METHOD
As a routine part of the Jefferson Longitudinal Study of Medical Education,16 a follow-up study of graduates is conducted every five to six years to investigate the graduates' professional activities, research productivity, career satisfaction, perceptions of pertinent issues in medical education and practice, and other educational outcomes. In May 1998, a 33-item questionnaire, which had been pilot tested by 30 faculty and 196 senior medical students,17 was mailed to all graduates of Jefferson Medical College (JMC) between 1987 and 1992 (n = 1,272). The final version of the questionnaire consisted of items addressing five aspects of changes in the U.S. health care system: medical education, quality of care, patient referral, costs of care, and ethical and socio-political issues.17 Physicians were asked to respond to each question by using a five-point Likert-type scale (from strongly disagree = 1 to strongly agree = 5) (copies of the questionnaire are available from authors). To measure young physicians' support for unionization to preserve their influence, we asked them to respond to the statement: Physicians should unionize to maintain the influence of their profession.
Three follow-up reminders (about three weeks apart) were sent to non-respondents. We used correlational analysis (bivariate and multivariate), t-test, chi-square, and z-test for proportions for statistical analyses.
RESULTS
Of the 1,272 physicians to whom surveys mailed, 835 returned useable questionnaires (66% response rate). Of the respondents, 578 (69%) were men and 257 (31%) women; the average age was 35.8 years (SD = 3.4 years). The primary areas of practice of the respondents were family medicine (14%), hospital-based specialties (12%), general surgery and surgical subspecialties (11%), general internal medicine (10%), medical subspecialties (10%), emergency medicine (6%), orthopedic surgery (6%), pediatrics (5%), obstetrics and gynecology (4%), and psychiatry (3%). The remainder were in miscellaneous specialties.
Comparisons of respondents and non-respondents. We compared selected demographic and performance variables between those who responded to the survey (n = 835) and those who did not (n = 437). We observed no significant difference between the two groups with regard to gender (31% men versus 33% women), age (35.8 versus 35.9 years), salaried faculty appointments (14% versus 12%) or practice as a generalist (29% versus 34%).
Also, we found no substantial difference between the respondents and the non-respondents on academic performance measures such as scores on the three parts of the National Board of Medical Examiners licensing examinations (effect-size estimates = .06, .02, and .01, respectively). Similarly, averages of clinical competence ratings made by residency program supervisors at the end of the first postgraduate training year in three areas, data-gathering and processing skills, interpersonal skills and attitudes, and socio-economic aspects of patient care,18,19 were the same in the two groups (effect-size estimates = .03, .05, and .05, respectively). These findings suggest that the physicians who responded to the survey can be considered representative of the entire study population with respect to the aforementioned variables.
Physicians' support for unionization. Of the 820 physicians who responded to the item that physicians should unionize to maintain the influence of their profession, 119 (15%) strongly agreed, 226 (28%) agreed, 221 (26%) expressed no opinion, 183 (22%) disagreed, and 71 (9%) strongly disagreed. Therefore, 43% of the study sample supported unionization (strongly agree and agree combined), and 31% did not support unionization (strongly disagree and disagree combined).
Bivariate correlations. The measure of support for unionization yielded statistically significant correlations with responses to 15 items on the questionnaire. Supporters of unionization were less likely to agree with the statement that the evolving trend toward managed care in the United States will ultimately improve the quality of health care (r = -.22, p < .01), were less likely to concur that HMOs or managed care physicians have the same dedication to their patients as do physicians in fee-for-service systems (r = -.21, p < .01), and did not hold a positive view of changes occurring in the health care delivery system (r = -.21, p < .01). Supporters of unionization also disagreed with the idea that the future of health care should be based on the needs of society rather than on the satisfaction of physicians (r = -.17, p < .01), and were less likely to endorse the concept that physicians should take an active role in supporting rather than resisting changes in the health care system (r = -.12, p < .01).
The physicians who supported unionization were more likely to agree that changes in the health care system were impairing physicians' independence (r = .25, p < .01), to agree that capitation would reduce physicians' motivation to closely monitor their patients on an ongoing basis (r = .18, p < .01), to believe that managed care organizations restrain physicians' freedom to provide optimal care to their patients (r = .18, p < .01), to agree that involvement of nurse practitioners in the diagnosis and treatment of diseases would compromise the quality of care (r = .17, p < .01), to endorse physician-assisted suicide (r = .17, p < .01), to approve referring emotionally disturbed patients to mental health specialists (r = .13, p < .01), to believe that the primary goal of managed care organizations is to manage costs not to manage care (r = .13, p < .01), to endorse government's role in regulation of policies that influence the quality of care (r = .13, p < .01), to agree that the control of health care by insurers will ultimately lower the quality of health care (r = .12, p < .01), and to believe that the role of primary care physicians as gatekeepers in the health care system is driven by financial factors (r = .11, p < .01).
Multivariate regression analysis. To investigate the unique contributions of the variables in predicting young physicians' support for unionization, we used the multivariate stepwise regression algorithm in which support for unionization to maintain influence was the dependent variable (criterion measure) and responses to the 15 survey items that had statistically significant correlations with the responding physicians' opinions of unionization were the independent variables (predictors). The summary results of the regression analysis are reported in Table 1.
The most significant predictor was the physicians' belief that changes in the health care system are impairing physicians' independence, followed by approving referral of emotionally disturbed patients to mental health professionals, and endorsing legalization of physician-assisted suicide. Other significant predictors included physicians' disagreement with the statement that changes in the health care system are positive, and their negative views of the involvement of nurse practitioners in diagnosis and treatment. Physicians' disagreement with the statement that physicians in managed care render the same quality of care as their independent counterparts and their disagreement with the statement that the needs of society have precedence over the satisfaction of physicians were also significant predictors. The multivariate R was .38 (p < .01), indicating that about 14% of the variance of willingness to join labor unions could be accounted for by variation in these seven predictors.
Gender, practice area, and performance measures. Equal proportions of men and women supported the concept of unionization. Statistically significant differences were found among physicians in different areas of practice. Family medicine physicians demonstrated the lowest support rate (33 of 116, 28%). The highest support rates were found among physicians in general surgery and surgical subspecialties, including orthopedic surgery (64 of 136, 47%), medical subspecialties (40 of 86, 47%), and pediatrics (18 of 38, 47%), followed by physicians in hospital-based specialties (anesthesiology, pathology, and radiology, 45 of 97, 46%), and psychiatry (13 of 28, 46%). The support rate was 36% (31 of 85) for physicians in general internal medicine. The z-test for proportions indicated that the differences in support rates between family medicine physicians and the other aforementioned specialties, with the exceptions of general internists and psychiatrists, were statistically significant (p = < .05).
No significant relationship was observed between support for the concept of unionization and academic appointment, board certification status, performance in medical school (first- and second-year grade-point averages and performances on core clerkship examinations in the third year of medical school), scores on Parts 1, 2, and 3 of the National Board of Medical Examiners examinations, or ratings of post-graduate clinical competence in data gathering and processing skills, interpersonal skills and attitudes, and socioeconomic aspects of patient care. No significant correlation was found between physicians' support for unionization and the amount of educational debt in medical school.
DISCUSSION
Previous studies have confirmed the cyclical nature of physicians' attitudes toward unionization.15,20 In a 1972 survey of 367 Medical Economics readers, 61% agreed that physicians should form a union and only 28% disagreed.21 In a 1987 spot check of readers, however, only a third were in favor of physicians' unions.22 Our survey of young physicians, in which 43% of the surveyed physicians supported unionization, confirms the resurgence of interest in physicians' unions that was also demonstrated by the AMA's recent decision. Our data were collected prior to the AMA's House of Delegates' vote in support of unionization. Had the survey been performed after this vote, physicians' support for unionization might have been even greater.
Recent statements by the AMA and others have suggested reasons for physicians' renewed interest in unionization. Expansion of managed care health plans, according to Dickey, is among the reasons … because managed care health plans have obtained a significant amount of economic leverage over physicians. This leverage has enabled health plans to assume substantial control over medical decision-making for patients, drive down the incomes of many physicians, and to threaten the viability of some physicians practices.23 Also, according to Laskin, the rapid growth in managed care, with its increasing effects on the quality of patient service, the doctor-patient relationship, the economics of practice, and professional decision-making, is seeing greater numbers of frustrated practitioners seeking ways to regain a voice in the medical marketplace.24
The results of our multivariate regression analyses confirmed the afore-mentioned propositions. In our study, the strongest predictor of young physicians' support for unionization was the physicians' belief that their independence had been impaired by changes in the health care system. Their agreement that physician-assisted suicide should be legalized, possibly another indication of desire for increased independence in decision-making, was also a significant predictor of support for unionization.
Other significant predictors were related to physicians' attitudes toward providing appropriate care for their patients: those supporting physicians' unions felt that a primary care physician should refer a suicidal patient to a psychiatrist and disagreed with the statement that physicians involved in managed care plans have the same dedication to their patients as did fee-for-service physicians. Physicians' negative views of the changes in the health care system and their negative perceptions of the quality of care provided by managed care delivery systems also contributed significantly to their support of the concept of unionization.
More physicians who supported unionization agreed that the involvement of nurse practitioners in the diagnosis and treatment of some diseases will compromise the quality of care. This finding may reflect greater degrees of altruism and concern for the patient: physicians may genuinely feel that nurse practitioners' treatment of some illnesses may result in adverse outcomes. It may also be a manifestation of these physicians' concern that their spheres of influence and incomes will be adversely affected if nurse practitioners participate in patient care.
Finally, those physicians who disagreed with the statement that the future of health care should be based on the needs of society and not on the satisfaction of physicians were more likely to be in favor of unionization. This disturbing finding may indicate that physicians who favor unionization do so, at least in part, out of self-interest. It may also reflect these physicians' lack of trust in the ability of policymakers to define specific societal needs.
While support for unionization among young physicians was unrelated to gender, academic achievement, performance on licensing examinations, ratings of clinical competence in residency training, or educational debt, their practice specialties were related to their support for unionization. Surgeons, medical subspecialists, pediatricians, and hospital-based specialists were the most likely to support, and family physicians were the least likely to support, physicians' unionization. The present study did not allow for defining the reasons that family physicians expressed less inclination to unionize. Although not statistically significant, fewer general internists also endorsed unionization. One may speculate that these findings could be related to the role of family physicians and general internists in managed care's concept of primary care physicians as gatekeepers. Family physicians and general internists are also less dependent upon costly technology for the diagnosis and management of their patients than are specialists, and they may not feel that managed care interferes with their ability to care for their patients. Additional data are needed to substantiate these speculations.
There have been two prior studies of physicians' attitudes toward unionization. In 1976, Engel and Schulman surveyed a sample of 200 housestaff at a large medical center.25 They found that advocacy of unionization was not determined solely by degrees of altruism or the residents' perceptions of how autonomously they functioned, but was also influenced by age and family back-grounds. Those who perceived themselves to be highly autonomous and those whose fathers were non-professionals were lowest in their advocacy of unionization. The more altruistic residents advocated unionization more strongly if they had professional fathers. Engel and Schulman25 suggested that housestaff whose fathers had not achieved professional status might feel that unionization would detract from their own rise in status. Our study did not examine this area. Among those with less altruistic attitudes, older housestaff were less supportive of unions.25 If this previously reported association of less support for unionization by older housestaff is applicable to practicing physicians, our finding of 43% of the young physicians in support of unionization is probably an overestimate for all practicing physicians.
In 1980, Klover and colleagues26 sought to identify the issues around which physicians had unionized or might unionize by studying a random sample of 231 members of a county medical society who represented all specialties. A total of 54 non-union physicians and 15 union physicians were asked open-ended questions about unionization. The highest number of union members (39%) responded that interference in professional judgments by outsiders was their main reason for joining a union. Other issues were government pays more attention to unions than individuals (23%), and patient considerations (15%). No union members ranked economic considerations as important. However, when both union and non-union members were asked what factors might influence physicians to join unions in the future, the most frequently identified issue was price controls (33%), followed by national health insurance (24%).26 Our study did not specifically address price-control issues. However, the negative endorsement of nurse practitioner's care by those who favored unionization could be interpreted as concern about possible declines in income.
The strongest predictors of young physicians' support for unionization in our study were desire for independence and concern for the best patient care. In a previous survey, a substantial majority of respondents indicated that today's physicians believe that control of health care by insurance companies leads to substandard care and that managed care systems restrain them from providing optimal patient care.17,27 Thus, given the ongoing changes in the U.S. health care system, it may be reasonable to conclude that interest in physicians' unionization will continue to increase. Whether physicians' unions will be able to address and counter the frustrations of physicians is unclear. It remains to be seen whether this era, like the 1980s, will be one in which physicians' unions virtually disappear after the turbulence subsides, or whether it will begin a new age in which most physicians sustain support for unionization.
The advantages of our study include the large sample size, the gender composition of the sample, and the broad specialty and geographic distributions of the participants. Comparisons of the study sample with national data on physicians' characteristics and distributions28 suggest that our study sample resembles the population of young practicing physicians in the United States (ages 44 and younger) with regard to gender (32% of women physicians in that age group nationally), and some specialties, including internal medicine and medical subspecialties, obstetrics/gynecology, and psychiatry. However, our findings may not be fully generalizable to the entire population of young physicians in the United States due to the fact that the study participants graduated from a single medical school. Likewise, the opinions of these relatively young physicians may not necessarily reflect those of older physicians.
Despite these limitations, our results indicate that in a sample of young physicians approval of physicians' unions is associated with factors that reflect their frustration with market-driven policies that compromise quality of care and factors that negatively affect their autonomy, incomes, and personal satisfaction. Thus, the support of young physicians for unionization may derive from a complex combination of factors that include both altruism and self-interest.
REFERENCES
1. Boyer MM. Physicians examining physician union. Pennsylvania Med. 1997;100:12-5.
2. Mangel JS. Unionism and professionalism-physician, do no harm: one person's view. J. Health Care Finance. 1998;24:78-83.
3. Marcus SA. Doctors need union representation. N Engl J Med. 1981;305:525.
4. Marcus SA. Trade unionism for doctors. An idea whose time has come. N Engl J Med. 1984;311:1508-11.
5. Marcus SA. The black physician and medical unionism. JAMA. 1986;78:921-3.
6. Stobo JD. Medicine. A profession or trade union? Arthritis & Rheum. 1991;34:641-5.
7. Wilsford D. The cohesion and fragmentation of organized medicine in France and the United States. J. Health Polit Policy & Law. 1987;12(3):481-503.
8. Riffer J. Physicians unions fight loss of control. Hospitals. 1986;60(2):82.
9. Rudd P, Waitzkin H. Letter: Physicians working for the union. N Eng J Med. 1976;295:54.
10. Woods D. Unions moving into MD offices and clinics. CMAJ. 1974;111:862-4,874.
11. Urlich, S. Will your appendectomy be performed by a member of the AFL-CIO? This button is fictitious, but doctors' unions are very real. Modern Hosp. 1973;121(4):63-7.
12. Keith SN. Collective bargaining and strike among physicians. J Natl Med Assoc. 1984;76:1117-21.
13. Willett DE. More on doctors' unions: are private practitioners de facto employees? N Engl J Med. 1985;312:1067.
14. Dworkin JB, Extejt MM, Demming SR. Unionism in hospitals, or what's happened since PL 93-360? Health Care Manage Rev. 1980;5(4):75-81.
15. Budrys G. When doctors join unions. Ithaca, NY: Cornell University Press, 1997.
16. Hojat M, Gonnella, JS, Veloski JJ, Erdmann JB. Jefferson Medical College's longitudinal study: A prototype for assessment of changes. Education for Health: Changes in Training and Practice. 1996;9:99-113.
17. Hojat M, Veloski JJ, Louis DZ, Xu G, Ibarra D, Gottlieb JE, Erdmann JB. Perceptions of medical school seniors of the current changes in the U.S. health care system. Eval & Health Professions. 1999;22:169-83.
18. Hojat M, Veloski JJ, Borenstein BD. Components of clinical competence ratings: an empirical approach. Educ Psychol Meas. 1986;46:761-9.
19. Hojat M, Borenstein BD, Veloski JJ. Cognitive and non-cognitive factors in predicting the clinical performance of medical school graduates. J Med Ed. 1988;63:323-5.
20. Phan C. Physician unionization: The impact on the medical profession. J Legal Med. 1999;20:115-40.
21. Paxton HT. Union now? Yes, say 61% of surveyed doctors. Med Econ. 1972;4:31-43.
22. Bumke D. Where doctors unions are finding new life. Med Econ. 1987;21:63-83.
23. Dickey N. Report of the Board of Trustees. Chicago, Illinois: American Medical Association. June 1997, BOT Report 41:216.
24. Laskin DM. Are we ready to get organized? J Oral Maxillofac Surg. 1999;57:641.
25. Engel GV, Schulman ML. The relationship of unionization to professionalism in medicine. J Med Ed. 1976;51:132-4.
26. Klover J, Stephens DB, Luchsinger VP. Contemporary perceptions of unionization in the medical profession: a study of attitudes of unionized and non-union physicians. J Collective Negotiations. 1980;9(2):107-116.
27. Hojat M, Gonnella JS, Erdmann JB, et al. Physicians' perceptions of the changing health care system: comparisons by gender and subspecialties. J Commun Health. 2000;25:455-71.
28. Pasko T, Seidman B. Physician Characteristics and Distribution in the US. Chicago, Il: American Medical Association, Division of Survey and Data Resources, 1999 (Tables A-5 and B-2).