Marriage of one physician to another, which creates a so-called dual-doctor family (8), probably has substantial effects. Conflicts between work and family roles are common for physicians (9, 10). These conflicts may be compounded in dual-doctor families because medicine has traditionally demanded single-minded dedication regardless of whether either spouse has a "full-time back-up for the rest of life's activities" (2). Studies of nonphysicians in dual-earner marriages suggest that conflicts between work and family roles may be substantial for both spouses (11, 12) and that women may be affected disproportionately (13). Studies of physicians have suggested that the effects of marriage to another physician are complex (14-20). For example, women in dual-doctor families are more likely than other female physicians to interrupt their careers (14, 19), but both partners in a dual-doctor family may also enjoy greater affluence as well as interpersonal support from shared professional experiences (2, 14, 17). Past studies, however, have not comprehensively examined the effects of dual-doctor marriage in general or in recently trained physicians.

In this study, we explored how marriage to another physician affects younger physicians. We sought to learn how physicians in dual-doctor families differ from other physicians in their professional and family lives and their perceptions about career and family and to determine whether these differences were more pronounced for women than men. In a survey of a large, diverse group of married physicians, we addressed three areas: number of hours worked and annual income, number of children and child-rearing arrangements, and perceptions of work and family.

Methods

Top Methods Results Discussion Author & Article Info References

Study Sample and Survey Methods

We surveyed 2200 physicians who graduated between 1980 and 1990 from the medical schools of Case Western Reserve University and the University of Cincinnati. By using computer-generated random numbers, 100 physicians were selected from each of the 11 classes of each school. This sample included physicians residing in every state except South Dakota; 39% resided in Ohio, 7% resided in California, and less than 5% lived in each of the remaining 47 states. Sixty-eight physicians (3%) had the same address as another physician in the sample. Surveys were sent by mail to all physicians in May 1995, and three mailings went to nonrespondents during the subsequent year. The survey was approved by the institutional review board of University Hospitals of Cleveland. We were unable to contact 122 of the 2200 physicians by mail or telephone. We excluded the responses of physicians who reported that they had graduated before 1980 or after 1990 (n = 11) or were born before 1950 (n = 67). The remaining 2000 physicians form the study sample.

Survey Measures

We composed a comprehensive questionnaire to inquire about physicians' professional and family lives. A preliminary version of the questionnaire was evaluated by faculty and postgraduate fellows at Case Western Reserve University and was revised on the basis of their comments.

The final questionnaire contained 76 close-ended items. Individual items inquired about sociodemographic characteristics and physicians' professional experience, including practice setting, the average number of hours worked per week, and personal income in 1995. Income was characterized by five response categories in $50 000 increments.

Physicians' perceptions about their careers and how they were affected by spouse and family were addressed by several items (Appendix). Items inquired about perceived career success, precedence of the respondent's career relative to the spouse's career, sharing of work interests, perceived conflict between their professional and family lives, and the respondent's view of how his or her work was affected by family.

The occupation and work experience of spouses were characterized according to the spouse's baccalaureate and postbaccalaureate degrees, average number of hours worked per week outside the home, 1995 income, and the respondent's estimate of success in achieving the spouse's career goals. Parenting experiences were characterized by items about the number of children, arranging work schedule to fulfill childcare responsibilities, perceived role as "primary caregiver with respect to child-rearing," and success in meeting the respondent's goals for his or her children's welfare.

Statistical Analysis

We sought evidence of the construct validity of key measures by testing expected associations. Strong associations were found (P < 0.001) in expected directions between such variables as hours worked and personal income, hours worked and perceived sacrifice in hours worked because of family considerations, and being the primary caregiver with respect to child-rearing and arranging work schedules to fulfill childcare responsibilities. For example, 68% of respondents who worked fewer than 40 hours per week earned less than $100 000 per year compared with 27% of those who worked 40 or more hours per week.

To evaluate the possibility of nonresponse bias, we contacted 88 randomly selected physicians who did not respond to our survey and asked them to respond by telephone to 15 items from the full questionnaire. To examine the generalizability of our findings, we compared selected sociodemographic and professional characteristics of the study sample to those of the 213 966 physicians listed in the American Medical Association Physician Masterfile for 1996 and who were born after 1949 and graduated from medical school between 1980 and 1990.

Most analyses were restricted to married physicians because we were primarily interested in the effects of being married to another physician rather than being married to a nonphysician. Because gender affects many professional and family roles (3, 6, 7, 10, 14, 17), and because we found strong associations between gender and many characteristics of professional and family life, analyses were conducted separately for male physicians and female physicians. Differences between groups were evaluated by using the chi-square test for categorical variables and the Kruskal-Wallis test for continuous variables. To simplify presentation of results in tables, response categories were combined when this maneuver did not alter interpretation of the relevant item.

Results

Top Methods Results Discussion Author & Article Info References

Surveys were completed by 1208 (60%) of the 2000 eligible physicians. The respondents' mean age was 37.5 years (range, 30 to 45 years), 62% of respondents were men, and 90% were white. Respondents represented a diverse array of specialties, about half were in private practice, and most reported annual personal incomes of $100 000 or more (Table 1). Twenty-six percent of respondents were married to other physicians, 32% were married to nonphysicians working outside the home, and 28% were married to nonphysicians at home; 10% had never married, and 4% were divorced or separated.

Male physicians and female physicians differed in many ways (Table 1). Compared with male physicians, female physicians were more likely to be married to other physicians, worked fewer hours, and earned less money. Female physicians were more likely than male physicians to be primary or equal caregivers for their children and to have arranged their work schedules to care for children. They were also less likely than male physicians to report success in achieving their career goals. Although most married male and female physicians often felt conflict between their professional and family lives, few male physicians reported substantial limitations in professional life for family reasons. In contrast, 25% of female physicians reported substantial limitations.

Specialty, Hours Worked, and Income

Physicians in dual-doctor families did not differ significantly in specialty from other physicians of the same sex (data not shown). Female physicians in dual-doctor families worked fewer hours than female physicians married to nonphysicians (mean difference, –3.78 hours [95% CI, –7.52 to –0.04 hours]). The hours worked by male physicians did not differ according to whether they were married to a physician (Table 2). Both men and women in dual-doctor families reported that their physician spouses worked more hours than the nonphysician spouses. In addition, physician couples worked more total hours than one-physician couples. Both men and women in dual-doctor families had lower personal incomes than physicians married to nonphysicians, but men and women in dual-doctor families had spouses with higher incomes; thus, their total family incomes were substantially higher than the family incomes of physicians married to nonphysicians.

Number of Children and Child-rearing Arrangements

Among male physicians, those with a nonphysician spouse at home had the most children, but among female physicians, those in dual-doctor families had the most children (Table 3). Although caregiving responsibilities for children were predominantly borne by women, both male and female physicians in dual-doctor families played greater roles in the care of their children.

Perceptions of Work and Family

Physicians in dual-doctor families did not differ significantly from other physicians of the same sex in believing themselves to be successful in achieving their career goals; for all groups, the percentage of respondents who believed that they were successful ranged from 69% to 88% (P > 0.05) (Table 4). In addition, compared with other physicians and their spouses, men and women in dual-doctor families more often reported that their spouses were successful in achieving career goals. Finally, almost all physicians ( 90% regardless of marital group or sex; P > 0.2) reported success in achieving their goals for their children's welfare.

Compared with female physicians who were not married to other physicians, women in dual-doctor families more often reported substantial limitations in professional life for family reasons; the percentage of women who reported such limitations was 33% for female physicians married to physicians, 19% for female physicians married to nonphysicians employed outside the home, and 15% for female physicians married to nonphysicians at home (P = 0.02). Such limitations were rarely reported by male physicians, regardless of whether they were married to physicians. Nonetheless, frequent feelings of conflict between professional and family lives were equally common for all groups of physicians (the percentage of respondents who reported feeling conflict ranged from 48% to 65%; P > 0.2).

Most male and female physicians reported enjoyment and satisfaction in discussing and sharing work interests with their spouses, but this finding was most frequent for physicians in dual-doctor families (Table 4). Both men and women in dual-doctor families were also most likely to gain personal career advancement from sharing work interests with their spouses, and they least often reported that their career had taken precedence over their spouse's career.

Evaluations of Generalizability and Nonresponse Bias

Modest but statistically significant differences were found (P < 0.001) between respondents and the national sample in sex (38% women compared with 29% women, respectively), mean age (38.5 years compared with 38.0 years), ethnicity (90% white compared with 82% white), and specialty (21% compared with 15% practicing internal medicine and 11% compared with 7% practicing pediatrics). These differences between the respondents and the national sample were similar for male and female physicians.

In our evaluation for nonresponse bias, usable responses were obtained from 68 of the 88 randomly selected nonrespondents (77%). Nonrespondents and respondents to the initial survey were similar in terms of mean age (38.4 years and 38.5 years, respectively) and sex (35% and 38%, respectively, were women). Nonrespondents were less likely to be married (77% compared with 88% of respondents; P = 0.01); among married physicians, however, nonrespondents and respondents did not differ significantly (P > 0.2) in many other characteristics, including the proportion of respondents who were married to another physician, average number of children, specialty, personal income, and type of practice.

Discussion

Top Methods Results Discussion Author & Article Info References

In our survey, 22% of male physicians and 44% of female physicians were married to other physicians. Both men and women in dual-doctor families differed in their professional and family lives from other married physicians; physicians of either sex earned less money, more often played a major role in child-rearing, and more often arranged their work schedules to fulfill childcare responsibilities. Compared with physicians with nonphysician spouses, women (but not men) in dual-doctor families worked fewer hours and more often limited their income, hours, and availability to patients because of family. Nonetheless, neither men nor women in dual-doctor families reported being significantly less successful than other physicians in achieving their career goals or their goals for their children's welfare. In fact, marriage to another physician had distinct benefits, including more frequent enjoyment and satisfaction from shared work interests, greater success in achieving spouses' career goals, and higher family income.

In general, physicians in dual-doctor families differed from other physicians regardless of whether the other physicians' spouses worked outside the home. Thus, our findings probably indicate effects specific to marriage between physicians as well as effects attributable to both spouses working (21). Not surprisingly, however, differences were often greatest between physicians in dual-doctor families and physicians married to nonphysicians at home. For example, for both male and female physicians, those married to nonphysicians at home had the highest personal incomes and were the least involved in child-rearing, whereas those in dual-doctor families had the lowest personal incomes and were the most involved in child-rearing.

Many of our findings undoubtedly reflect choices made by the physician respondents that are influenced by different degrees of constraint and freedom. Physicians in dual-doctor families may have felt constrained by their spouses' careers, so they chose positions with more limited responsibilities and less compensation in order to assume childcare and other family responsibilities. Another possibility is that physicians married to physicians may have felt free to make choices they desired (for example, to devote more time to family) because their spouses' income and sharing of family responsibilities made it easier for them to achieve such goals. This interpretation is consistent with our finding that both men and women in dual-doctor families were no more likely than other physicians to report frequent conflict between their professional and family lives, even though they were more likely to arrange their work schedules to fulfill childcare responsibilities. Our study did not determine the relative importance of constraint versus freedom, and both may apply for many physicians.

The Importance of Gender

Like many previous studies (6, 7, 10, 14, 20, 22-37), our findings highlight substantial differences between young male and female physicians in almost every aspect of professional and family life (and their perceptions about the two) that we examined, differences that remain despite greater parity in hourly wages (6). In fact, these differences between male and female physicians were greatest for women in dual-doctor families: Compared with men and with women whose spouses were not physicians, women in dual-doctor families worked the fewest hours, earned the least money, made the greatest professional adjustments for child-rearing, and most often reported limitations in their professional lives because of family.

Limitations

Potential problems with the internal validity of our findings may have resulted from nonresponse bias, inaccuracy in physicians' responses to our survey, and chance. We believe that nonresponse bias is unlikely because a survey of a random sample of nonrespondents showed that married nonrespondents were similar to married respondents. To provide evidence of the accuracy of physicians' responses to our survey, we sought and found expected associations between survey items (that is, evidence of construct validity). Finally, it is unlikely that chance accounted for many of the associations that we observed: The associations were consistent, and most had P values below 0.001. The study's statistical power to detect differences was limited, however. For example, women in dual-doctor families least often felt successful in achieving their career goals (Table 4); although this difference was not statistically significant, it may be important and merits further study.

The generalizability of our findings to other physicians has not been established. Our sample of recent graduates from two medical schools differed somewhat from U.S. physicians in the American Medical Association Physician Masterfile in age, ethnicity, and specialty. However, the differences were modest to inconsequential (such as age) and were unlikely to have affected our major findings.

Conclusions

Three conclusions based on our findings stand out. First, the results of our study can inform the choices that physicians make. Young physicians may be interested to learn that marriage between physicians is associated with higher family income, more enjoyment and satisfaction from shared professional interests, and more involvement in child-rearing for both partners. However, even in dual-doctor families, the professional and family lives of both male and female physicians continues to reflect dominant gender roles. For female physicians in particular, marriage to another physician often involves choices involving limitations in personal income and professional life in favor of filling family roles. Although these choices may, in fact, be possible because of the greater income of a dual-doctor marriage, they may also highlight the strength of normative societal pressures; the resulting professional limitations merit consideration and discussion.

Further studies of physicians in dual-doctor families should be performed. In particular, qualitative research would illuminate the tensions and tradeoffs between choices and constraints, would uncover the extent to which normative societal pressures—stereotyping and expectations from family, colleagues, or superiors—have shaped professional compromises and would help determine physicians' overall satisfaction with professional and personal arrangements.

Finally, viewed in the context of the changing demography of medicine, our findings may also inform policies and plans for physician training and employment. Although some physicians may maintain conventional expectations for long hours and high incomes, many young physicians have chosen to work fewer hours, to earn less money, and to seek greater flexibility in their work to fulfill child-rearing roles. Physician training programs (38), workforce policy, and public expectations should recognize and adapt to the diverse and changing needs of young physicians to balance professional and family roles.

Appendix

Survey Items Inquiring about Respondents' Views of Their Professional Life and the Interaction of Their Professional Life and Family Life

1. In terms of your goals, how well have you succeeded in your career?

(Response options: a 5-point scale ranging from "total success" to "total failure.")

2. Which of the following statements best applies to your situation? (Choose one.)

My career has taken precedence over my spouse's/partner's career.

My spouse's/partner's career has taken precedence over mine.

Neither of our careers has taken precedence.

3. My spouse's and my shared professional/work interests have advanced my career.

(Response options: not at all, a little, a fair amount, or very much.)

4. My spouse's and my shared professional/work interests have been a source of satisfaction for me.

(Response options: not at all, a little, a fair amount, or very much.)

5. My spouse/partner and I enjoy discussing professional/work issues.

(Response options: not at all, a little, a fair amount, or very much.)

6. How often do you feel conflict between your professional and nonprofessional lives?

(Response options: never, rarely, sometimes, often, or all the time.)

7. To what degree have you made sacrifices in the number of hours you work for the sake of your marriage/partnership/children?

(Response options: not at all, a little, a fair amount, or very much.)

8. To what degree have you made sacrifices in your earned income for the sake of your marriage/partnership/children?

(Response options: not at all, a little, a fair amount, or very much.)

9. To what degree have you limited your availability to patients for family reasons?

(Response options: not at all, a little, a fair amount, or very much.)