Abstract
BACKGROUND:
Physician satisfaction is linked to positive patient outcomes. Mothers form an increasing fraction of the obstetrics and gynecology (ob/gyn) workforce.
OBJECTIVE:
Define factors that affect physician satisfaction among ob/gyn physicians who are also mothers.
METHODS:
We constructed and validated a Redcap survey and invited members of online ob/gyn-mom groups to participate. Characteristics of participants’ professional and personal lives were evaluated for possible association with the satisfaction outcomes. Comparison testing was performed using Chi-squared test or Fisher’s exact test for categorical variables, Student’s t-test for parametric variables, and Wilcoxon Rank-Sum test for non-parametric variables.
RESULTS:
Responses were received from 232 participants. A majority reported being unsatisfied with their time to spend with children (66%), partner (70%), and on personal hobbies/activites (75%). Eighty-percent rate professional morale as very/somewhat positive. Women who rated their morale as very/somewhat positive worked fewer hours per week than women with neutral/negative responses (43.6 vs 49.7, p = 0.01). Women with positive morale were also less likely to work over 50 h/week (39.5% vs 56.8%, p = 0.04).
CONCLUSIONS:
Ob/gyn physician-mothers have high professional morale but are dissatisfied with time for extra-professional activities. Longer clinical hours correlate with dissatisfaction based on several measurements.
Introduction
Physician satisfaction has been linked to improved physician efficacy [1] quality of care, and patient satisfaction [2–4]. Dissatisfaction has been associated with increased physician turnover and early retirement [5] as well as decreased patient compliance [2]. Physician job satisfaction also influences the number and quality of future medical school applicants [6].
Female gender may be associated with lower physician job satisfaction [7, 8] but the association of satisfaction with age and gender has been understudied [9]. Popular press has discussed the pressures associated with parenthood (particularly motherhood) while working as a physician [10] but very few formal studies address job satisfaction among physician-parents [11, 12]. This issue becomes increasingly salient given the changes in the physician workforce over the last two decades. There is an increased number of women working as physicians, an increased emphasis on work-life balance, and more two-physician or two-working parent households. The combination of job strain and family circumstance has been associated with adverse patient health outcomes [13], further emphasizing the need for knowledge that may help improve physician-parental job satisfaction. Armed with such information, individuals may be better able to make career choices that will improve their quality of life.
The purpose of the current study is to investigate the experiences of ob/gyn physicians who are also mothers and to identify factors that correlate with satisfaction. We hypothesize that ob/gyn physician-mothers will report certain common experiences and that certain practice models (e.g. less than full time, group practice) will be favored as more conducive to work-life balance and job satisfaction.
Materials and methods
We constructed a REDCap survey and invited members of an online ob/gyn-mom group to participate. The 5-page, 40-question survey was developed by the authors and tested for validation on a small group of potential respondents to check for usability, comprehension, and technical functionality. Please note that since some of the questions addressed similar topics, and in the interest of brevity, we have chosen to omit some of the questions in the results section. Full survey answers available upon request. Links to the survey were posted in the Facebook groups “Ob/Gyn Moms” and “M in Maternal-Fetal Medicine”. These are private groups and membership is approved after confirmation of physician and maternal status. Informed consent was obtained prior to survey initiation and participants were free to skip questions. Answers were stored in a REDCap (Research Electronic Data Capture) online, secure database. The survey was not advertised outside of these groups and the survey was kept open for 2 weeks with periodic reposting and reminders on these pages. Respondents were able to leave questions blank if desired and were able to review and change answers if desired. Items were not randomized but adaptive questioning was used with conditional responses to other items to reduce the number of non-applicable questions viewed by each participant.
Participants were asked about demographics (age, gender, marital status), life-variables (spouse/partner occupation, child-care method, number of children), and job-variables (type of employment, specialty/ sub-specialty, hours/week, trainees, research). Participants were also asked questions about overall morale, level of satisfaction with time for their children, time for their spouse, and time for non-work activities. Characteristics of participants’ professional and personal lives were evaluated for possible association with the satisfaction outcomes. Answer options were multiple choice and provided a non-response option. Questions were structure in the following format: How would you rate your own professional morale? (very positive, somewhat positive, neutral, somewhat negative, very negative) I am satisfied with the amount of time I have available to spend with my children (strongly agree, somewhat agree, neutral, somewhat disagree, strongly disagree)
REDCap allows investigators to check the completeness of surveys post-hoc. Although this survey was “open” in that a password was not required, each respondent provided an email address, which was used to check for duplicate surveys. Survey instructions also told participants to fill out only one survey per person.
Results are reported as frequency (%) or mean (standard error). Comparison testing was performed using Chi-squared test or Fisher’s exact test for categorical variables, Student’s t-test for parametric variables, and Wilcoxon Rank-Sum test for non-parametric variables. A significance value of 0.05 was used. The Institutional Review Board reviewed our study and considered it exempt. This paper is written in accordance with the CHERRIES guidelines for reporting of Internet surveys [14].
Results
We received 224 unique visitors of which 221 proceeded with the survey and 193 completed all the questions. The groups in which the survey was posted have a total of about 2200 members. This gives a view rate of 10%. There were three visitors to the survey page who elected not to complete the survey, for a participation rate of 98.6%. There were no other surveys that were terminated early. Overall demographic characteristics are shown in Table 1.
Demographic characteristics of study responders
Demographic characteristics of study responders
In addition to the demographic and employment variables described above, the median age of responders was 38 (SD 5.8) with a range of 29 to 62. Median number of children was 2 (SD 0.9) with a range of 1 to 6. One-hundred-fifty (64.6%) respondents reported working with trainees and 98 (42.2%) were involved in research. Median clinical hours per week was 45 (SD 14) and range was 6 to 100. Median non-clinical hours per week was 5 (SD 10) and range was 0 to 50. One-hundred seventy-five reported doing deliveries either for their group or their group and other groups and 125 report at least some in-house call.
Respondents reported the following factors (in decreasing order of frequency) as contributing most to overall satisfaction: patient relationships, intellectual stimulation, financial benefits, interactions with other medical professionals, education of trainees, prestige of medicine, and research opportunities. The following factors were rated (in decreasing order of frequency) as contributing most to dissatisfaction: long hours/lack of family and personal times, liability/defensive medicine, non-clinical paperwork, electronic medical record implementation, health care reform uncertainty, government regulations, reimbursement issues, lack of clinical autonomy, managed care, and pressures of running a practice. Table 2 details overall responses to other questions about work and life satisfaction.
Overall responses to work/life satisfaction questions
Women who rated their morale as very or somewhat positive worked fewer hours per week than women with neutral or negative responses (43.6 vs. 49.7, p = 0.01). Women with positive morale were also less likely to work over 50 h/week (39.5% vs. 56.8%, p = 0.04). Figure 1 illustrates the overall satisfaction based on number of hours worked.
Professional Morale in Relation to Work Hours.
Women who said they would chose medicine again as a career were more likely to have a physician-spouse (29.8% vs 7.7%, p = 0.01) and were more likely to use a nanny/au-pair for childcare (45.1% vs 13.9%, p = 0.01). Longer clinical hours, especially over 50 hours/week, correlated with dissatisfaction regarding time spent with children, spouse, and on personal activities. Women with older children were more likely to be satisfied with the amount of time available to spend on personal activities and interests. No other variables correlated with any satisfaction outcome. Table 3 shows variables between respondents who were very/somewhat satisfied and other responses.
Predictive characteristics of job satisfaction
*very or somewhat satisfied with overall career/job.
The results confirmed our hypothesis that ob/gyn physician-mothers will report certain common experiences. For example, Ob/gyn physician-mothers have high professional morale but are dissatisfied with time for extra-professional activities, including time with family. It is interesting to note that more women are more satisfied with time available to spend with children, less satisfied with time to spend with spouse, and least satisfied with personal time. This finding suggests that children and spouse are prioritized over self-care, which most working mothers know to be true. A recent study among ophthalmologists also found that although women in this field were satisfied, they wished they had more time for family [15].
Regarding the second part of our hypothesis, that certain practice patterns would be more conducive to satisfaction, women who reported being satisfied overall and on specific measures worked fewer hours and were less likely to work over 50 hours per week than less satisfied respondents. It should be noted that even among the respondents who were overall satisfied, the average hours per week still exceed 40. Christopher et al. found a similar trend in a study of physicians across all specialties [16]. Their study did not substratify by gender or parenthood status. The overall mean for all respondents in their study still exceeded 50 hours/week, but physicians working fewer hours were found to spend a greater proportion of that time in direct patient care. The authors speculated that less time spent in non-clinical administrative duties might partially explain the increase in satisfaction among these physicians. This theory would support our finding that “non-clinical paperwork” was rated highly as a factor contributing to dissatisfaction.
Other investigators have used different techniques to assess the association of work-hours and satisfaction, and results have generally been consistent. A recent study in this journal found similar results among working parents in non-medical fields, noting that a good balance between work and leisure were of importance of working parents’ perceived health a stress [17]. Other studies have found that part-time physicians report less burnout and higher satisfaction than full-time physicians [18] and that women physicians have 1.6 times the odds of reporting burnout compared with males [19]. The odds of burnout in women increased by 12–15% for each additional 5 hours worked per week over 40 hours per week [19]. In these studies, burnout was characterized by emotional exhaustion, a sense of depersonalization, feelings of diminished personal accomplishment, and a negative sense of self-value and ability.
Occupation of spouse and childcare method may also contribute to satisfaction among ob/gyn mother-physicians. Having a physician spouse and using a nanny/au-pair for childcare increased the likelihood of choosing medicine as a career again. This finding was unexpected. It is obvious that when in a long-term relationship, having a partner who is supportive of one’s career goals would improve satisfaction, and we speculate that physician-spouses may be better able to understand and support the rigors of a medical career. Although we were not able to find studies specifically addressing the impact of partner/spouse occupation on physician satisfaction, several investigators have reported on the “medical marriage” and their findings lend support to our speculation [20, 21]. We further suspect that the nanny/au-pair method of childcare may correlate with being in a “medical marriage”. These findings are somewhat consistent with previous studies. Warde et al. found that major factors associated with parental satisfaction were a supportive spouse, salaried practice setting, and marriage to a spouse working in a profession. They also found that marriage to a spouse working as a homemaker correlated with satisfaction. In that study, number of hours was not directly correlated to satisfaction score but did relate to intervening variables, such as marital conflict [12]. Another study demonstrated that women reported feelings of guilt about their performance as mothers and doctors [11]. It should be noted that findings in these types of studies are likely to change over time as social and medical attitudes and practice patterns evolve.
Almost all respondents felt that motherhood had made them better physicians but only half felt that being a physician has made them better mothers. Furthermore, many respondents answered that unequal treatment of women-physicians and parent-physicians is a problem in medicine. Unequal treatment of women in medicine is not a new concept and we will not further address it here, except to note that almost half of current medical students are women; thus, more work is needed on solutions to this problem. We were unable to find other work on the challenges faced specifically by physician-parents, but this area may be fruitful for further research.
Our study has some obvious flaws inherent to the study design including those inherent to all Internet surveys, mainly selection bias, which has been extensively described elsewhere [22]. We must also acknowledge a possible effect of the cognitive-dissonance free-choice paradigm, which states that humans tend to change their attitudes to align with previous choices. To state this principle another way, people tend to view their chosen choices more favorably [23]. Applying this theory to our study, respondents would be more likely to rate their career choices favorably. Furthermore, we chose to use a quantative approach to this study, which limits nuanced interpretation of individual results. We chose this approach over a qualitative questionnaire secondary as the results are often more conducive to interpretation. The study is also limited by our query of only mothers in one field of medicine. It would be interesting to study the experience of fathers in this field but there are no online groups that would facilitate query of this population. Another limitation is that the subspecialty of maternal-fetal medicine is over-represented as compared to other subspecialities.
Nevertheless, we had 200 ob/gyn physician-mothers who were willing to provide important information about satisfaction and lifestyle variables. Both our view rate and participation rate are excellent for this type of survey [22]. We feel that the subject matter was quite conducive to this approach and we were able to show important differences in particular variables that correlate with increased satisfaction.
Conclusion
Physician satisfaction is important for patient care and patient satisfaction. We hope that the results of this study will provide physician employers and individual physicians with useful knowledge when designing practices and making career decisions. Examples of these choices might include (but are not limited to): carefully assessing expected hours prior to choosing a position or focusing on ways to contain hours-worked within the optimal range.
Conflict of interest
The authors report no conflicts of interest.