Review of: Eisenberg DM, Davis RB, Ettner SL, et al. Trends in alternative medicine use in the United States. JAMA 1998; 280 (18): 1569–1575; and Eisenberg DM, Kessler RC, Foster C, et al. Unconventional medicine in the United States. NEJM 1993; 328(4): 246–252.
The current article challenges the findings and conclusions of the above-mentioned publications with Dr. David Eisenberg as a first author. The JAMA article is a follow-up to the earlier article. Both of these studies reveal bias and a lack of adequate statistical and scholarly analysis which compromise their value for the medical community. In spite of their flaws and weaknesses, they are significant contributions to the ongoing debate regarding the relative value of alternative medicine (AM) and Western scientific medicine (WSM). To better understand the scientific and intellectual parameters of this debate, we will discuss the historical precedents that underlie this debate and the problems it poses for the medical community and medical consumers.
The 1993 NEJM article reported results from a 1990 telephone survey of 1539 respondents which had been designed to assess their use of health care practices in America1. One-third (34%) of the persons interviewed had used unconventional therapy during the prior year and 10% had seen providers for therapy. The majority (83%) of those who had used AM methods also had treatment from a conventional medical doctor.
In 1997, 2055 randomly selected subjects were questioned with comparable key questions. Forty-two percent of persons surveyed had used at least one of sixteen alternative therapies during the prior year. The authors’ conclusion was that “Alternative Medicine use and expenditures increased substantially between 1990 and 1997. . . .”2 The therapies that increased the most during the period 1990–1997 were herbal medicine, massage, megavitamins, self-help groups, talk energies, energy healing, and homeopathy. Total 1997 out-of-pocket expenditures for alternative therapies were estimated to be 27 billion dollars, which the article claims is about the same amount as that projected for “all U.S. physician services.” These findings led to the conclusion that U.S. federal agencies, academic institutions, and private foundations should become more aggressive in promoting AM and in funding related scientific research, educational curricula, and quality control monitoring.
The studies were partly funded by the Fetzer Institute. Its founder, John Fetzer “privately explored interests in the unseen elements of life, meditation, prayer, philosophy and other ways of healing.”3 The founder’s value system seems to be consistent with the mystical belief system of Dr. David Eisenberg, whose popular book, Encounters with Qi,4 was inspired by his immersion in Chinese ancient medicine.
An editorial in a 1998 NEJM observed that “there is much confusion about the definition of Alternative Medicine.”5 The Eisenberg group’s 1993 and 1998 publications define alternative medicine as “interventions neither taught widely in medical schools nor available in medical schools not available in U.S. hospitals.” is followed a few lines later by the sentence “The majority of U.S. hospitals.” Medical schools and hospitals were the two national institutions most intimately associated with medical legitimacy in the United States during the twentieth century, and provided the lion’s share of medical care.
If the findings are accurate regarding AM’s popularity and financial clout, the modern medical establishment is being challenged in a parallel way to that which occurred in the country during the first decade of the twentieth century. According to the historian Paul Starr, a struggle for power and authority took place between mainstream scientific medicine and the precursors of contemporary alternative medicine, including Thomsonians, homeopaths, osteopaths and chiropractors.6 The competition surfaced just at the time that the groundbreaking Flexner report, Medical Education in the United States and Canada (1910) appeared.7 Aided by the financial support of the Carnegie Foundation, Flexner toured the most important medical teaching institutions in the United States and Canada. He recorded detailed descriptions of dirty, poorly equipped laboratories, inappropriate use of medication, the absence of explicit and universal standards of education, and rampant commercialism. The Flexner report called for massive reform, including the establishment of licensing regulations, clear requirements for graduation from medical schools, and the closing of the worst, most incompetently run schools and hospitals. The report became a clarion call for changes in medicine in North America. National political leaders supported and financed its demand for reform.
There ensued a battle between scientific medicine and the promoters of AM for prestige, power and resources. History has shown that the medical personnel educated according to scientific standards attained increasing authority, legitimacy, and political influence. Later in the twentieth century they also became the dominant providers of medical services. The same trend occurred in all developed countries. Physicians were strong advocates for medical practice and education based on the findings of anatomy, biology, physics, physiology, and mathematics.
Now, as just suggested, if Eisenberg et al.’s JAMA survey of the period between 1990–1997 is credible, AM is gaining in a race for financial resources and popular prominence at the end of the century.
However, the 1998 Eisenberg survey is riddled with ambiguities and other defects. The first paragraph begins with an internal contradiction. The definition of alternative medicine as “interventions not taught widely in U.S. medical schools now offer courses on alternative medicine.” is followed a few lines later by the sentence “The majority of U.S. medical schools now offer courses on alternative medicine.” This claim is supported by reference to another 1998 article of which Eisenberg was a second author.8 This publication surveyed 94% of U.S. medical schools and found that the majority of them did indeed include a course on AM. A critical reader might conclude that the criterion “widely taught” is not met by survey results (most schools have one short, elective course), were it not for the main inference the authors draw from both papers: AM is immensely important for the education of all medical students. The authors use their data in ways that suggest they are indirectly lobbying for increased funding for AM to expand its prominence in medical school curricula.
The criterion that AM’s methods are absent from U.S. hospitals is also questionable, since many of the AMA publication’s “medical” therapies such as prayer, vitamins, and massage, while not usually considered to be primary interventions, certainly are amply represented in most standard medical institutions.
The 1998 article has much more serious flaws. Most of the treatment techniques claimed as belonging to AM’s domain, and included in the estimate of national expenditures for its use, have little or nothing to do with medicine. Unlike the NEJM paper of 1993, the 1998 paper distinguished between “more alternative” and “less alternative” approaches. The first category includes homeopathic medicine, folk remedies, and energy healing, while the second includes biofeedback, relaxation, and hypnosis. While the latter group’s methods are considered more acceptable in the realm of standard practice, all of the sixteen “principal therapies” considered, varying from “guided imagery” (4.5%) to chiropractic (11%) to herbal medicine (12%), have effects mainly through suggestion, making the distinction irrelevant. Likewise neglected is the fact that the efficacy of neither group has been proven. Furthermore, there was not even confirmation that any medical illness actually existed in the patients questioned by phone.
Gorski has pointed out that over half of the users of AM in the 1993 study employed it without supervision or even contact with a medical professional.9 He also noted that both the 1993 and 1998 papers failed to make a distinction between AM used exclusively and AM as a complement to medical care. So a patient’s use of a vitamin C tablet to ward off a possible coryza could have the same strength in the data analysis as the use of colonic irrigation or the strapping on of magnets to the abdomen for the treatment of a biopsy-confirmed bowel carcinoma. More useful information would have been the percent of AM users with serious illness who deprived themselves of medical care.
In 1998, Astin reported that of the 54% of a sample of over 1000 patients who reported that they always had been “highly satisfied” with their medical practitioners, 39% used alternative therapies.10 About the same number (40%) of the total group who turned to AM for help did so because of negative attitudes toward scientific biomedicine, so animosity toward traditional medicine did not predict higher AM use. Also, only 4.4% of this randomly selected group of a thousand patients relied solely on AM and would be suitable for inclusion for a comparison with scientific medicine. Using this more restricted definition casts doubt on AM’s alleged large economic influence, and deflates the Eisenberg group’s claim that more financing is needed for AM patient utilization, research, and education.
The paper’s economic analysis, which shows that out-of-pocket expenses for AM and standard medicine during 1997 were about the same, is likewise a questionable conclusion.11 This anaylsis defended sample size in terms of percentage change between 1990 and 1997, but not the change in percent of utilization which pertains to the hypothesis of main interest. The analysis and survey as a whole used ratio weights, rather than the more appropriate multivariate analysis, to control for factors responsible for change. Important variables such as immigration status, insurance coverage, and perception of success of conventional medical treatments were not included. Any of these factors could have had a major influence on the outcome results. Most recent immigrants come from countries with non-Western medical traditions. Also, the uninsured and those unhappy or unfamiliar with biomedical care may be more likely to utilize alternative methods.
The survey noted that AM generated higher out-of-pocket costs, but this comparison does not include the avoided costs of not using expensive conventional treatment for which there is no insurance. Finally, the large expenses of the pharmaceutical companies and those of the NIH and other research institutes that fund biomedical research were not included in the comparison.
The flaws and ambiguities in both Eisenberg et al. articles raise doubts about the validity of their findings. These defects and the fact that one of the sponsoring foundations is devoted to funding research investigating metaphysical healing cast doubt on the investigators’ objectivity. Their methods suggest an unwitting bias toward establishing a bandwagon effect through magnifying the economic importance of AM. On the other hand, that the Journal of the American Medical Association and the New England Journal of Medicine publish such flawed studies indicates that AM has indeed attained U.S. prominence and become a cultural phenomenon. Whether AM will have any benefits for our patients remains an untested and unanswered question.
- Eisenberg DM, Kessler RC, Foster C, et al. Unconventional medicine in the United States. NEJM. 1993; 328(4): 246–252.
- Eisenberg DM, Davis RB, Ettner SL, et al. Original contribution, trends in alternative medicine use in the United States. JAMA. 1998; 280 (18): 1569–1575.
- John E. Fetzer Institute, Dec. 9, 1999, www.fetzer.org.
- Eisenberg D. Encounters with Qi. New York: W. W. Norton; 1995.
- Holland JC. Editorial. Use of alternative medicine—a marker for distress? NEJM. 1998; 340(22): 1758–1759.
- Starr P. The Social Transformation of American Medicine. New York: Basic Books; 1982.
- Flexner A. Medical Education in the United States and Canada. Carnegie Foundation for Advancement of Teaching; 1910.
- Wetzel MS, Eisenberg DM, Kaptchuk TJ. Courses involving alternative medicine at US medical schools. JAMA. 1998; 280(9): 784–787.
- Gorski T. Do the Eisenberg data hold up? Sci Rev Alt Med. 1999; 3(2): 62–69.
- Astin JA. Why patients use alternative medicine. JAMA. 1998; 279(19): 1548–1553.
- Berry B. Commentary on Eisenberg article. Personal communication.