ORIGINAL CONTRIBUTION

Personal Values of Family Physicians, Practice Satisfaction, and Service to the Underserved
B. Clair Eliason, MD; Clare Guse, MS; Mark S. Gottlieb, PhD
Background: Personal values are defined as desirable

goals varying in importance that serve as guiding principles in peoples lives, and have been shown to influence specialty choice and relate to practice satisfaction. We wished to examine further the relationship of personal values to practice satisfaction and also to a physicians willingness to care for the underserved. We also wished to study associations that might exist among personal values, practice satisfaction, and a variety of practice characteristics.
Methodology: We randomly surveyed a stratified prob-

cient = 0.14, P = .002). Those involved in teaching medical trainees were more satisfied than those who were not involved (P = .009). Some value-type ratings were found to be positively associated with caring for the underserved. Those whose practices consisted of more than 40% underserved (underserved defined as Medicare, Medicaid, and indigent populations) rated the tradition (motivation tomaintaincustomsoftraditionalculture and religion)value type significantly higher (P = .02). Those whose practices consisted of more than 30% indigent care rated the universalism (motivation to enhance and protect the well-being of all people) value type significantly higher (P = .03).
Conclusions: Family physicians who viewed benevolence as a guiding principle in their lives reported a higher level of professional satisfaction. Likewise, physicians involved in the teaching of medical trainees were more satisfied with their profession. Family physicians who rate the universalism values highly are more likely to provide care to the indigent.

ability sample of 1224 practicing family physicians about their personal values (using the Schwartz values questionnaire), practice satisfaction, practice location, breadth of practice, demographics, board certification status, teaching involvement, and the payor mix of the practice.
Results: Family physicians rated the benevolence (moti-

vation to help those close to you) value type highest, and the ratings of the benevolence value type were positively associated with practice satisfaction (correlation coeffi-

Arch Fam Med. 2000;9:228-232 In addition to practice satisfaction, it is probable that the personal values of physicians are associated with other important characteristics of practice. Limited access to medical care for the underinsured has been identified as a problem in our health care system.7 This situation exists despite an adequate physician workforce. Overspecialization, economic pressures of physicians, maldistribution of physicians, insurance, and even public policy have worked against many patients and have limited their access to medical care.7,8 Current market forces may be changing this, but access to care for many remains limited.9 There is a need to understand all factors that may limit access to medical care for patients, including the personal values of physicians. Are there personal values that are associated with physicians who extend care to the underserved? If so, what are these values? In this study we examine the personal values of practicing family physicians, their professional satisfaction, and other characteristics of the practice, including the physicians willingness to

From the Department of Family and Community Medicine, University of Illinois College of Medicine at Rockford (Dr Eliason); and the Department of Family and Community Medicine, Medical College of Wisconsin, Milwaukee (Ms Guse and Dr Gottlieb).

out have been reported as serious problems for physicians. 1 Some reports have stated that many physicians would not choose to practice medicine again if given a choice.2,3 The increased corporatization of medicine, prolonged training, and high expectations among patients and organizations seem to be contributing to this dissatisfaction. However, surveys indicate that many physicians remain quite satisfied with their work.4 Greater understanding of factors contributing to practice satisfaction should be helpful to physicians who face many challenges in their professional career. Personal values are defined as desirable goals varying in importance that serve as guiding principles in peoples lives.5 We previously reported on the personal values of a group of exemplary family physicians and found a positive correlation between their ratings of benevolence and practice satisfaction; ie, those who were motivated to serve and help others were more satisfied with their work.6
ARCH FAM MED/ VOL 9, MAR 2000 228

ISSATISFACTION and burn-

WWW.ARCHFAMMED.COM

Downloaded from www.archfammed.com at STANFORD Univ Med Center, on November 8, 2009 2000 American Medical Association. All rights reserved.

MATERIALS AND METHODS

INSTRUMENT The Schwartz values questionnaire and instrument has been validated in multiple cultures and has been described previously.5,6,10,11 The respondent rates the importance of 56 personal values from 1 (opposed to my values) to 7 (of supreme personal importance). The value types are defined by their principal motivation and by the specific represented values. Using value types rather than single values has increased the reliability of the instrument.5 A brief definition of each value type and its dimension is included in Table 1. The value types have been organized into a 2-dimensional structure, with the self-transcendent values dimension opposing the self-enhancement values dimension, and the conservation values dimension opposing the openness to change values dimension. Schwartz has shown a dynamic relationship between the value types.5,6 Actions taken in pursuit of some values may be compatible or incompatible with other value types.10,12 The practice satisfaction portion of the survey consists of 3 questions on a semantic differential scale from 1 to 7. The stem of the 3 questions is For me the practice of medicine is . . . with available responses anchored by the following word pairs: bad-good, unfulfilling-fulfilling, and unenjoyable-enjoyable. In addition, we asked a series of questions about demographics, the respondents involvement with teaching, the scope of their practice, organizational type of practice, payor sources of the practice, and date of last recertification. SURVEY We obtained from the American Academy of Family Physicians (Leawood, Kan) 1224 physician names and

addresses that had been randomly drawn from 12 groups according to age group (25-39 years, 40-54 years, or 55-70 years), rural and nonrural status, and access to the underserved (high and low access, 102 physicians in each stratum). In the high-access group, 40% to 70% of the patients in the practice were represented by Medicare or Medicaid or were indigent, compared with the low-access group, in which 0% to 30% of the patients were represented by Medicare or Medicaid or were indigent. The confidential survey, along with a return envelope, was mailed to the physicians 3 times, about 6 weeks apart, in the summer of 1996. The second and third mailings were only to those who had not responded previously. ANALYSIS The 10 value types were computed by averaging the ratings for the associated specific values. Value-type ratings were adjusted to account for generally high or generally low rating tendencies according to the method described by Schwartz.5 A measure of satisfaction was constructed by averaging the responses to the 3 job satisfaction questions. These measures were combined, having previously shown a reliability of of .84 as determined by Cronbach .6 For the analyses, physicians providing care to the underserved were defined as those whose sum of the percentage midpoints of the response categories for Medicare, Medicaid, and indigent patients was 40% or higher. Partial correlations were used to examine the association between value types and job satisfaction while controlling for rating tendency and confounding variables such as sex and age. Analysis of variance and regression were used to find the relationship between demographic and practice patterns and value-type ratings, as well as job satisfaction and care to the underserved. Statistical analyses were done with the Stata Statistical Software package.13

care for the medically underserved. We hypothesized that there would be a positive association between the ratings of benevolence and practice satisfaction ratings among family physicians. We also hypothesized that family physicians who care for the underserved would have higher ratings of the self-transcendent values because self-transcendent values emphasize enhancing the welfare of others and oppose self-enhancing values. Furthermore, we wished to examine associations that might exist among the value-type ratings of family physicians, practice satisfaction, and a variety of practice characteristics such as location of practice (rural vs nonrural), type of practice, board certification, sex, and the age of the family physician.
RESULTS

Seven hundred twelve usable surveys were returned (58% response rate). Response rates were not significantly different among the 12 sampling strata. Table 2 gives the demographics and practice patterns. The mean valuetype ratings from highest to lowest and the correlation with practice satisfaction ratings are presented in Table 3. The benevolence value type was rated the highest and the power value type was the lowest. The association of the ratings
ARCH FAM MED/ VOL 9, MAR 2000 229

of the benevolence value type with the ratings of practice satisfaction was statistically significant (P = .002, correlation coefficient = 0.14). Both the self-determination value type and the hedonism value type had a negative or inverse association with the practice satisfaction ratings that reached statistical significance (P = .04, correlation coefficient = 0.008 and P = .05, correlation coefficient = 0.01, respectively). The power value-type ratings were also negatively associated with practice satisfaction, but the association was not statistically significant. About half (52%) of the family physicians surveyed provided 40% or more care to the underserved (Medicare, Medicaid, and indigent patients). After controlling for age and sex, rural practices (P = .005) and solo practices (P = .04) were significantly more likely to provide 40% or more care to the underserved. Tradition was the only value type to be significantly associated with service to the underserved (P = .02) when age, sex, solo practice, and rural practice were controlled. However, family physicians whose practice consisted of 30% or more indigent patients rated the universalism value type significantly higher than other family physicians (P = .03). A variety of associations were found between valuetype ratings, demographics, and practice patterns (see
WWW.ARCHFAMMED.COM

Downloaded from www.archfammed.com at STANFORD Univ Med Center, on November 8, 2009 2000 American Medical Association. All rights reserved.

Table 1. Value Dimensions, Value Types, and Definitions

Value Dimension Self-enhancement Value Type Power Definition A desire for social status and control over people and resources Personal success by demonstrating competence Pleasure or sensuous gratification for oneself Excitement and challenge in life Able to choose independent thought and action Motivation to enhance and protect all people Motivation to help those close to you Desire to maintain customs of culture and religion Restraint of actions likely to violate social norms Desire for order and stability for self and society

Table 2. Demographics of Family Physicians Surveyed

Response Rate,* % 52 29 41 9 19 3 32 67 27 8 27 61 58 78 31 51 79 15 4 52 83 31 35 34 74 1978 1949-1994 1992 1950-1996

Characteristic Rural practice Organizational arrangement of practice Solo practice Primary care group practice 2-Person practice Multispeciality group practice Unknown Member of medical school faculty Participate in teaching of medical trainees Routinely perform the following services Routine obstetric deliveries High-risk obstetric deliveries including cesarean delivery Any obstetric service Inpatient pediatrics Outpatient surgery Inpatient adult medicine Inpatient surgery Staff hospital intensive care Any inpatient care unit or critical care unit None of the above 30% Covered by indigent or charity care 40% Covered by Medicare, Medicaid, or indigent care Male sex Age group, y 25-39 40-54 55-70 Board certification within last 7 y Year graduated from medical school Median Range Year of board certification Median Range

Achievement Hedonism Openness to change Self-transcendent Stimulation Self-direction Universalism Benevolence Conservation Tradition Conformity Security

Table 4). Practice satisfaction was found to be associ-

ated with teaching medical trainees (P = .009). Those teaching medical trainees also rated the security (P = .004) and hedonism value types lower (P = .004) than those not involved in teaching. We found significant value-type associations with both sex and age, although this was not the focus of our study. Men rated the conformity value type and the tradition value type higher than women (both P.001). Women rated the universalism value type higher than men (P.001). Older physicians rated the security value type higher (P.001) and the stimulation value type lower (P = .002).
COMMENT

*The population was 712 for a 58% response rate. Sex missing in 4 cases. Year graduated missing in 14 cases; year of board certfication missing in 24 cases.

PRACTICE SATISFACTION The positive association of the benevolence value type and practice satisfaction (P.001, correlation coefficient = 0.14) is consistent with other research involving physician satisfaction and confirms our previous findings in exemplary physicians.6 In a work satisfaction survey of physicians and dentists, Lewis et al14 found that their service to humanity and direct patient care were highly satisfying elements of their professional life. Richardson and Burke15 studied stress and job satisfaction among physicians and found the major source of satisfaction to be their relationship with their patients and colleagues. Mawardi16 found high satisfaction among physicians associated with helping patients solve problems and developing relationships with patients and their families. Reams and Dunstone17 found in a qualitative study that physicians who focused mainly on patient care and less on the business aspect of medicine were more highly satisfied. Findings by Pastor et al18 were similar in a study of satisfaction among rural physicians, who derived most of their job satisfaction from patient care, while income was not a source of satisfaction.
ARCH FAM MED/ VOL 9, MAR 2000 230

The 67% of the respondents in our survey who participated in some teaching were as a group more highly satisfied (P = .009) than those who did not teach. Teaching of medical trainees is a helping activity that is consistent with the benevolence value type, although we did not find that those involved in teaching rated benevolence higher than those not involved in teaching. Hall et al19 found higher satisfaction among emergency room physicians who were involved with teaching residents. Lloyd et al,20 however, reported increased depression among Canadian emergency room physicians who were involved with medical education. Whether this was because they were overextended because of increased responsibilities is unclear. The self-direction value type was the second highestrated value type in our study, indicating that physicians value their ability to choose and be independent. Those family physicians who rated the values of self-direction highest were somewhat less satisfied (P = .04) with their practice of medicine. Many outside forces have imposed limitations on physicians decision making. Those phyWWW.ARCHFAMMED.COM

Downloaded from www.archfammed.com at STANFORD Univ Med Center, on November 8, 2009 2000 American Medical Association. All rights reserved.

Table 3. Mean Ranking of Value Types (Highest to Lowest) and Correlation With Practice Satisfaction
Value type Benevolence Self-direction Conformity Achievement Security Universalism Hedonism Tradition Stimulation Power *Partial correlation. n 706 698 704 703 700 701 704 695 699 703 Mean 5.3 4.7 4.7 4.5 4.5 4.5 4.0 3.7 3.6 2.3 Correlation* 0.14 0.08 0.06 0.05 0.02 0.02 0.07 0.03 0.04 0.06

Table 4. Practice Pattern Association* With Value Type and Practice Satisfaction Ratings
Value Type (P ) None + Conformity, .03 Universalism, .01 None None None Hedonism, .02 + Universalism, .02 Conformity, .05 + Power, .01 None + Universalism, .03 + Tradition, .02 None None Security, .004 Hedonism, .006 Self-direction, .03 + Benevolence, .04

P
.002 .04 .11 .18 .61 .67 .05 .50 .25 .12 Practice Pattern Rural practice Solo practice Primary care group practice 2-Person group practice Multispeciality group practice Routinely include in practice Obstetrics

sicians who are able to accept many of these limitations and work within existing frameworks will likely be more satisfied. Previous studies indicate that too many regulations and restrictions lead to dissatisfaction among physicians.21,22 ACCESS TO CARE The relationship between the personal values of physicians and access to care has been little explored in the medical literature. We hypothesized that physicians would provide more care to the underserved if they emphasized the self-transcendent value types compared with the self-enhancement value types depicted in the Schwartz values dimensions.5,6 In this values dimension structure, self-enhancing values oppose the self-transcendent value types. The analysis of our data did show an association of family physicians whose patients were made up of more than 40% Medicare or Medicaid recipients or recipients and/or indigent patients and the traditional value type (P = .02). The traditional value type is represented by personal values, including the following categories: humble, accepting my portion in life, devout, respect for tradition, and moderate. In the values dimension it is adjacent to the benevolence value type. A small group of family physicians in our survey whose practices consisted of more than 30% indigent patients rated the universalism value type higher (P = .03) compared with other family physicians. These findings provide general support to our hypothesis that self-transcendent value types will be rated higher by physicians who extend care to the indigent. LIMITATIONS AND OTHER ASSOCIATIONS The correlations between the value types and practice satisfaction are low and provide little explanation for the variation in practice satisfaction. Nevertheless, the associations found between practice satisfaction and the value types of benevolence, self-determination, and hedonism are consistent with other research and were statistically significant. The value type and practice pattern associations found in Table 4 are interesting and possibly meaningARCH FAM MED/ VOL 9, MAR 2000 231

Inpatient care Intensive care unit or critical care unit 30% of patients covered by indigent or charity care 40% of patients covered by Medicare, Medicaid, or indigent care Managed care 50% or more vs 20% or less Member of medical school faculty Participate to some extent in teaching medical trainees Board certified within last 7 y

*Adjusted for age and sex. Practice satisfaction was not significant for all except those who participate in teaching medical trainees ( P = .009). Analysis of variance and regression analysis are used to find significant relationships between practice patterns and value types and/or practice satisfaction, but not between value type and practice satisfaction. + indicates a positive association; indicates a negative association.

ful, but should be interpreted cautiously. About 5% of statistical association (P = .05) will occur by chance. Because we did not make prior hypotheses about specific associations in this part of the study, these associations must be interpreted more cautiously. Those associations with a lower P value (P.005) are more likely to be significant. The negative associations between the high ratings of the value types of security and hedonism with the teaching of medical trainees is statistically significant (P = .004 and P = .006, respectively) and is consistent with the Schwartz values dimension.5,6 Teaching is generally an activity for which there is less personal economic gain and that involves actions that assist others. Thus, one might expect a negative association with the value types of hedonism and security. A possible limitation in this study is the potential for self-reporting bias. A written score on a questionnaire does not necessarily translate to actual deeds. The confidentiality of the questions would weigh against this because the family physicians had no apparent personal gain. Schmitt et al11 have also demonstrated 6-week test/ retest reliability of the value type indices, ranging from 0.70 to 0.90. The definition of providing care to the underserved in this study resulted in an underserved group that included Medicare patients. The literature demonstrates that Medicare patients who are white, educated, upper class, and have supplemental insurance are not underserved.23 There is evidence, however, that signficant segments of the Medicare population are underserved.
WWW.ARCHFAMMED.COM

Downloaded from www.archfammed.com at STANFORD Univ Med Center, on November 8, 2009 2000 American Medical Association. All rights reserved.

This includes the following: (1) those without supplemental insurance,23,24 (2) those from core urban and rural areas,23,25,26 (3) minority elders such as African Americans,23,25,26 (4) low-income, and sometimes middleincome Medicare recipients,23,26,27 (5) elders whose mobility is impaired,28 and (6) those who are less well educated.23 Because we were not able to further identify these subgroups of Medicare patients, the access to care for this group may range from poor to good. For these reasons, our ability to find a significant relationship between the self-transcendent values and providing care to the underserved may have been reduced.
CONCLUSION

tion and the benevolence value-type rating have consistently shown a positive association at least in primary care specialties. Finally, additional research would be helpful to further clarify the relationships that might exist among personal values, practice satisfaction, and the provision of medical care to the underserved. Accepted for publication July 16, 1999. Reprints: B. Clair Eliason, MD, Department of Family and Community Medicine, University of Illinois College of Medicine at Rockford, 1601 Parkview Ave, Rockford, IL 61107-1897.
REFERENCES
1. Musick JL. How close are you to burnout? Fam Pract Manage. 1997;4:31-46. 2. Lewis CE, Prout DM, Chalmers EP, Leake B. How satisfying is the practice of internal medicine? Ann Intern Med. 1991;114:1-5. 3. Greenberg B. Life balance isnt that easy. Fam Pract. 1997;4:16. 4. Chuck JM, Nesbitt TS, Kwan J, Kam SM. Is being a doctor still fun? West J Med. 1993;159:665-669. 5. Schwartz SH. Universals in the content and structure of values: theoretical advances and empirical tests in 20 countries. In: Zanna MP, ed. Advances in Experimental Social Psychology. Vol 25. Orlando, Fla: Academic Press Inc; 1992: 1-65. 6. Eliason BC, Schubot DB. Personal values of exemplary family physicians: implications for professional satisfaction in family medicine. J Fam Pract. 1995;45: 251-256. 7. Bartman BA, Moy E, DAngelo LJ. Access to ambulatory care for adolescents: the role of a usual source of care. J Health Care Poor Underserved. 1997;9:214226. 8. Patrick DL, Stein J, Porta M, Porter CD, Ricketts TC. Poverty, health services and health status in rural America. Milbank Q. 1988;66:105-136. 9. Cornelius LJ. The degree of usual provider continuity for African and Latino Americans. J Health Care Poor Underserved. 1997;8:170-185. 10. Schubot DB, Cayley Jr W, Eliason BC. Personal values related to primary care specialty aspirations. Fam Med. 1996;28:723-728. 11. Schmitt MJ, Schwartz SH, Steyer R. Measurement models for the Schwartz values inventory. Eur J Psychol Assess. 1993;9:107-121. 12. Schwartz SH, Verkasalo M, Antonovsky A, Sagiv L. Value priorities and social desirability: much substance, some style. Br J Soc Psychol. 1997;36:3-18. 13. Stata Corp. Stata Statistical Software: Release 5.0. College Station, Tex: Stata Corp; 1997. 14. Lewis JM, Barnhart FD, Howard BL, Carson DI, Nace EP. Work satisfaction in the lives of physicians. Tex Med. 1993;89:54-61. 15. Richardson AM, Burke RJ. Occupational stress and job satisfaction among physicians: sex differences. Soc Sci Med. 1991;33:1179-1187. 16. Mawardi BH. Satisfactions, dissatisfaction and causes of stress in medical practice. JAMA. 1979;214:1483-1486. 17. Reams HR Jr, Dunstone DC. Professional satisfaction of physicians. Arch Intern Med. 1989;149:1951-1956. 18. Pastor WH, Huset RA, Lee MC. Job and life satisfaction among rural physicians: results of a survey. Minn Med. 1989;72:215-223. 19. Hall KN, Wokeman MA, Levy RC, Khoury J. Factors associated with cancer longevity in residency-trained emergency physicians. Ann Emerg Med. 1992;21: 291-297. 20. Lloyd S, Streiner D, Shannon S. Burnout, depressions, life and job satisfaction among Canadian emergency physicians. J Emerg Med. 1994;12:559-565. 21. Arnetz BB. Physicians view of their work environment and organization. Psychother Psychosomat. 1997;66:155-162. 22. Skolnik NS, Smith DR, Diamond J. Professional satisfaction and dissatisfaction of family physicians. J Fam Pract. 1993;37:257-263. 23. Blustein J, Weiss LJ. Visits to specialists under Medicare: socioeconomic advantage and access to care. J Health Care Poor Underserved. 1998;9:153169. 24. Blustein J, Hanson K, Shea S. Access to Health Care: Part 3: Older Adults. Hyattsville, Md: National Center for Health Statistics; 1997. Data from the National Health Survey, Series 10. 25. Millar JS, Scheffler SA, Murray CK, Bratzler DW. Comparison of influenza immunization rates for Oklahoma Medicare patients: 1995, 1996 and 1997. J Okla State Med Assoc. 1998;91:509-513. 26. Moy E, Hogan C. Access to needed follow-up services: variations among different Medicare populations. Arch Intern Med. 1993;153:1815-1823. 27. Lee AJ, Gehlbach S, Hosmer, Reti M, Baker CS. Medicare treatment differences for blacks and whites. Med Care. 1997;35:1173-1189. 28. Miller RH. Access to ambulatory care among noninstutionalized, activity-limited persons 65 and over. Soc Sci Med. 1992;34:1237-1247.

This study found associations between practice satisfaction and teaching medical trainees, as well as between practice satisfaction and ranking of the benevolence value type or the self-direction value type. Two value types were found to be associated with providing care to the underserved, universalism and tradition. These associations raise interesting questions as to cause and effect that cannot be answered definitively by a cross-sectional study such as this one. Does the involvement in teaching contribute to practice satisfaction or do satisfied physicians gravitate toward teaching? Are values mutable and, if so, would changes toward increased benevolence or self-direction increase professional satisfaction? Or could the problem of underserved populations be addressed by a shift in physician values toward universalism and tradition? These questions will challenge future researchers. At this juncture, practicing physicians who are dissatisfied with their work may benefit by identifying values that are important to them and looking for positions that will allow them to follow those values. It might also behoove them to become more involved in providing leadership at all levels within health care organizations and thereby have a hand in shaping the environment in which they work. Most would agree that we need more physicians who are satisfied with their work and who have motivation to be of service to their patients and to help provide care to the underserved of our society. This study suggests that such physicians would give high ratings to the selftranscendent values. Practicing physicians who are dissatisfied with their work may benefit by focusing on activities that involve helping and serving their patients, not on business and financial issues. The current medical climate, in which practice guidelines, nonphysician case managers, and regulatory efforts are prevalent, will likely lead to increased dissatisfaction among physicians who rate the self-direction value type high. Medical organizations that employ physicians may find that their physician employees are more satisfied if they are not overregulated. Family physicians in turn need to be more assertive in providing leadership in health care organizations at all levels or they may find themselves increasingly dissatisfied as they lose their autonomy. We have previously recommended6 and recommend again consideration of personal values in the selection of medical students, because practice satisfacARCH FAM MED/ VOL 9, MAR 2000 232

WWW.ARCHFAMMED.COM

Downloaded from www.archfammed.com at STANFORD Univ Med Center, on November 8, 2009 2000 American Medical Association. All rights reserved.