|Home | About | Journals | Submit | Contact Us | Français|
To evaluate personal and professional factors associated with marital and parental satisfaction of physicians.
A survey was sent to equal numbers of licensed male and female physicians in a Southern California county. Of 964 delivered questionnaires, 656 (68%) were returned completed. Our sample includes 415 currently married physicians with children, 64% male and 36% female.
Ratings of marital and parental satisfaction were measured on a 5-point Likert scale, 5 being extremely satisfied. Prevalence of work and home life factors was also evaluated. The mean score for marital satisfaction was 3.92 (range 1.75–5.0). Approximately half of the physicians reported high levels of marital satisfaction (63% of male physicians and 45% of female physicians). The gender difference disappeared after adjusting for age differences. Two factors were associated with high marital satisfaction: a supportive spouse (odds ratio [OR] 10.37; 95% confidence interval [CI] 2.66, 40.08) and role conflict (OR 0.61; 95% CI 0.42, 0.88). The mean score for parental satisfaction was 3.43 (range 1.0–5.0), and approximately two thirds of both male and female physicians reported at least moderate levels of parental satisfaction. The major factors associated with parental satisfaction were a supportive spouse (OR 2.24; 95% CI 1.32, 3.80), role conflict (OR 0.35; 95% CI 0.23, 0.53), salaried practice setting (OR 2.14; 95% CI 1.21, 3.81), marriage to a spouse working in a profession (OR 2.14; 95% CI 1.21, 3.81), and marriage to a spouse working as a homemaker (OR 2.33; 95% CI 1.20, 4.56). Number of hours worked was not found to be related to either satisfaction score, but rather to an intervening variable, role conflict.
For physicians with children, our study indicates that minimizing the level of role conflict and having a supportive spouse are associated with higher levels of marital and parental satisfaction. Working in salaried positions and marriage to a spouse who is either working in a profession or who is a stay-at-home parent are also related to high parental satisfaction.
Physicians can lead very full lives combining their roles as caregiver, spouse, and parent. Multiple roles provide the opportunity to be personally satisfied in many aspects of their lives simultaneously, and difficulty in one area can be compensated by success in another. Although there is rarely time for boredom, the demanding nature of a physician’s work is bound to influence family life and may lead to strained family relationships or altered career paths.1, 2 In the case of married physicians with children, the situation may be exacerbated by a working spouse with a second set of career demands and the need for both parents to be actively involved in childrearing.
Little is written about the professional and personal factors that may influence the relationships of physicians with their spouse and children. To our knowledge, no studies have evaluated physicians’ relationships with their children. Divorce rates may indicate the quality of marital relationships. Among physicians, divorce occurs less often than it does among nonphysicians,3, 4 and female physicians may be more at risk of divorce than male physicians.3–5 However, physicians may stay in unhappy marriages for other reasons, such as financial need or social status.6 Marital satisfaction scales have been used to explore the quality of physicians’ marriages, but the relation between family life and work is complicated and difficult to study.6–10 A small survey of family practice residents found a correlation between marital satisfaction and psychological stressors.7 Two studies of physicians focused on sources of conflict in medical marriages.6, 8 A more recent study evaluated a broader range of predictors of marital satisfaction, but was hampered by a low response rate.9
In this cross-sectional study of married Southern California physicians with children, our purpose was to ascertain which personal and professional factors are associated with marital and parental satisfaction. This information will be useful to medical students and young physicians choosing career paths, to physicians grappling with how best to combine work and family, and to health services administrators as they seek to improve productivity and retain qualified physicians.
In 1988 a questionnaire was mailed to all female physicians (706) and an equal number of systematically selected male physicians (every sixth) from a Southern California county, as they were listed in a roster of licensed physicians from the California Department of Consumer Affairs. Subjects were selected in this manner because we were interested in studying gender differences and because males outnumbered females in a ratio of approximately 6 to 1. A follow-up mailing was sent 4 weeks later to nonresponders. All questionnaires were coded to ensure confidentiality. The Department of Consumer Affairs provided the age and sex of physicians who were nonresponders (those who received the questionnaire, but did not return it) and nonreceivers (those whose questionnaires could not be delivered by the postal service).
Physicians were included in this study if they had a correct mailing address and reported being currently married with children. The survey questions covered seven general areas: professional characteristics; spouse characteristics; family characteristics; household help characteristics; career changes for marriage and children; and attitudes—role conflict, marital satisfaction, and parental satisfaction.
Marital satisfaction was evaluated with a series of eight questions inquiring about the quality of the marriage, and each question was scored on a 5-point Likert scale (5 indicating high satisfaction). This instrument was selected because it had previously been used and validated among physicians.7, 11 Parental satisfaction was evaluated using three similar questions about the quality of the respondent’s relationship with his or her children. We judged this instrument to have face validity on the basis of extensive literature review and the opinion of experts in the field. For both satisfaction scales, factor analysis indicated that the items in each scale comprised one factor. Internal consistency was assessed and revealed an α score of .95 for the marital scale and .76 for the parental scale.
We used χ2, Student’s t, and two-group Wilcoxon tests to assess gender differences in the variables we studied, depending on the type and distribution of the variable under consideration. Owing to the multiple comparisons being made, we chose p≤ .01 as the level of significance for bivariate analyses. All possible subsets regression was used to screen for important predictors for each of the two outcome variables—marital and parental satisfaction. We included the following variables as predictors in each analysis: age, gender, specialty type, work hours, annual combined income, practice setting, spouse occupation, spouse work hours, spouse support for the physician’s career, spouse household responsibilities, presence of household assistance, number and ages of children, and frequency of felt frustration. The final regressions included variables selected for the best subset according to Mallow’s Cp. Mallow’s Cp is a direct function of the residual sum of squares that incorporates a penalty for each variable added to the model to protect against overfitting.12 The final regressions, which controlled for important covariates, consisted of two multiple logistic regression analyses, one modeling high marital satisfaction and the other modeling high parental satisfaction as the reference groups. We chose to dichotomize these into high versus not high marital and parental satisfaction because of the skewed distribution of these satisfaction scales.
Of the 1,412 questionnaires that were mailed, 964 (68%) were delivered, and the remaining 448 (32%) were not received because of incorrect address. Of the 964 questionnaires delivered, 656 (68%) were returned either partially or totally completed (responders). Fifty-three percent of respondents were male; 85% of these men were married, and 90% had children. Of those women responding, 64% were married and 73% had children. Among these 656 physicians, 492 (75%) were currently married, and 415 of these married physicians had children. This latter group formed our study sample. To evaluate for bias, we compared the age and gender of our sample with those of both the nonreceiver and nonresponder groups and found no differences with the exception that responders were older than nonreceivers (mean respective age 47 and 40 years, p < .001). To evaluate whether our sample was sociodemographically representative of all physicians in this country, we compared our data with the American Medical Association (AMA) database for age-gender categories and selected specialties,13, 14 and found our sample to be similarly distributed to the AMA sample (Table 1)
Gender differences in descriptive characteristics are shown in Table 2. The women physicians in our sample were younger than the men. Similar proportions of men and women physicians were currently working, with female physicians being more likely than their male peers to work part-time (less than 40 hours per week) and less likely to work long hours (more than 60 hours per week). We also found differences in the distribution of specialties practiced by gender. The most notable differences were that women were more likely than men to practice pediatrics, while men were more likely to be surgeons. Among working physicians, women were more likely than men to be practicing in settings other than private practice and to earn less money.
The characteristics of spouses varied by the gender of the physicians in our sample. For male physicians, half of the spouses did not work, and of those who did work, only 12% worked full-time. In comparison, nearly all of the spouses of female physicians worked (95%), and the majority of them (59%) worked full-time. Regarding household responsibilities assumed by physicians’ spouses, 82% of the male physicians’ spouses performed most or all of the duties at home compared with 5% of the female physicians’ spouses. In general, the majority of physicians in our sample received high levels of support for their careers from their mates, although the male physicians were more likely to receive higher levels of support than were the female physicians. The majority of all physicians employed some type of help at home, but more women than men did so.
Female physicians had significantly fewer children than male physicians, and the median number of children was 2 for all physicians (not shown). The male physicians were more likely to have older children than the female physicians.
Our measure of role conflict (frustration from the competing demands of career, marriage, and family) indicated that female physicians were more likely to experience moderate to high levels of frustration than were male physicians (87% vs 62%).
Table 3 presents the results of our two major dependent variables, marital and parental satisfaction. These scores are both shown by approximate tertiles. The ranges of scores represented by these tertiles are not exactly the same because the physicians in our sample reported somewhat higher scores for marital satisfaction (mean 3.92, range 1.75–5.0) than for parental satisfaction (mean 3.43, range 1.0–5.0). A moderate association was found between marital and parental satisfaction (r= .48). Women physicians were less satisfied with their marriages than the men, but this difference disappeared after stratifying for age (not shown). There were no significant gender differences in mean parental satisfaction scores.
The results of bivariate analysis between the two outcome variables of marital and parental satisfaction and possible related factors are shown in Table 4. Greater marital satisfaction was associated with greater spousal support for career, higher spouse work hours, and less role conflict. Parental satisfaction was associated with practice type (working in HMO or other settings, and not working in solo or group practice), spouse occupation as a professional or homemaker, higher spouse work hours, greater spousal support for career, and less role conflict. Neither marital nor parental satisfaction was related to physicians’ gender, age, specialty, work hours, income, household responsibilities, presence of household help, or number of younger children. In addition, marital satisfaction was not related to practice type or spouse occupation.
Tables 5 and and66 summarize the results of multiple logistic regression analysis for the whole sample and for men and women separately. Variables for which no results are shown were not included in the final runs. In Table 5, high marital satisfaction was associated with a high level of career support from one’s spouse and a lower level of role conflict. Separate analyses for men and women revealed the same results.
In Table 6, high parental satisfaction was associated with role conflict and with being younger than 45 years for all groups analyzed. Spouse support for career and marriage to a spouse who is a working professional were associated with higher family satisfaction in the analyses for the whole group and for female physicians. Similar trends were noted among male physicians, but these were not statistically significant. We found three other factors to be associated with high parental satisfaction only in the analysis for the whole group: practice setting other than private practice (mostly salaried positions), marriage to a homemaker, and marriage to a spouse who contributes little to household responsibilities.
In this study of married physicians with children, we evaluated two important aspects of a physician’s life, marriage and parenthood. Our study is one of a few examining the effect of work and home life on marital satisfaction of physicians and the first to examine parental satisfaction of physicians.
Approximately half of the physicians in our sample were extremely or very satisfied with their marriages. Similar ratings were found among a group of Texas physicians. In that same sample of physicians, those who reported high levels of marital satisfaction also had higher levels of work satisfaction, lower levels of work stress, and higher levels of family competence.15 The links between marital satisfaction and work life issues suggest that maintaining a healthy balance is good for professional life as well as family life.
High marital satisfaction was associated with two factors: a supportive spouse and lower levels of role conflict. These results are similar to what is found in other research studies of marital satisfaction among physician and non-physician groups.7, 16, 17 The perception of a spouse’s ability to give and receive support has been shown to predict marital quality for both partners in a marriage.18
In this study role conflict is defined as “the perceived frustration resulting from the competing demands of career, marriage, and family.” We found role conflict to affect marital satisfaction in an inverse relation: i.e., lower levels of role conflict strongly predicted higher levels of marital satisfaction. However, we did not find physician work hours to influence marital satisfaction. Although this may seem contrary to expectations, others have also found that it is not the number of hours worked that affects satisfaction with marriage,9 but rather the role conflict experienced by the physician.8 In a group of practicing physicians, Gabbard et al. found no difference in marital satisfaction between physicians who worked 40 to 50 hours per week and those who worked 60 or more hours. However, the group working longer hours more frequently complained about “the lack of time for fun, family, and self.”8
We hypothesize that role conflict is an intervening variable between marital satisfaction and many other factors related to gender, age, work, and home life. These numerous factors appear to affect the level of role conflict experienced by a working parent. In other studies we and others have found that women are more likely to experience role conflict than men,19–25 and younger male physicians are more likely to do so than older male physicians.25 Role conflict in working parents has been associated with excessive hours at work, scheduling, physically or psychologically demanding work, housework inequality, and child care difficulty.2, 22, 26, 27
Interventions in the workplace that decrease role conflict such as vacation time, flexible work hours, and equitable part-time employment opportunities will assist physicians who are struggling to combine work and family. Many studies of physician and nonphysician working parents have shown that reasonable work hours and a work environment flexible to family responsibilities improve not only family life satisfaction, but also mental health, job satisfaction, and work productivity.2, 22, 26–28 Health services administrators should make these options available for physicians as they seek to improve productivity and retain qualified personnel. It is a win-win situation for administrators as well as for physicians and their families.
The physicians in our sample tended to be less satisfied with their roles as parents than with their marriages. Because childrearing requires continuous parental attention as children develop physically and emotionally, for many couples, raising children may be more complex and demanding than maintaining their marital relationship. Again we found lower role conflict and a supportive spouse were strongly related to parental satisfaction. These results place even more emphasis on the importance of making family-friendly work options viable for physicians.
The location and type of medical practice also appear to affect a physician’s relationship with his or her children. Physicians in salaried positions, such as in staff-model HMO and Veterans Administration hospital settings, more frequently reported high parental satisfaction than did physicians in private fee-for-service practice. This may be because hours, call schedules, and income associated with salaried positions are more predictable. Some women physicians prefer working arrangements such as part-time and salaried positions so that they can spend more time with their children.29–31 Of note, the predominant concept of HMO when this survey was conducted was the staff model and not the more contemporary concept of “HMO without walls.”
This study also indicates that the occupation of a physician’s spouse may make a difference in parental satisfaction. Physicians married to a professional parent more commonly experienced high parental satisfaction than did physicians whose spouses worked at nonprofessional jobs. For the dual-career couple, the financial needs of a family are often better accommodated by two incomes than by one. The psychological burden of being the “sole breadwinner” may be eased, thereby freeing up emotional energy for the children. In addition, the professional spouse may be more intellectually satisfied than the nonprofessional and more apt to enrich the home environment with new ideas and unique experiences.32
Physicians who were married to a stay-at-home parent more commonly experienced high parental satisfaction than did physicians whose spouses worked at nonprofessional jobs. For those physicians whose spouse is a homemaker, improved parental satisfaction may result from the parents having their separate, well-defined roles. However, as only 1% of our female sample was married to homemakers, our results may not accurately represent this group.
It is noteworthy that we did not find specialty related to marital or parental satisfaction. This may be because both men and women medical students choose specialties with family considerations in mind,33 thereby minimizing the effect of specialty on marital or parental satisfaction of working physicians.
In the intervening time between when the study was conducted and this publication, managed care has become more prevalent in most parts of the country and the proportion of practicing physicians who are women has increased. Early results from the Physician Worklife Study in l997 indicate that with managed care has come more time pressure and less job satisfaction.34 Women in particular report more job stress, lower perceived wellness, and more burnout than men. Insufficient organizational attention paid to personal time was found to be a key explanatory variable for these findings.35 We expect that these issues would increase role conflict, our major predictor variable. These data suggest that repeating our study today could result in even lower levels of both marital and parental satisfaction, especially in women physicians, owing to the added time pressure and job stress of the current medical environment.
As in all studies that utilize questionnaires or interviews, our study is limited to data obtained from physicians’ self-reports. Self-report bias might have especially affected reports of perceptions (e.g., regarding marital and parental satisfaction, frustration with role conflict, and spouse support for career). We suspect that because of social response bias, physicians would have reported higher marital and parental satisfaction than might have actually been.
Our study is limited in its ability to make causal inferences by its cross-sectional design. We also did not evaluate the marital and parental satisfaction of physicians’ spouses, but prior research indicates that marital satisfaction scores of physicians and their spouses are highly correlated.9 Other limitations of this study center on issues of generalizability. Because the study group included only married physicians with children, results can only be applied to a similar group. Although this represents the majority of physicians (three quarters of male physicians and half of female physicians), our sample excludes single, widowed, and currently divorced parents because our independent and dependent variables examined three interrelated domains: career, marriage, and family. Our sample was from one county and was not selected randomly, although it was selected in an unbiased manner. The similarity in demographic characteristics between our sample and the AMA physician database shows that our data should be representative of married physicians with children who live outside Southern California (see Table 1). It is possible that selection bias was created by the nonresponders and nonreceivers. We believe the higher than expected nonreceiver rate was due to the inaccurate mailing addresses from our original list. The younger age of the nonreceiver group may indicate that many of these are young physicians in training who move frequently. The study area included a medical school and major university teaching hospital.
In conclusion, it appears that once the choice of medical specialty is made, minimizing the level of role conflict and having a supportive spouse are the two most important factors influencing marital and parental satisfaction among physicians. Among working physicians with children, role conflict appears to be the intervening factor between both marital and parental satisfaction and conditions at work and home, such as work hours, child care, and household responsibilities. Furthermore, working in a salaried medical position may benefit parental satisfaction. Our findings support the growing body of literature that family-friendly work options are good for working parents, their families, and their employers.
This work was supported in part by grants from the UCLA Academic Senate Committee on Research, UCLA Stein/Oppenheimer Endowment, and the Long Beach Chapter of the American Medical Women’s Association.