Alternative Medicine
Page 1

AN MJBA-280406. 93014. AU Eisenberg-David-M. Kessler-Ronald-C. Foster-Cindy. Norlock-Frances-E. Calkins-David-R. Delbanco-Thomas-L.

IN From the Division of General Medicine and Primary Care, Department of Medicine, Beth Israel Hospital and Harvard Medical School, Boston (D.M.E., T.L.D.). The Institute for Social Research, University of Michigan, Ann Arbor (R.C.K., C.F.). The Division of General Medicine, Department of Medicine, New England Deaconess Hospital and Harvard Medical School, Boston (D.R.C.). Chicago College for Osteopathic Medicine, Chicago (F.E.N.). Address reprint requests to Dr. Eisenberg at the Division of General Medicine and Primary Care, Beth Israel Hospital, 330 Brookline Ave., Boston, MA 02215. Supported by a contract with the John E. Fetzer Institute and a grant from the Nathan Cummings Foundation.

TI Special Article: Unconventional Medicine In The United States -- Prevalence, Costs, And Patterns Of Use.

SO The New England Journal of Medicine. 1993 Jan 28. 328 (4). pp 246-252.

PU Copyright 1993 by the Massachusetts Medical Society. PD 930128. PT Article (ART). IS 0028-4793. AB Abstract

Background. Many people use unconventional therapies for health problems, but the extent of this use and the costs are not known. We conducted a national survey to determine the prevalence, costs, and patterns of use of unconventional therapies, such as acupuncture and chiropractic.
Methods. We limited the therapies studied to 16 commonly used interventions neither taught widely in U.S. medical schools nor generally available in U.S. hospitals. We completed telephone interviews with 1539 adults (response rate, 67 percent) in a national sample of adults 18 years of age or older in 1990. We asked respondents to report any serious or bothersome medical conditions and details of their use of conventional medical services; we then inquired about their use of unconventional therapy.

Results. One in three respondents (34 percent) reported using at least one unconventional therapy in the past year, and a third of these saw providers for unconventional therapy. The latter group had made an average of 19 visits to such providers during the preceding year, with an average charge per visit of $27.60. The frequency of use of unconventional therapy varied somewhat among sociodemographic groups, with the highest use reported by nonblack persons from 25 to 49 years of age who had relatively more education and higher incomes. The majority used unconventional therapy for chronic, as opposed to life-threatening, medical conditions. Among those who used unconventional therapy for serious medical conditions, the vast majority (83 percent) also sought treatment for the same condition from a medical doctor; however, 72 percent of the respondents who used unconventional therapy did not inform their medical doctor that they had done so. Extrapolation to the U.S. population suggests that in 1990 Americans made an estimated 425 million visits to providers of unconventional therapy. This number exceeds the number of visits to all U.S. primary care physicians (388 million). Expenditures associated with use of unconventional therapy in 1990 amounted to approximately $13.7 billion, three quarters of which ($10.3 billion) was paid out of pocket. This figure is comparable to the $12.8 billion spent out of pocket annually for all hospitalizations in the United States.

Conclusions. The frequency of use of unconventional therapy in the United States is far higher than previously reported. Medical doctors should ask about their patients' use of unconventional therapy whenever they obtain a medical history. (N Engl J Med 1993;328:246-52.).
Unconventional, alternative, or unorthodox therapies are difficult to define, because they encompass a broad spectrum of practices and beliefs. As Murray and Rubel have written, "Many are well known, others are exotic or mysterious, and some are dangerous" *RF 1 *. From a sociological standpoint, unconventional therapy refers to medical practices that are not in conformity with the standards of the medical community *RF 2 *. Here we define unconventional therapies as medical interventions not taught widely at U.S. medical schools or generally available at U.S. hospitals. Examples include acupuncture, chiropractic, and massage therapy.

Studies based on samples in limited geographic areas suggest that the use of unconventional therapy is widespread *RF 3,4,5 *. In particular, unconventional therapies are frequently used by patients with cancer, *RF 6,7,8,9,10,11 * arthritis, *RF 11,12,13 * chronic back pain, *RF 3,14 * the acquired immunodeficiency syndrome, *RF 15 * gastrointestinal problems, *RF 16,17 * chronic renal failure, *RF 18 * and eating disorders *RF 19 *. Little is known, however, about the overall prevalence, cost, and patterns of use of unconventional therapy in the United States *RF 1,20 *.

To improve our understanding of the use of unconventional therapy, we conducted a national telephone survey focusing on 16 interventions found, on the basis of pilot research, to be representative of unconventional therapies used commonly in the United States. Our study focused on the following questions: What is the extent of use of unconventional therapy in the United States? How much is spent annually on these therapies, including out-of-pocket and third-party payments? What sociodemographic factors distinguish users of unconventional therapy from nonusers? For what medical conditions do people most commonly use unconventional therapy? And to what extent are medical doctors responsible for or informed about the use of unconventional therapy by their patients?

Methods Sample We conducted our survey by telephone between January 18 and March 7, 1991. The sample was selected by means of random-digit dialing *RF 21 *. We limited eligibility to English-speaking persons, 18 years of age or older, in whom cognitive or physical impairment did not prevent the completion of the interview. We designed the survey with a target sample of 1500. Assuming an estimated prevalence of use of unconventional therapy between 10 and 50 percent, we calculated that 1500 interviews should produce estimated prevalence rates with 95 percent confidence intervals of 2 to 3 percent.

Of the initial sample of 5158 telephone numbers, 38 percent were nonworking, and 13 percent were not assigned to households. We declared 221 respondents ineligible because they did not speak English (97), because of cognitive or physical incapacity (96), or because they were temporarily unavailable (28). Among the remaining 2295 eligible respondents, 1539 completed the interview, 653 declined to participate (81 of them before we could establish eligibility), and 103 began the interview but stopped before completing all questions. These figures correspond to a 67 percent overall response rate among eligible respondents. Only one respondent per household was eligible to be interviewed. This person was selected by computer randomization from the list of household members given by the first household member contacted. Persons with responses substantially different from the remainder of the sample (for example, those with frequent visits to a provider of unconventional therapy at no cost) were contacted again by a supervisor for verification or clarification of their responses. Since we asked respondents about the use of unconventional therapy during the 12 months before the interview, we considered the results representative of 1990.

The Interview We described the interview to the respondents as a survey by investigators from Harvard Medical School that was designed to assess the health care practices of Americans. We made no mention of unconventional therapy while recruiting the respondents. The interviews, which averaged 25 minutes in length, began with questions on the respondents' health, health worries, days in bed at home or in the hospital, and indicators of functional impairment caused by health problems. We then asked the respondents about their interactions with medical doctors during the past 12 months. A "medical doctor" was defined early in the interview as "a medical doctor (M.D.) or an osteopath (D.O.), not a chiropractor or other nonmedical doctor." Throughout the remainder of the interview we used the term "medical doctor." We use the same term in this report when referring to a respondent's provider of conventional medical care.

We next assessed the respondents' medical problems. The interviewers stated: "Now I'm going to read a list of conditions. Please tell me if you have had any of these conditions in the past 12 months." The interviewers then asked about 24 medical conditions and offered a follow-up question, "What other important conditions did you have?" The 24 conditions included common symptoms (such as back problems, digestive problems, dizziness, headache, and allergies), as well as specific diagnoses (such as high blood pressure, diabetes, and cancer). Only 8 percent of the respondents reported conditions not included in our list.

The respondents were then asked to identify the three (or fewer) "most bothersome or serious" health problems from the list they had just provided. These health problems are referred to here as "principal medical conditions." We asked the respondents whether they had seen their medical doctor for each principal medical condition during the past 12 months and what their perceptions of these interactions had been.