the 99 patients randomized provided evaluable data. Four of the original 99 patients were excluded in the analyses because of changes in medications during the seven-day study period that greatly affected their report of nausea (two received chemotherapy, one received zoledronic acid, and one had substantial changes in psychotropic medications to control anxiety). One patient was removed from the study immediately after randomization because he wanted the acupressure bands despite randomization to the control condition. Two patients who provided consent died before taking part in the intervention phase of the study and another was hospitalized and unable to complete the study measures. Three patients simply failed to return the post-intervention measures. Three of the remaining 88 patients did not complete all of the post-intervention measures but did complete at least two days of the nausea diary. Nausea averages were calculated on just the data provided by these patients and they were included in the analyses.
Of these 88 patients, 42 were males, 69 were Caucasian, 14 were African American, three were Hispanic and two were Asian. These patients ranged in age from 20-85 years, with a mean of 52.9. Most of the patients (64.8%) were married. Forty-one percent had graduated from or attended college, and an additional 47% were high school graduates. Twenty-nine patients were randomized to Arm 1 (standard care) with an equal number randomized to Arm 3. Arm 2 had 30 patients. All study subjects received their treatments as outpatients. Forty-eight of the 88 patients took some type of antiemetic medication during the two-day baseline period. An examination of between-group differences at baseline revealed that the three study arms differed significantly on their baseline average nausea, P = 0.01, meanArm 1 = 2.1, meanArm 2 = 2.8 meanArm 3 = 3.1. This important clinical difference between groups is handled statistically by the use of change scores in all analyses.
The analysis of variance of the change scores (baseline minus post) in average nausea severity approached but did not reach conventional levels of statistical significance, P = .08. The mean change score for the three groups were: Arm 1 = 0.10, Arm 2 = 0.65, Arm 3 = 0.75 (See Figure 2. Note: A positive score indicates a decrease in nausea).
Additional ANOVAs examining the efficacy of the intervention showed no significant differences among the three groups in the change scores of QOL, P = 0.93, in changes in the daily amount of antiemetic medication taken, P = 0.35, nor in changes in the occurrence of vomiting, P = 0.23 (Table 1).
Overall, 59 evaluable patients received the acupressure bands (Arms 2 and 3), and 29 (Arm 1) did not. T-tests revealed that patients receiving the bands had a significantly greater reduction in average nausea than those who did not (meanbands = 0.70, meanno bands = 0.10; P = 0.01). This equates to a 23.8% decrease in nausea in the band groups compared to a 4.8% decrease in the control group, a 19% difference. Further analyses examining the effect of bands versus no bands found no significant differences in change scores of QOL, P = 0.93, in changes in the daily amount of antiemetic medication taken, P = 0.78, or in changes in the occurrence of vomiting, P = 0.17.
No significant difference in expected efficacy of the acupressure bands after receiving the informational intervention were observed in a t-test comparing Arm 2 and Arm 3 patients (meanneutral = 3.17, meanenhanced = 3.31; P = 0.56). Similarly, the two groups did not differ on whether this expectation was changed by the information materials provided (meanneutral = 2.66, meanenhanced = 2.67; P = 0.96). The expected efficacy of the acupressure bands in the 59 patients receiving the bands was not significantly correlated with change in nausea (r = -0.07; P = 0.61) or change in occurrence of vomiting (Spearman's rho = -0.08; P = 0.56).
The initial measure of expected nausea, taken at the time of consent, was correlated significantly with average nausea during the two-day baseline period (r = 0.21; P = 0.05), and the correlation between this initial expectancy measure and average nausea in the five-day intervention period closely approached statistical significance (r = 0.205; P = 0.055). Overall, the second assessment of expected nausea made just following randomization and receipt of the informational intervention was significantly lower that the first, with a mean decrease of 0.25 (P =0.04, from 4.10 to 3.85) for the 88 patients. The difference between the three groups in change in expected nausea was not statistically significant, P = 0.30. Unlike the first assessment of expected nausea, the second assessment did not correlate significantly with average nausea during either of the two assessment periods, both P-values > 0.81. Similarly, the change score in expected nausea did not correlate significantly with change in nausea in either the 29 patients in the control group or in the 59 patients receiving the acupressure bands, (both, r < 0.02, P > 0.88).
Overall satisfaction with the acupressure bands appeared to be high, with 36 of the 57 patients wearing the band for over 48 hours (the longest of the five “duration of time worn” categories on the feedback questionnaire) and an additional 10 patients wearing the bands at least 24 hours. Patient satisfaction with the bands as assessed by the feedback question asking whether they would recommend that other patients wear bands when receiving radiation therapy was also positive. The mean response on this 5-point scale (anchored at one end by 1 = “Strongly do not recommend” and at the other end by 5 = “Highly recommend”) was 3.7.
The findings did not support the primary hypothesis that acupressure band with efficacy-enhancing information would reduce radiation therapy-induced nausea as compared with acupressure band with neutral information concerning band efficacy. The analysis examining this was an ANOVA of the change in nausea from baseline among the three study arms (the above two conditions along with the no band control group). There was no statistically significant difference among the group mean change scores. Analyses examining the differences between the three groups in change scores of QOL, changes in the daily amount of antiemetic medication taken, and changes in the occurrence of vomiting, were all insignificant.
On the other hand, we found that acupressure wrist bands, regardless of type of accompanying information, reduced radiation-induced nausea. Patients receiving the bands had a significantly greater reduction in average nausea compared to those who did not, equating to a 23.8% decrease in nausea in the combined band groups compared to a 4.8% decrease in the control group. This is a 19% decrease in nausea and is slightly better than the 15 -17% change in nausea set as a clinically significant improvement in many NCI-sponsored multicenter nausea trials.18,19,32 No differences between the patients receiving the acupressure bands compared to the control group were observed in study outcomes related to QOL, antiemetic medication, or vomiting.
The principal hypothesis of the study was that an informational intervention designed to positively affect patients' expectations of the effectiveness of their nausea control regimen by providing positive information regarding the acupressure bands would be helpful in controlling nausea. Our experiment to empirically test this hypothesis compared levels of nausea in patients receiving such positive information (Arm 3) compared to patients receiving neutral information (Arm 2). The failure of the study to support this hypothesis was likely due to the fact that our informational manipulation was not successful in changing these effectiveness expectancies. We observed no differences in expected efficacy in the 29 patients receiving the bands with expectancy enhancing information and the 30 patients receiving the bands with the neutral information. We asked patients if their expectation of band efficacy was changed by the information provided. There was virtually no difference between groups (meanneutral = 2.66, meanenhanced = 2.67). The fact that both of these means were below the midpoint on a 5-point scale that had “Did not change my thoughts” centered under the three, suggests that the informational intervention in both groups was largely ineffective in changing efficacy expectations regarding the acupressure bands.
The failure of our intervention to change either efficacy or nausea expectations may be related, at least in part, to our choice of patient sample. An important eligibility criterion for our study was that patients must have pre-existing nausea, and it is very possible that because nausea expectancies in this group were based upon actual experience, they may have been very difficult to overcome, especially by the relatively modest manipulation used in the present study. Future research could address this issue by using cancer treatment naïve patients and/or stronger expectancies manipulations.
In this study, the benefit from wearing the acupressure bands was not related to patients' expected efficacy, in sharp contrast to the findings from the multicenter study described earlier that found expected efficacy of the acupressure bands in reducing chemotherapy-related nausea to be significantly related to actual efficacy.18 It is possible that the discrepancy in results is due to the different patient populations. In the present study, patients were receiving radiation therapy and all had nausea from prior radiation treatments. By contrast, patients in the multicenter study had not yet received treatment and all received chemotherapy. These may be important factors in understanding the relationship between expectancies related to nausea development. In particular, as stated above, the nausea expectancies related to an event that an individual has not yet personally experienced, (e.g., receiving chemotherapy for the first time) may be more malleable than expectancies related to outcomes from events that the patient has experienced before, e.g., repeated radiation treatments.
Considering that patients in the present study all had nausea from previous radiation treatments, it can be presumed that they had very strong expectancies of experiencing nausea. It is possible that our acupressure band efficacy expectation measures are not sensitive enough to pick up on relevant changes related to the likelihood of nausea occurring when an individual believes that its likelihood is very high. Evidence to this effect has been previously reported by our group.33 In that report, acupressure bands plus standard care, compared with standard care alone, were effective in reducing nausea in breast cancer patients who had very strong expectancies for experiencing chemotherapy-related nausea. But, as in the present study, we found that while the bands were helpful, the effectiveness of the bands in controlling nausea was not related to responses on the expected efficacy measure.
It is possible that the acupressure bands were effective in our study for reasons not related to expectancy. One possibility is that the acupressure bands are indeed efficacious, as suggested by a growing body of literature showing that stimulation of acupuncture points can indeed reduce nausea.24 It has, after all, been a staple of Chinese medicine for centuries.34 While our data supports this interpretation, we are reluctant to accept it at face value because of other reports in the literature, including our own, linking the efficacy of acupressure bands to expectancy/placebo effects.18,25,26 We are also not aware of any physiological mechanisms by which stimulation of an acupuncture point on the wrist can affect development of nausea.
Limitations of the study include the possibility that our positive findings are merely an artifact due to report bias on the patients' part or some other non-specific factor related to the study that was not adequately controlled for because the study was not blinded. This lack of blinding concerning the band versus non-band conditions is a clear limitation of the study, as is the fact we did not control for or assess the size, site, and dose per fraction of the radiation fields patients received and other things patients may have done to control their nausea. It is also possible that our findings are a statistical artifact related to the fact that despite randomization, patients receiving the acupressure bands had higher levels of nausea than the control group patients during the baseline period. Other than mere chance and the fact that we have a relatively small sample size, we have no explanation why this difference at baseline occurred. We used change scores from pre- to post-intervention in all of our analyses as a method of controlling for this difference in baseline nausea but this method, as with all statistical methods to control for an unequal randomization in a small study, is imperfect. Results of these analyses should therefore be interpreted cautiously. Future studies should control for these limitations, if possible.
The present study supports previous experimental research showing that acupressure bands are effective in controlling nausea and extends these positive findings to the population of patients experiencing radiation therapy-related nausea. It is noteworthy that prior to entering the study, all patients experienced nausea and that this nausea was not completely controlled by their physician-provided antiemetic regimen. Nausea for cancer treatments, once it occurs, is often very difficult to control and yet these simple bands were effective in that effort. It is a low-cost, non-intrusive, safe adjunct to standard antiemetics that has great patient acceptance.
This work was supported by Grants RSG 01071-010PBP and MRSG-04-233-01-CPPB from the American Cancer Society and 1R25-CA102618-01A1 from the National Cancer Institute.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errorsmaybe discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.