More and more physicians in the United States, especially

Do Osteopathic Physicians Differ in Patient Interaction from Allopathic Physicians? An Empirically Derived Approach Timothy S. Carey, MD, MPH; Thomas M. Motyka, DO; Joanne M. Garrett, PhD; Robert B. Keller, MD Colleges of osteopathic medicine teach osteopathic principles, which provide a different approach to and interaction with patients than principles taught in allopathic medical schools. The authors examined whether osteopathic primary care physicians interactions with patients reflect the principles of osteopathic medicine when compared with allopathic physicians interactions. The principles of osteopathic medicine were adapted to elements that could be measured from an audio recording. This 26-item index was refined with two focus groups of practicing osteopathic physicians. Fifty-four patient visits to 11 osteopathic and 7 allopathic primary care physicians in Maine for screening physicals, headache, low back pain, and hypertension were recorded on audiotape and were dual-abstracted. When the 26-item index of osteopathic principles was summed, the osteopathic physicians had consistently higher scores (11 vs. 6.9; P =.01) than allopathic physicians, and visit length was similar (22 minutes vs. 20 minutes, respectively). Twenty-three of the 26 items were used more commonly by the osteopathic physicians. Osteopathic physicians were more likely than allopathic physicians to use patients first names; explain etiologic factors to patients; and discuss social, family, and emotional impact of illnesses. In this study, osteopathic physicians were easily distinguishable from allopathic physicians by their verbal interactions with patients. Future studies should replicate this finding as well as determine whether it correlates with patient outcomes and satisfaction. From the University of North Carolina at Chapel Hill, where Dr Carey is a professor in the School of Medicine and Dr Garrett is associate director and professor in the Departments of Medicine and Social Medicine, and from the Maine Medical Assessment Foundation, where Dr Keller is executive director and an orthopedic surgeon. Dr Motyka is in private practice in Chapel Hill, NC. This study was supported by funding from the American Osteopathic Association. Address correspondence to Timothy S. Carey, MD, MPH, The Cecil G. Sheps Center for Health Services Research, 725 Airport Rd, CB No. 7590, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7590. E-mail: tim_carey@unc.edu More and more physicians in the United States, especially those in primary care, are graduates of colleges of osteopathic medicine. Osteopathic physicians generally practice in parallel with the more numerous graduates of allopathic medical schools. However, osteopathic medicine comes from a distinctly different professional tradition. The first school of osteopathic medicine was founded by Andrew Taylor Still, MD, DO, in 1892 in Kirksville, Missouri. There are currently 20 colleges of osteopathic medicine in the United States.1 Approximately 5% of physicians in the United States are graduates of colleges of osteopathic medicine, accounting for more than 100 million patient visits annually.2,3 In the past, relatively little empiric research has examined ways in which osteopathic physicians differ from allopathic physicians other than by using osteopathic principles and practice, which include osteopathic manipulative treatment (OMT). However, the tradition of viewing osteopathic medicine as a distinct profession is deeply rooted within the culture of colleges of osteopathic medicine. We took advantage of an ongoing cohort study, the Maine Osteopathic Outcomes Study (MOOS), to examine whether a distinctive characteristic of osteopathic medicine might be that osteopathic physicians have different communication patterns than allopathic physicians. Osteopathic principles, initially elucidated by Dr Still, include the unity of the body; self-regulatory and self-healing systems; the relationship between structure and function; and a rational treatment approach.4 These initial, broadly based treatment principles have been modified in the twentieth century. There remains an emphasis on a structural approach: the osteopathic principles of practice, which osteopathic medical educators believe demonstrate distinctive differences between osteopathic and allopathic physicians. These revised principles of osteopathic medicine include the following: Osteopathic medicine is the practice of rational medicine based on the medical sciences. Osteopathic medicine treats the individual as a whole. Osteopathic medicine recognizes the body s ability to be self-regulating and self-healing. Osteopathic medicine acknowledges the structure-function interrelationship. JAOA Vol 103 No 7 July 2003 313

Osteopathic medicine endorses the use of OMT. Osteopathic medicine emphasizes a close and personal relationship between physician and patient. Osteopathic medicine recognizes that health care requires intelligent collaboration between the lay public and physicians. Osteopathic medicine relies on all recognized modalities of medical services. Osteopathic medicine emphasizes disease prevention. Osteopathic medicine endorses the application of all services of modern scientific medicine that are needed to meet the needs of all people.5 Osteopathic philosophy has continued to develop since the inception of this study. Recently revised statements of osteopathic principles would not have influenced physician behavior in this study and were not in existence for use as the basis of this investigation.6 We operationalized these principles into draft items that could be identified during typical physician-patient encounters. We then sought to validate these operationalized osteopathic principles by determining whether they could be measured during physician-patient treatment visits. Rather than attempt to examine all aspects of physician-patient communication during office visits, we chose to examine those components of verbal interaction that we thought would most likely identify differences between the two medical professions. Our hypothesis was that osteopathic physicians would use these verbal interactions more than allopathic physicians. Methods We conducted two focus groups with experienced osteopathic physicians in Philadelphia and Florida. Participants in the focus groups (6 to 8 physicians per group) were not connected with the primary data collection for the cohort study or the study of physician-patient communication in Maine. The physicians in the focus groups, recruited through local colleges of osteopathic medicine and professional societies, were osteopathic physicians, practiced primarily in primary care, and saw patients greater than 50% of the time spent practicing. Recruitment started with a lead letter followed by phone contact. Each group discussion lasted 2 hours. The focus groups worked with the investigators (T.M. and T.C.) in operationalizing the modern osteopathic principles into items that could be measured during audiotaped office visits. Content from the first focus group was used to inform the discussion in the second focus group, leading to a list of candidate items. Each osteopathic principle was mapped to a physician-patient interaction. For example, recognizes the body s ability to be selfregulating and self-healing maps to items physician discusses what patient can do to improve own condition and discusses body s self-healing ability or reassures that condition will improve on its own. We excluded spinal manipulation from the communication list, as few allopathic physicians perform OMT, and OMT is considered a treatment modality, not a communication tool. We were interested in identifying distinguishing characteristics of osteopathic practice other than OMT. The list of candidate items was shared with several experts in osteopathic medicine (eg, associate deans, curricular directors) and was refined based on their comments. Pretesting was performed at the internal medicine and family practice clinics at the University of North Carolina at Chapel Hill the six patients pretested resulted in minimal change in the study process. We conducted our study using a subset of volunteer physicians participating in the MOOS. The MOOS is a prospective cohort study of acute low back pain in patients seen in primary care osteopathic and allopathic practices in Maine. The study of physician-patient communication was separate from the data collection of the cohort study. The physician participants (N 18) were told that we were examining communication characteristics of the two professions, but they were not informed of emphasis on osteopathic principles or which behaviors we were studying. Patients were asked at the beginning of their visits whether they would allow us to audiotape the visits. If willing, they signed an informed consent form, and the tape recorder in the room was activated. The research protocol was approved by the institutional review board. An omnidirectional microphone was used for clarity. A total of 54 audiotapes were obtained from 11 osteopathic physicians and 7 allopathic physicians. Four types of visits were eligible for taping: acute low back pain, headache, hypertension, and health maintenance. These conditions were chosen to represent a range of visits that are most consistent with osteopathic intervention (back pain) to those least consistent with osteopathic intervention (health maintenance). The audiotapes were sent to the University of North Carolina at Chapel Hill. Two reviewers (one research assistant, one physician) abstracted each audiotape using an explicit form, addressing communication characteristics of interest. Each reviewer listened to each tape twice. The two abstractions were performed independently. Differences between abstractors were resolved at adjudication meetings, and tapes were reviewed a third time if necessary. The data were entered and analyzed in a blinded fashion. Abstractors and data analysts did not know which tapes came from osteopathic physicians and which came from allopathic physicians. In a global assessment, abstractors were asked at the end of each audiotape abstraction process to assess whether they thought the physician was an osteopathic or an allopathic physician. Statistical analysis was performed using a standard package (Stata, Stata Corp, College Station, Texas). Tests for means and proportions were adjusted for the cluster effect, as each physician contributed more than one interview to the evaluation. The individual items were examined separately and summed to create a draft index. 314 JAOA Vol 103 No 7 July 2003

Table 1 Patient Characteristics Osteopathic Allopathic Physicians Physicians (N = 32) (N = 22) Women, % 72 77 Hypertension visits, % 19 27 Low back pain visits, % 31 27 Headache visits, % 25 23 Health maintenance, % 25 23 Results Demographic characteristics of the 54 patients (32 patients of osteopathic physicians and 22 patients of allopathic physicians) are provided in Table 1. Typical for primary care, most (70%) of the patients were women. Average number of physician years in practice was 11.7. Distribution of the four diagnoses (hypertension, low back pain, headache, health maintenance) among the osteopathic and allopathic groups was similar. Average length of visit determined by timing the audiotape was similar between the two professions, but slightly longer (22 minutes vs. 20 minutes) in the osteopathic group. This difference was not statistically significant. We could determine from the tapes when OMT was used during the visits. The osteopathic physicians used OMT for visits other than back pain. Osteopathic manipulative treatment was used in 1of 8 general examination visits, 2 of 6 hypertension visits, 4 of 8 headache visits, and 9 of 10 back pain visits. Table 2 indicates the 26 items that were coded as being present or absent in the audiotapes. The number of subjects analyzed is less than 54 for some of the variables, as some of the visits were for preventive care and, therefore, no complaint was specified. Twenty-three of the 26 items were used more commonly by the osteopathic physicians compared with the allopathic physicians. Four comparisons reached statistical significance. Exact P values are provided when the test of significance was less than or equal to.05. Even among many of the items that did not reach statistical significance, the trend indicated more use of the behavior by the osteopathic physicians compared with the allopathic physicians. The only communication characteristics that appeared not to be used more by the osteopathic physicians were patients using the physicians first names (not done by either profession); touching the patients, but not during the examination (probably not reliably assessed, given the measurement method using audiotapes); discussing medical literature with the patients; and discussing health issues related to work. Only one item, discussion of the medical literature, had a trend toward greater use in the encounters with allopathic physicians. As these assessments were based on audiotapes, it was difficult to determine when the patient was being touched and when not; this was included as an exploratory variable. Statistical tests were conducted for each comparison, taking into account that the data reflected a clustered sampling design, ie, the patients were clustered within physicians (this design effect must be accounted for in statistical testing). We call the overall results the osteopathic principles of practice index (OPPI). For those patients who had complete responses to all items as being present or absent, we were able to assemble an index of responses indicating osteopathic behavior. Of the 26 candidate items, the osteopathic physicians had an average of 11 positive responses, and the allopathic physicians had an average of 6.9. This difference was statistically significant (P.01) even after accounting for the cluster effect of having multiple patients per physician. In an exploratory analysis, we examined the diagnosis-specific scores. We expected that the osteopathic behaviors would be more evident in the mechanical diagnoses (headache and low back pain) than for hypertension. Although the numbers were small, the index scores for low back pain and headache (12.6 and 10.1, respectively) were higher than for hypertension (9.5). The allopathic scores within each of these diagnoses were similar (low back pain, 7.2; headache, 7.2; hypertension, 6.5). When we summed the available items within the general medical examination patients, we also found that patients of osteopathic physicians scored higher on the index than patients of allopathic physicians (9.5 vs. 6.8, respectively). Physician gender has also been demonstrated to influence communication.7 We found that summary scores were similar between male and female allopathic physicians (6.6 and 7.5, respectively), but that male osteopathic physicians had higher scores than osteopathic female physicians (11.9 vs. 7.6, respectively). Such interactions should be examined in future studies. Our sample size was too small to examine interactions across physician and patient gender and specialty. Figure 1 demonstrates the average score for each of the 18 physicians. While the osteopathic physicians did have higher scores than the allopathic physicians, there was some overlap, as some allopathic physicians had scores similar to osteopathic physicians. The one outlier osteopathic physician with a low score was based on a single interview. We asked the abstractors to estimate whether each physician they were listening to was an osteopathic physician or an allopathic physician. In some cases, blinding was not possible because OMT was performed. When physicians were osteopathic physicians, the abstractors were correct approximately 90% of the time. When the physicians were allopathic physicians, the abstractors were correct 70% of the time. That is, 30% of the time, both abstractors thought physicians were osteopathic physicians when they were allopathic physicians. JAOA Vol 103 No 7 July 2003 315

Table 2 Comparison of Osteopathic Physicians and Allopathic Physicians Patient Interactions According to Osteopathic Principles of Practice Osteopathic Physicians Allopathic Physicians (N = 32) (N = 22) P Discuss preventive measures specific to the complaint (N = 41) 79 47.03 Discuss general/unrelated health measures 88 73 Discuss family/social issues unrelated to health 66 55 Discuss health issues in relation to family life 56 32.01 Discuss health issues in relation to social activities 38 14.05 Discuss health issues in relation to work 53 55 Discuss patient s emotional state 66 32.02 Physician uses patient s first name 31 9.05 Patient uses physician s first name 0 0 Physician discusses his or her personal experience, not including professional experience with other patients 28 18 Touches patient but not during examination 3 0 Review of systems includes unrelated areas (N = 41) 71 53 Examination includes unrelated areas (N = 41) 54 24 Asks Anything else I can do for you? or equivalent 19 9 Asks Do you have any questions? or equivalent 38 14 Prescribes no medications (including over-the-counter medications) 38 14 Recommends herb/nutritional/physical or other non-drug alternative medications, not including osteopathic manipulative treatment 59 45 Discusses what patient can do to improve own condition 69 55 Physician discusses body s self-healing ability or reassures that condition will improve on its own 19 5 Physician discusses musculoskeletal cause or consequence related to patient s condition 63 32 History included musculoskeletal review of systems 66 50 Discussion of literature or scientific basis of treatment 9 18 Inquires about alternative modes of therapy patient may have used 25 14 Asks for patient s opinion on cause of problem 19 9 Asks for patient s opinion about treatment 13 14 Explains cause of problem or reasoning behind treatment 50 36 Mean total of yes responses (total = 26) 11 6.9.01 316 JAOA Vol 103 No 7 July 2003

Mean Score 20 15 10 5 0 DO (n = 24) Provider Type MD (n = 17) Figure. Mean scores for each physician on the osteopathic principles of practice index, based on 41 observations with some overlap in points. Discussion Our index items indicating characteristic communication styles of osteopathic physicians were derived from the theoretical basis of osteopathic principles. Using focus groups of practicing osteopathic physicians and experts in the field, we operationalized these theoretical constructs to communication characteristics that could be identified using audiotaped interviews. Our findings indicate that osteopathic physicians seem to use communication characteristics with their patients that are similar to theoretically derived osteopathic principles and that distinguish them from allopathic physicians. Our findings provide some initial validation that osteopathic principles distinguish osteopathic physicians from allopathic physicians, not just in terms of training but also in terms of the content of communication between physician and patient. These communication characteristics are clearly not unique to osteopathic physicians. Allopathic physicians commonly used these same communication techniques, and several allopathic physicians in our study had OPPIs similar to those of the osteopathic physicians in our sample. We had hypothesized that the osteopathic index would be highest in back pain, intermediate in headache, and lowest in hypertension. Hypertension is less of a mechanical or structural diagnosis compared with headache and back pain, and we thought that the osteopathic behaviors would be less manifested. Our data support this, indicating that more mechanical diagnoses were somewhat more likely to be associated with communications congruent with osteopathic principles. The strengths of our study are that our abstraction instrument was theoretically based and that our subjects (physicians) were blind to the items in which we were interested. While the physicians were told that we were studying communication between physicians and patients, the exact types of communication were unknown to the subject physicians. Our study does have several limitations. Osteopathic manipulative treatment was used with some patients, which would have indicated to audiotape reviewers that it was an osteopathic physician they were hearing. Our sample size was small: 54 physician-patient encounters across 18 physicians. Our generalizability to the US population of osteopathic physicians is limited in that the study physicians all practiced in a single state (Maine) and most osteopathic physicians were graduates of a single college of osteopathic medicine. Therefore, it is possible that the osteopathic principles may be more or less manifest in other parts of the United States with more heterogeneous physician populations. However, informal discussion with medical educators indicates that osteopathic principles are consistently emphasized in osteopathic undergraduate medical education. The osteopathic physicians in our study used OMT for multiple diagnoses, perhaps indicating special interest in this particular group in the manual treatment aspects of the osteopathic medical profession. Results of our study indicate that communication patterns differ between osteopathic physicians and allopathic physicians. We cannot correlate these communication patterns with improved clinical outcomes or differences in patient satisfaction. Other components of the MOOS focus on the difference in treatment patterns and outcomes of the care of acute low back pain. This work represents the first time that communication patterns have been evaluated between these similar, but distinct, medical professions. Our hypothesis that osteopathic physicians have different communication patterns did receive limited support in this small study. Future work should have larger sample size, stratify by physician gender and years in practice, examine physicians from multiple colleges of osteopathic medicine, and correlate communication patterns with patient satisfaction. By examining different communication patterns with patients, osteopathic and allopathic physicians can learn from each other with the goal of improved medical care of all patients. During the latter half of the twentieth century, the osteopathic medical profession was sometimes ambivalent regarding its level of distinctiveness from allopathic medicine in ways other than the use of OMT.8,9 Given that other health care practitioners (physical therapists, doctors of chiropractic) use manipulative therapy and that some allopathic physicians are beginning to use manual treatment in practice, basing a distinct JAOA Vol 103 No 7 July 2003 317

profession only on OMT might be called into doubt.10 Our finding of different communication patterns, which adhere to the theoretical basis of osteopathic principles, reinforces the notion that osteopathic physicians represent a profession that can be distinguished from allopathic physicians. Osteopathic physicians seem to have a communication style with patients that is more personal and somewhat more holistic in that issues relating to family, social activities, and patient emotions seem to be more commonly incorporated into visits. Additional research examining the presence of other distinguishing features of osteopathic and allopathic medical practice and their implications regarding patient outcomes of function, satisfaction, cost, and use can further illuminate this important and little-studied area. References 1. Osteopathic Medicine. Chicago, Ill: American Osteopathic Association; 2001. Available at: http://www.aoa-net.org/consumers/omed.htm. Accessed June 10, 2003. 2. AOA Fact Sheet. Chicago, Ill: American Osteopathic Association; 2001. Available at: http://www.aoa-net.org/consumers/omed.htm. Accessed June 10, 2003. 3. Osteopathic Medicine and Managed Care. Chicago Ill: American Osteopathic Association; 2001. Available at: http://www.aoa-net.org/consumers/mgdcare.htm. Accessed June 10, 2003. 4. Ward RC, ed. Foundations of Osteopathic Medicine. Baltimore, Md: Williams & Wilkins; 1997. 5. Sirica CM, ed. Osteopathic Medicine: Past, Present, and Future. New York, NY: Josiah Macy Jr Foundation; 1996: pp63-69. 6. Ward RC, ed. Foundations of Osteopathic Medicine. 2nd ed. Philadelphia, Pa: Lippincott Williams & Wilkins; 2002. 7 Roter DL, Hall JA. How physician gender shapes the communication and evaluation of medical care. Mayo Clin Proc. 2001;76:673-676. 8 Gevitz N. The DO s: Osteopathic Medicine in America. Baltimore, Md: John Hopkins University Press; 1982. 9. Gallagher RM, Humphrey FJ, Micozzi MS, eds. Osteopathic Medicine: A Reformation in Progress. New York, NY: Churchill Livingstone; 2001. 10. Johnson SM, Kurtz ME. Diminished use of osteopathic manipulative treatment and the consequent impact on the uniqueness of the osteopathic profession. Acad Med. 2001;76(8):821-828. UNIQUE Academic / Clinical Opportunity for Family Medicine Physician Texas College of Osteopathic Medicine, at the University of North Texas Health Science Center at Fort Worth is accepting applications for a boardcertified Family Medicine Physician. Applicants must be Board Certified in Family Practice by the AOBFP and possess a commitment to teaching and scholarly activities. The department is vibrant, progressive faculty consisting of 27 FTE DOs/MDs, seven PAs, six medical educators, a Family Practice residency program, a PA Studies program, several clinics for both residents and students, Divisions of Sports Medicine, Emergency Medicine, Research, and Rural Health offices of predoctoral training and preceptorship training. Experience in the practice of culturally sensitive medicine is a must. Bilingual Spanish / English is a plus. An academic rank of assistant or associate professor will depend upon experience. Competitive salary, excellent benefits. If you are innovative and excited about practicing in a dynamic new medical environment, then contact me about an immediate or anticipated future career opportunity in academic Family Medicine. Send your CV to: Elizabeth Palmarozzi, D.O. Acting Chairman, Department of Family Medicine University of North Texas Health Science Center at Fort Worth Texas College of Osteopathic Medicine 3500 Camp Bowie Boulevard Fort Worth, Texas 76107-2699 (817) 735-2440 FAX (817) 735-5089 The UNT Health Science Center enjoys a smoke-free campus and is an EEO/Affirmative Action Institution UNIVERSITY of NORTH TEXAS HEALTH SCIENCE CENTER at Fort Worth 318 JAOA Vol 103 No 7 July 2003