Alternative Medicine Research in Clinical Practice: A US National Survey

doi:10.1001/archinternmed.2009.49

Arch Intern Med. Author manuscript; available in PMC 2010 January 11.

Published in final edited form as:

Arch Intern Med. 2009 April 13; 169(7): 670–677.

doi: 10.1001/archinternmed.2009.49

PMCID: PMC2804465

NIHMSID: NIHMS165972

Alternative Medicine Research in Clinical Practice

A US National Survey

Jon C. Tilburt, MD, MPH, Farr A. Curlin, MD, Ted J. Kaptchuk, Brian Clarridge, PhD, Dragana Bolcic-Jankovic, MA, Ezekiel J. Emanuel, MD, PhD, and Franklin G. Miller, PhD

Author Affiliations: Department of Bioethics, National Institutes of Health, Bethesda, Maryland (Drs Tilburt and Miller); Division of General Internal Medicine and the Program in Professionalism and Bioethics, Mayo Clinic, Rochester, Minnesota (Dr Tilburt); Section of Internal Medicine and the McClean Center for Clinical Medical Ethics, University of Chicago, Chicago, Illinois (Dr Curlin); Osher Research Center, Harvard Medical School, Boston, Massachusetts (Mr Kaptchuk); and Center for Survey Research, University of Massachusetts, Boston (Dr Clarridge and Ms Bolcic-Jankovic).

Correspondence: Jon C. Tilburt, MD, MPH, Division of General Internal Medicine, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (Email: tilburt.jon/at/mayo.edu)

The publisher's final edited version of this article is available at Arch Intern Med

See other articles in PMC that cite the published article.

Abstract

Background

Little is known about whether federally funded complementary and alternative medicine (CAM) research is translating into clinical practice. We sought to describe the awareness of CAM clinical trials, the ability to interpret research results, the acceptance of research evidence, and the predictors of trial awareness among US clinicians.

Methods

We conducted a cross-sectional mailed survey of 2400 practicing US acupuncturists, naturopaths, internists, and rheumatologists.

Results

A total of 1561 clinicians (65%) responded. Of the respondents, 59% were aware of at least 1 major CAM clinical trial; only 23% were aware of both trials. A minority of acupuncturists (20%), naturopaths (25%), internists (17%), and rheumatologists (33%) were “very confident” in interpreting research results (P<.001). Fewer acupuncturists (17%) and naturopaths (24%) than internists (58%) and rheumatologists (74%) rated the results of randomized controlled trials as “very useful” (P<.001). Twice as many internists (53%) and rheumatologists (64%) rated patient preferences as “least important” compared with acupuncturists (27%) and naturopaths (31%) (P<.001). In multivariate analyses, for clinicians aware of at least 1 trial, male sex (odds ratio [OR], 1.30 [95% confidence interval {CI}, 1.05–1.62]), prior research experience (OR, 1.45 [95% CI, 1.13–1.86]), institutional or academic practice setting (ORs, 1.98 [95% CI, 1.01–3.91], and 1.23 [95% CI, 0.73–2.09], respectively), and rating randomized trials as “very useful” (OR, 1.46 [95% CI, 1.12–1.91]) (P<.001) for clinical decision making were positively associated with CAM trial awareness. Acupuncturists, naturopaths, and internists (ORs, 0.15 [95% CI, 0.10–0.23], 0.15 [95% CI, 0.09–0.24], and 0.18 [95% CI, 0.12–0.28], respectively) were all similarly less aware of CAM trial results compared with rheumatologists.

Conclusion

For clinical research in CAM to achieve its social value, concerted efforts must be undertaken to train clinicians and improve the dissemination of research results.

Complementary And Alternative medicines (CAM) are widely used, but until recently there have been few rigorous empirical studies of the safety and efficacy of CAM treatments. In response, the National Institutes of Health (NIH) has invested more than US $2 billion over the past decade to support scientific research on CAM.¹^,² For this investment to achieve its anticipated social value,³^,⁴ clinical research must be translated into improvements in clinical and public health practice—a process fraught with obstacles.⁵^,⁶

Prior beliefs and attitudes of clinicians toward CAM and CAM research may limit the extent to which they translate new empirical data regarding CAM into their clinical practice.⁷^–⁹ Clinicians’ practice philosophies, their attitudes regarding the efficacy of CAM therapies, their experience with those therapies, and their judgments regarding the reliability of different information sources all appear to shape their views of CAM.¹⁰^–²¹ There is also evidence that CAM clinicians differ from mainstream clinicians in their attitudes toward the role of research results in clinical decision making.¹³^,²²^,²³ Still, other factors such as prior training, practice context, and professional norms may all influence the extent to which evidence from CAM research influences clinical practice.

For evidence from clinical research to have an impact on medical practice, health care professionals must first be aware of the research. Once aware, health care professionals must be able to interpret these findings, judging both their validity and their implications. Finally, they must apply the scientific evidence to their own practices.⁵ While not directly measuring the clinical impact of CAM research, this study sought to address the following questions regarding the translation of CAM research into the clinical practices of conventional and alternative health care professionals: (1) To what extent are clinicians aware of findings from prominent CAM trials? (2) Do clinicians feel prepared to interpret research results? And (3) how highly do clinician audiences for CAM research value research evidence relative to other factors in clinical decision making? To clarify the obstacles that may hinder CAM research evidence from influencing clinical practice, this study also explored the characteristics, experiences, and attitudes of clinicians that are associated with being aware of CAM research.

METHODS

PARTICIPANTS

We mailed a self-administered, 12-page questionnaire to stratified random samples of 1200 practicing US physicians (600 general internists and 600 rheumatologists) and 1200 licensed US CAM providers (600 acupuncturists and 600 naturopaths). Between May 20 and June 1, 2007, each sampled health care professional received a cover letter, an informational sheet, a letter of support from a relevant professional society, and the questionnaire. After 6 weeks, a second packet was sent to nonrespondents. A $20 incentive accompanied the initial mailing. Among the 2400 health care professionals sampled, 1561 responded (65% response rate): 440 acupuncturists (73%), 442 naturopaths (74%), 334 general internists (56%), and 345 rheumatologists (58%). Among acupuncturists and naturopaths, respondents and nonrespondents did not differ by age or practice setting, but respondents were slightly more likely to report female sex (P =.04). Among rheumatology and internal medicine samples, respondents did not differ from nonrespondents by age, practice setting, or sex.

Physician samples were obtained from the American Medical Association Physician Masterfile—a comprehensive database of US physicians. The CAM provider samples were obtained from national databases of licensed acupuncturists and naturopaths compiled from state boards. Acupuncturists were selected because the NIH funds many clinical trials in acupuncture that might bear on the practices of these CAM providers. Similarly, naturopaths routinely recommend herbs, vitamins, and other dietary supplements commonly studied in NIH trials. At the time the sample was drawn (fall 2006), 13 states licensed naturopathic physicians and 44 states licensed acupuncturists.

Eligibility was defined as holding an active license for the designated profession, current involvement in patient care, and ability to read English. After verifying as many addresses as possible, we sent questionnaires to a random list of 600 eligible health care professionals (without replacement) from each professional group.

QUESTIONNAIRE DEVELOPMENT

We devised measures of health care professional knowledge, attitudes toward and experiences of CAM, and attitudes toward research evidence as follows: (1) We compiled key theoretical constructs from the literature and explored those constructs in focus groups using a semistructured moderator’s guide devised and implemented by an experienced social scientist and coinvestigator (B.C.). (2) A convenience sample of health care professionals participated in two 90-minute focus groups—one for alternative health care professionals (4 naturopaths and 5 acupuncturists) and one for conventional health care professionals (2 rheumatologists and 4 general internists). (3) Videotaped group interactions and observer notes were analyzed, and themes were used to revise a draft survey instrument. (4) Eight additional health care professionals (2 internists, 2 rheumatologists, 3 acupuncturists, and 1 naturopath) completed the revised survey instrument and participated in a 90-minute, face-to-face cognitive interview to further clarify the questionnaire. (5) We then revised the survey based on these interviews. (6) Finally, we conducted a pretest of 40 health care professionals (10 from each respective professional group) to get rough estimates of group response (21 responded), to test usability, to check initial distributions, and to obtain participant comments that might indicate problems.

The final questionnaire included the following 5 domains: (1) participant characteristics, (2) experience with research and CAM therapies, (3) awareness of CAM trials, (4) acceptance of research evidence, and (5) factors that influence judgments about hypothetical studies.

DEPENDENT VARIABLE

Because awareness is an absolutely necessary first step in translating research evidence into improved health outcomes,⁵ awareness of CAM trials was our primary dependent variable. Rather than asking general questions about familiarity with CAM research that may be vague and especially susceptible to social desirability, we asked respondents about their awareness of 2 specific prominent CAM studies. Between 2004 and 2006, 2 landmark, high-profile CAM studies were published that evaluated CAM modalities, which formed the basis for evaluating respondent awareness of CAM research.²⁴^,²⁵ We asked participants to read summaries of the 2 clinical trials that studied the use of CAM modalities for osteoarthritis of the knee. The first study, a randomized controlled trial (RCT) comparing adjunctive acupuncture with sham acupuncture for osteoarthritis of the knee, was published in the Annals of Internal Medicine in 2004.²⁴ That study found that acupuncture was superior to educational control for improving standardized measures of pain and function. The second study, a double-blind RCT of glucosamine with and without chondroitin, was published in the New England Journal of Medicine in 2006.²⁵ That study found no benefit over placebo for the primary outcome. However, in a secondary analysis the investigators found that the glucosamine-chondroitin combination was superior to placebo for participants with moderate to severe disease. With respect to each study, survey participants indicated whether they were familiar with the trial, whether the trials’ findings were consistent with their clinical experience, and whether they had changed their practices in light of the trial’s results. In subsequent analyses, we operationalized awareness of CAM trials in 2 ways: familiarity with at least 1 of the 2 studies or familiarity with both studies.

INDEPENDENT VARIABLES

Respondents’ ability to interpret research results was measured according to their confidence in interpreting research results in general (very confident, moderately confident, not very confident, or not at all confident).

We also measured health care professionals’ acceptance of research evidence in general (not related to CAM research specifically). Clinical experience, research results, and patient preferences are all legitimate considerations for clinical decisions.²⁶ However, to generate an indirect measure of the extent to which health care professionals are inclined to accept new scientific evidence as relevant for their clinical decisions, we asked participants to indicate the overall importance and relative role of each factor in their clinical decisions. Participants also rated the overall clinical usefulness of published results from RCTs.

Respondents reported their demographic information, practice characteristics, and their participation in any of the following 6 research activities: enrolled patients in a clinical study, been a coinvestigator, been a principal investigator, been a reviewer for a journal, designed a study, or served on an institutional review board. To measure respondents’ experiences with CAM, we asked them to indicate whether they had ever recommended each of the following: spinal manipulation (eg, chiropractic), acupuncture, energy medicine (eg, Reiki), meditation practices (eg, yoga), glucosamine ± chondroitin, and body work (eg, massage and Shiatsu). These types of CAM therapies are commonly used for musculoskeletal conditions and correspond to a spectrum of CAM treatment categories as defined by the NIH.²⁷

HUMAN SUBJECTS APPROVAL

This study was approved by the Office of Human Subjects Research of the National Institutes of Health and by the institutional review board of the University of Massachusetts–Boston. A cover letter outlined the voluntary nature of participation; a returned questionnaire was considered documentation of informed consent.

DATA MANAGEMENT AND STATISTICAL ANALYSIS

All data were double entered and 100% verified. In unadjusted bivariate comparisons of professional groups, we performed Pearson χ² tests for categorical variables and unpaired, 2-tailed t tests for continuous variables. Unless otherwise specified, responses to ordered categorical variables were dichotomized based on the distribution characteristics of responses. Then, 2 basic logistic regression models that included independent variables of age, sex, race, geographic region, practice setting, professional group, and research experience and variables indicating whether a clinician had recommended acupuncture or glucosamine were constructed using awareness of 1 CAM trial or awareness of both CAM trials, respectively, as the dependent variable. Then, because individual variables measuring attitudes toward evidence in clinical practice were highly correlated, these variable were tested in the base models individually to determine independent associations. All regression models used Hosmer-Lemeshow goodness-of-fit tests, which tests for low model fit. All analyses were performed with STATA Intercooled 8.0 (Stata-Corp, College Station, Texas) statistical software.

RESULTS

Of the 2400 clinicians who were sent questionnaires, 1561 responded (65%). Professional groups varied considerably in their personal and professional characteristics (Table1). A higher proportion of acupuncturists(28%)reported being Asian. Acupuncturists (63%) and naturopaths (63%) were more likely to be women, and a much higher proportion of rheumatologists reported having an academic practice setting(20%).Acupuncturists and naturopaths were more likely to be from the western region of the United States.

Table 1

Characteristics of 1561 Clinician Respondents, Shown by Professional Group

Experience in conducting research and recommending CAM modalities also varied widely (Table 2). Acupuncturists were the least likely to report personal experience in research (25%), followed by naturopaths (32%), general internists (46%), and rheumatologists (77%). A majority of respondents in all professional groups reported having recommended each of the listed CAM therapies except energy medicine, which few internists or rheumatologists (7% and 7%, respectively) had recommended.

Table 2

Experiences With Research and Recommending CAM Therapies Among 1561 Clinician Respondents

AWARENESS OF CAM TRIALS

There was wide variation in respondent awareness of the 2 presented CAM trials within and across professional groups (Table 3). Among all health care professionals, 37% were aware of the acupuncture trial published in the Annals of Internal Medicine in 2004²⁴ and half (49%) were aware of the glucosamine trial published in the New England Journal of Medicine in 2006.²⁵

Table 3

Awareness of CAM Clinical Trials and Ability to Interpret Research Results Among 1561 Clinician Respondents, Shown by Professional Group

Acupuncturists (46%) and rheumatologists (49%) were more likely to be aware of the acupuncture study than were naturopaths (30%) and general internists (22%). Among those who were aware, a modest proportion of clinicians, ranging from 19% (acupuncturists) to 30% (rheumatologists), reported changing their practice based on the study results among the 4 professional groups.

Internists (59%) and rheumatologists (88%) were much more likely to be aware of the glucosamine study than were acupuncturists (20%) and naturopaths (39%). Among those who were aware of this study, slightly greater proportions of respondents in the professional groups reported changing their practice based on the study results, ranging from 22% to 51%. Overall, 59% of all respondents (88% of rheumatologists, 60% of internists, 47% of naturopaths, and 48% of acupuncturists) were aware of at least 1 of the studies. Only 23% were aware of both trials (45% of rheumatologists, 17% of internists, 20% of naturopaths, and 15% of acupuncturists).

ABILITY TO INTERPRET RESEARCH RESULTS

A minority of respondents in all professional groups expressed the highest level of confidence in their ability to interpret research results; 20% of acupuncturists, 25% of naturopaths, 17% of internists, and 33% of rheumatologists were “very confident” in their ability to critically interpret research literature. A majority of respondents in all groups described themselves as “moderately confident,” including 59% of acupuncturists, 64% of naturopaths, 67% of internists, and 59% of rheumatologists (Table 3).

ACCEPTANCE OF RESEARCH EVIDENCE

With respect to the relative importance of different considerations in clinical decision making(clinical experience, patient preferences, and research results),we found that most respondents in all groups said that clinical experience is “very important” (Table4). Physicians were more than twice as likely as CAM providers to say research was “very important,” while CAM providers were more likely to say that patient preferences were “very important.” Moreover, physicians were 3 times more likely to rate results of RCTs as “very useful” than were CAM providers (most of the latter rated RCTs as “moderately useful”) (Table 4).

Table 4

Relative Importance of Research Evidence in Clinical Decisions Among 1561 Clinician Respondents, Shown by Professional Group

This same pattern emerged in participants’ rankings of the most and least important factors in their decision making. A majority of acupuncturists (79%) and naturopaths (81%) rated clinical experience as “most important,” whereas internists and rheumatologists were more evenly divided between those who rated clinical experience (50% each) and published research (37% and 43%, respectively) as the most important factor. Conversely, two-thirds of acupuncturists and naturopaths rated published research as “least important” (70% and 66%, respectively), while a majority of internists and rheumatologists rated patient preferences as “least important” (53% and 64%, respectively).

After controlling for age, sex, race, census region, and practice setting, being in the rheumatology professional group was independently associated with ratings of the importance of patient preferences (rating of patient preferences as “very important,” odds ratio [OR], 0.59 [95% confidence interval {CI}, 0.41–.85]).A similar pattern emerged with rating patient preferences as “least important.” In that analysis, after controlling for age, sex, race, census region, and practice setting, internal medicine and rheumatology specialties were both independently associated with rating patient preferences as least important(OR,2.76[95%CI,1.93–3.94], and OR, 4.03 [95% CI, 2.76–5.89], respectively).

ASSOCIATIONS WITH CAM TRIAL AWARENESS

In analyses using pooled unadjusted data from the 4 professional groups, we found that those who reported male sex, any research experience, being “very confident” in their ability to interpret research results, or having strong, favorable opinions about the role of research in their practice were all more likely to be aware of CAM trials (Table 5). In addition, awareness of CAM trials varied by professional group, practice setting, and region. For instance, higher proportions of rheumatologists, those with an academic or institutional practice setting, and those from the Northeast or Midwest regions reported being aware of CAM trials. In contrast, lower proportions of acupuncturists, those in solo practice, and those from the West region reported being aware of CAM trials (Table 5).

Table 5

Likelihood of CAM Trial Awareness, Stratified by Clinician Characteristics, Experiences, and Attitudes

Multivariate analyses showed that several characteristics and attitudes were independently associated with greater CAM trial awareness (Table 5). These included male sex (OR, 1.30 [95% CI, 1.05–1.62]), institutional practice setting (OR, 1.98 [95% CI, 1.01–3.91]), having research experience (OR, 1.45 [95% CI, 1.13–1.86]), and greater acceptance of evidence (rating of research as “very important,” OR, 1.40 [95% CI, 1.09–1.80]) (Table 5). Those with moderate or slight confidence in interpreting research results were less likely to be aware of CAM trials (ORs, 0.69 [95% CI, 0.52–0.93] and 0.53 [95% CI, 0.36–0.79], respectively). Acupuncturists, naturopaths, and internists (ORs, 0.15 [95% CI, 0.10–0.23], 0.15 [95% CI, 0.09–0.24], and 0.18 [95% CI, 0.12–0.28], respectively) were much less aware of CAM trial results than rheumatologists. In a second model using awareness of both CAM trials as the dependent variable, most associations remained the same; however, sex was not significantly associated with CAM trial awareness (OR, 1.29 [95% CI, 0.96–1.60]), but practicing in an academic setting was associated with being aware of both CAM trials (OR, 1.96 [95% CI, 1.22–3.15]). The P value of Hosmer-Lemeshow goodness-of-fit test to examine poor fit for the basic regression model (without attitudinal variables) was not significant (P=.44) and remained nonsignificant for subsequent models (including attitudinal variables) (P values ranged from .70 to .98), suggesting reasonable model fit.

COMMENT

In this national survey of 1561 conventional and CAM health care professionals, half said that they were aware of at least 1 of the 2 CAM trials presented, but only 1 in 4 was aware of both studies. Few clinicians were very confident in interpreting research results. Compared with those who were not aware of CAM trials, clinicians who were aware of CAM trials were much more likely to be rheumatologists, to be practicing in an institutional or academic setting, to have some research experience, to express greater ability to interpret evidence, and to report greater acceptance of evidence.

These results suggest that the ultimate clinical impact of clinical research in CAM likely depends on the training, attitudes, and experiences of the clinicians who could translate research results into clinical practice. For instance, CAM providers expressed much less research experience and less regard for trial results in their clinical decision making compared with their conventional colleagues, and they were also less aware of CAM trial results. In contrast, internists reported an intermediate level of research experience and very favorable regard for the role of research evidence in their practice, but they were no more aware of CAM trial results than their CAM counterparts. In contrast, clinicians who believe that they can make sense of research results may pay more attention to those results. Rheumatologists on average spend more time training and typically participate in research and learn skills of research interpretation during fellow-ship training. These experiences may confer greater confidence in interpreting results and may thereby lead to increased trial awareness.

Similarly, health care professionals must believe that scientific research is useful before they will be motivated to learn about it and accept it. Less than one-third of CAM providers indicated the highest level of regard for research results in their clinical decision making compared with two-thirds of conventional health care professionals. These findings may reflect differences in professional culture regarding what constitutes “evidence” and what makes one kind of evidence more dependable than another. Physicians may feel increasingly compelled to say that they practice “evidence-based medicine,” stressing measurable population-based health outcomes. ¹³^,²⁸^,²⁹ Providers of CAM may conceive of “evidence” in broader terms—including their own clinical experience—and place less emphasis on population-based empirical data.²⁸^,²⁹ In turn, it appears that CAM providers may emphasize patient preferences more in their clinical decision making compared with their conventional colleagues.

Despite the differences, all professional groups cite clinical experience as central to their clinical decision making. ³⁰ To have the maximal impact, evidence from clinical research (appropriately synthesized and graded) must be approached with an open mind and integrated with clinical experience and patient preferences.²⁶ If clinicians have entrenched views either in favor of or in opposition to a given therapy, it will be difficult for new evidence to refute such preconceived notions.³¹ Furthermore, scientists must disseminate research evidence in a clinically relevant manner so that clinicians can implement decisions consistent with that evidence.³²

This study has several limitations. First, as a cross-sectional survey it does not permit causal inferences. Second, despite our efforts to elicit candid opinions of respondents, the perceived desire to give socially acceptable responses could have differentially biased health care professionals’ responses. Third, self-reported attitudes and experiences may not reflect actual skills and behaviors and may not accurately reflect these clinicians’ views of CAM research in general, since we focused on just 2 studies. Furthermore, the differences in awareness between the 2 trials we asked about may be due to time delays or other factors not captured by this survey. Exactly whether or how respondents changed their practice in response to a clinical trial is also difficult to ascertain by survey. Finally, these results may not generalize to other conventional or CAM professional specialties. Nevertheless, these data provide an initial examination of the influence CAM research is having on clinical practice and factors that appear to be at play in that process.

In conclusion, data from this large national survey suggest that evidence from CAM research has the potential to make a difference in the practice of a broad range of professional practices, conventional and alternative, so long as health care professionals are aware of that evidence and have the experience, training, and opportunities to apply that evidence in the context of their specific practice. Nevertheless, significant barriers to CAM clinical research awareness exist. For clinical research in CAM (and conventional medicine) to achieve its potential social value, concerted efforts must be undertaken that more deliberately train clinicians in critical appraisal, biostatistics, and use of evidence-based resources, as well as expanded research opportunities, dedicated training experiences, and improved dissemination of research results.

Acknowledgments

Funding/Support: Funding for this research was provided by the National Center for Complementary and Alternative Medicine and the Department of Bioethics, National Institutes of Health. Dr Curlin is supported by grant 1 K23 AT002749 from the National Center for Complementary and Alternative Medicine.

Role of the Sponsors: The funder was not involved in data collection, analysis, or writing of the manuscript.

Footnotes

Author Contributions: Dr Tilburt had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Tilburt, Curlin, Kaptchuk, Clarridge, Emanuel, and Miller. Acquisition of data: Kaptchuk, Clarridge, and Bolcic-Jankovic. Analysis and interpretation of data: Tilburt, Curlin, Kaptchuk, Clarridge, Bolcic-Jankovic, Emanuel, and Miller. Drafting of the manuscript: Tilburt, Kaptchuk, Miller. Critical revision of the manuscript for important intellectual content: Curlin, Kaptchuk, Bolcic-Jankovic, Emanuel, and Miller. Obtained funding: Emanuel. Administrative, technical, and material support: Clarridge and Bolcic-Jankovic. Study supervision: Kaptchuk, Clarridge, and Miller.

Financial Disclosure: None reported.

Disclaimer: The views presented are those of the authors and do not necessarily represent the positions or policies of the National Institutes of Health or the Department of Health and Human Services.

Previous Presentation: This work was presented at the Society of General Internal Medicine Annual Meeting; April 11, 2008; Pittsburgh, Pennsylvania.

Additional Contributions: Development and implementation of the survey was performed by the Center for Survey Research, University of Massachusetts, Boston. Statistical support was provided by Summit Consulting, LLC, Washington, DC. MPA Media provided samples of CAM providers. We are grateful to the American College of Physicians, the National Certification Council for Acupuncture and Oriental Medicine, and the American Association of Naturopathic Physicians for endorsing the study. Kristine Hirschkorn, PhD, provided constructive insights in both reviewing the background literature and in reviewing an earlier draft of the manuscript. Richard Nahin, PhD, Laura Lee Johnson, PhD, Jack Killen, MD, and Margaret Chesney, PhD, provided critical feedback on the study design. This work was possible only with the leadership and commitment of the late Stephen Straus, MD.

REFERENCES

1. Institute of Medicine. Complementary and Alternative Medicine in the United States. Washington, DC: National Academies Press; 2005.

2. National Center for Complementary and Alternative Medicine. Expanding Horizons of Health Care, Strategic Plan 2005–2009. Bethesda, MD: NIH Publications; 2004.

3. Emanuel EJ, Wendler D, Grady C. What makes clinical research ethical? JAMA. 2000;283(20):2701–2711. [PubMed]

4. Miller FG, Emanuel EJ, Rosenstein DL, Straus SE. Ethical issues concerning research in complementary and alternative medicine. JAMA. 2004;291(5):599–604. [PubMed]

5. Glasziou P, Haynes B. The paths from research to improved health outcomes. ACP J Club. 2005;142(2):A8–A10. [PubMed]

6. Sung NS, Crowley WF, Jr, Genel M, et al. Central challenges facing the national clinical research enterprise. JAMA. 2003;289(10):1278–1287. [PubMed]

7. Boyle P. Getting Doctors to Listen: Ethics and Outcomes Data in Context. Washington, DC: Georgetown University Press; 2000.

8. Park CM. Diversity, the individual, and proof of efficacy: complementary and alternative medicine in medical education. Am J Public Health. 2002;92(10):1568–1572. [PubMed]

9. Vickers A. Evidence-based medicine and complementary medicine. ACP J Club. 1999;130(2):A13–A14. [PubMed]

10. Astin JA, Marie A, Pelletier KR, Hansen E, Haskell WL. A review of the incorporation of complementary and alternative medicine by mainstream physicians. Arch Intern Med. 1998;158(21):2303–2310. [PubMed]

11. Berman BM, Bausell RB, Lee WL. Use and referral patterns for 22 complementary and alternative medical therapies by members of the American College of Rheumatology: results of a national survey. Arch Intern Med. 2002;162(7):766–770. [PubMed]

12. Berman BM, Singh BB, Hartnoll SM, Singh BK, Reilly D. Primary care physicians and complementary-alternative medicine: training, attitudes, and practice patterns. J Am Board Fam Pract. 1998;11(4):272–281. [PubMed]

13. Boon H. Canadian naturopathic practitioners: holistic and scientific world views. Soc Sci Med. 1998;46(9):1213–1225. [PubMed]

14. Corbin Winslow L, Shapiro H. Physicians want education about complementary and alternative medicine to enhance communication with their patients. Arch Intern Med. 2002;162(10):1176–1181. [PubMed]

15. Ekins-Daukes S, Helms PJ, Taylor MW, Simpson CR, McLay JS. Paediatric homoeopathy in general practice: where, when and why? Br J Clin Pharmacol. 2005;59(6):743–749. [PubMed]

16. Hadley CM. Complementary medicine and the general practitioner: a survey of general practitioners in the Wellington area. N Z Med J. 1988;101(857):766–768. [PubMed]

17. Hirschkorn KA, Bourgeault IL. Conceptualizing mainstream health care providers’ behaviours in relation to complementary and alternative medicine. Soc Sci Med. 2005;61(1):157–170. [PubMed]

18. Kaczorowski J, Patterson C, Arthur H, Mith KS, Mills DA. Complementary therapy involvement of physicians: implications for practice and learning. Complement Ther Med. 2002;10(3):134–140. [PubMed]

19. Schmidt K, Jacobs PA, Barton A. Cross-cultural differences in GPs’ attitudes towards complementary and alternative medicine: a survey comparing regions of the UK and Germany. Complement Ther Med. 2002;10(3):141–147. [PubMed]

20. Visser GJ, Peters L. Alternative medicine and general practitioners in The Netherlands: towards acceptance and integration. Fam Pract. 1990;7(3):227–232. [PubMed]

21. Wearn AM, Greenfield SM. Access to complementary medicine in general practice: survey in one UK health authority. J R Soc Med. 1998;91(9):465–470. [PMC free article] [PubMed]

22. Hamilton E. Exploring General Practitioners’ attitudes to homeopathy in Dumfries and Galloway. Homeopathy. 2003;92(4):190–194. [PubMed]

23. Novak KL, Chapman GE. Oncologists’ and naturopaths’ nutrition beliefs and practices. Cancer Pract. 2001;9(3):141–146. [PubMed]

24. Berman BM, Lao L, Langenberg P, Lee WL, Gilpin AM, Hochberg MC. Effectiveness of acupuncture as adjunctive therapy in osteoarthritis of the knee: a randomized, controlled trial. Ann Intern Med. 2004;141(12):901–910. [PubMed]

25. Clegg DO, Reda DJ, Harris CL, et al. Glucosamine, chondroitin sulfate, and the two in combination for painful knee osteoarthritis. N Engl J Med. 2006;354(8):795–808. [PubMed]

26. Haynes RB, Devereaux PJ, Guyatt GH. Clinical expertise in the era of evidence-based medicine and patient choice. ACP J Club. 2002;136(2):A11–A14. [PubMed]

27. National Center for Complementary and Alternative Medicine. What is CAM? [Accessed December 30, 2008]. http://nccam.nih.gov/health/whatiscam/

28. Jackson S, Scambler G. Perceptions of evidence-based medicine: traditional acupuncturists in the UK and resistance to biomedical modes of evaluation. Sociol Health Illn. 2007;29(3):412–429. [PubMed]

29. Jagtenberg T, Evans S, Grant A, Howden I, Lewis M, Singer J. Evidence-based medicine and naturopathy. J Altern Complement Med. 2006;12(3):323–328. [PubMed]

30. Friedson E. A Study of the Sociology of Applied Knowledge. Chicago, IL: University of Chicago Press; 1998. pp. 158–184.

31. Kaptchuk TJ. Effect of interpretive bias on research evidence. BMJ. 2003;326(7404):1453–1455. [PMC free article] [PubMed]

32. Smith R. What clinical information do doctors need? BMJ. 1996;313(7064):1062–1068. [PMC free article] [PubMed]