There is significant and growing national interest in the introduction and integration of CAM instruction into allopathic medical education, in part supported by recent educational funding from the National Institutes of Health. Interestingly, rates of CAM use by medical students  were found to be higher than reported in the US general population in 1998 [2, 3] and 2004 . Tracking change in learner attitudes is one strategy to document successful and effective CAM instruction. However, evaluating CAM curricular impact is complicated by the perceived heterogeneity of trainees' baseline attitudes toward CAM, per se, and the application of CAM to medical practice. A further complication is the absence of reliable, practical, and valid measures of CAM learning outcomes. Although reports have documented attitudes of medical students [5, 6] and nursing, medical and pharmacy students and faculty , the surveys used measures that were not validated. Validation studies for two CAM attitude measures recently were reported. The first, the Integrative Medicine Attitude Questionnaire (IMAQ), compared internists attending a conference on holistic medicine to those attending an annual general professional meeting . The second, the CAM Health Belief Questionnaire (CHBQ), was validated by including the IMAQ and using three cohorts (n = 272) of medical students at one institution . The CHBQ was found to have coefficient alpha reliability = 0.75, and total CHBQ attitude scale scores positively correlated with total IMAQ scores (r = 0.71; p < .0005).
The objective of this study was to investigate and compare CAM attitudes, CAM use, and CAM information-seeking behaviours to derive priorities for CAM instruction in medical school and residency. The groups surveyed were (a) medical students, (b) medical and surgical interns at the start of their post-graduate training, and (c) faculty who teach or intend to integrate CAM into their courses or classes. We hypothesized that this selected group of faculty would have more positive attitudes, would themselves use CAM modalities at a high rate, and would use more CAM information resources than either interns or medical students. One of the two student cohorts was surveyed longitudinally in their second and third years of training to track attitude changes, if any, across the continuum of undergraduate medical training.
At the University of California, Irvine (UCI), CAM instruction for medical students at the time of the study was offered in year 1 as a 2-hour panel discussion with patients and CAM practitioners. Subsequent CAM activities were learning issues integrated into problem-based learning cases taught longitudinally across year 1. During one of 4 evidence-based medicine (EBM) classes, students were shown a variety of CAM databases and examples of CAM-related evidence and were required to perform information searches. In year 2 students interviewed at least one patient who used a CAM modality during a community preceptorship and presented the case to peers and faculty with evidence related to that CAM modality. In the third year, students received CAM instruction at noon lectures in two clerkships and were tested for their ability to counsel a patient on acupuncture and use of an herbal for osteoarthritis in the Family Medicine clerkship. Total hours of required CAM instruction across 4 years was approximately 8 hours.
The study was approved by the Institutional Review Board.
Student respondents included two medical student class cohorts (n = 355) at the University of California, Irvine (UCI), School of Medicine. There were two consecutive first-year (MS1; n = 170) and two consecutive second-year (MS2; n = 185) class cohorts. MS1 were surveyed during the first six weeks of medical school (fall of 2003 and 2004) as an in-class exercise and before exposure to any CAM instruction. One MS2 cohort was surveyed during the last six weeks of their second year (spring of 2002); and the other was surveyed midway through their second year (winter of 2002). The survey of both MS2 cohorts occurred after exposure to 3 hours of didactic CAM instruction in a required Patient Doctor course. One student cohort also was resurveyed at the end of their third year of medical school, i.e. three times in total, in year 1, 2 and 3 of medical school.
Respondents from the intern group were interns (n = 258) entering medical and surgical residencies at UCI in the academic years beginning 1 July 2002 and 2003.
Faculty respondents were faculty attending one of two workshops offered on evidence-based CAM instruction in November 2002 and October 2003 (n = 81). The faculty included nursing and physician faculty representing diverse medical disciplines that included primary care and subspecialties.