Worklife and Satisfaction of General Internists

Background  Prior studies have reported relatively low job satisfaction for general internists. We used data from a large US physician survey to assess correlates of satisfaction of general internists.

Methods  The Physician Worklife Survey was mailed to a national random stratified sample of 5704 US physicians. General internists were assessed for their satisfaction, training, patient mix, work hours, the likelihood of recommending their specialty to medical students, and job stability. We then compared them with a specialist sample (internal medicine subspecialists [IMSSs]) and a primary care sample (family physicians [FPs]). Logistic regression was used to model predictors of satisfaction, stress, and medical student recruitment.

Results  There were 2326 respondents (adjusted response rate, 52%): 450 (19%) were general internists; 502 (22%), FPs; and 438 (19%), IMSSs. General internists were less satisfied than were IMSSs with their relationships with colleagues and with patient care issues (P<.01 for both) and less satisfied than were FPs with community ties (P = .001). Global job, career, and specialty satisfaction were significantly lower for general internists vs FPs and IMSSs (P<.05). General internists spent proportionately more of their work week in the hospital than did FPs (20% vs 13%; P<.001) and more time providing outpatient care than did IMSSs (56% vs 42%; P<.001). General internists had more patients with complex medical and psychosocial problems than did FPs (P<.01) but fewer patients with complex medical problems than did IMSSs (P<.001). Higher satisfaction for general internists was associated with older physician age, less time pressure during office visits, fewer work hours, and fewer patients with complex psychosocial problems (P<.05 for all). General internists were less likely than were FPs to recommend their specialty to medical students (P<.001). Specialty satisfaction, female gender, and control of hassles predicted medical student recruitment by general internists.

Conclusions  General internists' role of caring for patients with complex problems is associated with lower levels of satisfaction than for IMSSs and FPs. Adjusting caseload for patient complexity, expanding time for office visits, and additional training in the care of patients with psychosocially complex problems may improve the job satisfaction of general internists and medical student recruitment into the specialty.

GENERAL INTERNISTS provide a large proportion of adult primary care in the United States; therefore, the future of this specialty is vital to the US health care system. Yet, job dissatisfaction among general internists is prevalent.^1-3 Prior studies^2,4-7 have revealed that general internists remain satisfied with their patient and colleague relationships and the intellectual challenge of the profession. However, they are dissatisfied with their income and the lack of control over their practices, with more time spent on administrative and business aspects and less time available to see patients.^2,4-7 A lack of general internist enthusiasm for their specialty and primary care has also been documented.^2,3,8 A national study² of generalist and subspecialist internists revealed that 40% would, if given the choice, select a career other than internal medicine and 40% discouraged medical students from entering their specialty. Zinn and colleagues⁸ found similar career satisfaction between academic general internists and family physicians (FPs). However, general internists were much less likely to encourage medical students to enter primary care than were FPs (36% vs 86%), an action known to affect student career choices.⁹

The advent of managed care brought irrevocable changes to the practice of medicine, with a subsequent focus on physician satisfaction. Physician satisfaction is an important determinant of patient adherence,¹⁰ patient satisfaction,¹¹ and physician turnover.¹² Meanwhile, the number of students choosing careers in internal medicine has declined, with a corresponding increase in numbers choosing internal medicine subspecialization and family practice. Many researchers and the American College of Physicians–American Society of Internal Medicine have called for a reaffirmation¹³ and a redefining of general internal medicine (GIM).^14,15

Previous data published from the Physician Worklife Study (PWS) have identified GIM as a specialty at risk. General internists had lower levels of satisfaction and higher numbers of patients with complex psychosocial and medical problems than did FPs and less time allotted for new patient and follow-up appointments than did internal medicine subspecialists (IMSSs).¹⁶ To our knowledge, there are no other large, recent, national studies assessing the satisfaction levels of general internists and worklife characteristics associated with satisfaction. We, thus, performed this detailed analysis of responses from the PWS to assess the satisfaction of general internists and to compare their satisfaction with that of IMSSs and FPs. We also sought to identify potentially modifiable factors negatively affecting the satisfaction and work life of general internists.

The development of the survey instrument, satisfaction facets, and global measures have been described elsewhere.^17-20 The Physician Worklife Survey is an 8-page 150-item mail survey that measures physician practice characteristics and aspects and outcomes of job satisfaction. The conceptual development of the satisfaction facets and global measures included using items from existing satisfaction measurement instruments (mainly the work of Stamps and Cruz²¹ and Lichtenstein²²), 6 physician focus groups, and a content analysis of physician open-ended data from a survey of large group practices.²³ An item pool was constructed and subjected to expert panel review for refinement. This measure was tested on a pilot group of 2000 physicians obtained from the American Medical Association's Masterfile (n = 835; adjusted response rate, 55%). A 36-item 10-facet satisfaction measure resulted after factor and reliability analysis. Reliabilities of the 10 facets of satisfaction ranged from 0.65 to 0.77, and the global satisfaction measure reliabilities ranged from 0.84 to 0.88. The satisfaction and global measures are composed of 2 to 5 individual items each and are measured on 5-point Likert scales. A modified global specialty satisfaction variable composed of 2 of the 3 original items was used in a regression analysis of factors predicting recommendation of specialty for general internists and had a reliability of 0.72. The facets include satisfaction with autonomy; relationships with patients, relationships with colleagues, relationships with staff; patient care issues; personal time; community; income; administration; and resources. Global satisfaction measures evaluated job, career, and specialty satisfaction. The 10 facets of satisfaction accounted for 58% of the variance seen in global job satisfaction. A complete listing of the individual items and associated facets is included in a prior publication from the PWS.¹⁶

The American Medical Association's Masterfile was then used to select a national sample of physicians in primary care (family medicine, GIM, and pediatrics) and subspecialties of internal medicine and pediatrics (n = 5704) after stratification by race, ethnicity, practice specialty, and 2 levels of regional participation in managed care (high and low). Four survey mailings were sent in 1996 to 1997 with cover letters from investigators and local and specialty medicine societies. Two hundred nonresponders were telephoned, revealing an 18% inaccurate address rate. Sampling weights were constructed from this information using the Lessler and Kalsbeek²⁴ technique. Nonresponse bias was assessed, looking for trends between survey instrument variables and return time of the survey, calculating Spearman rank correlation coefficients. Few meaningful associations were found (only 4 of 140 variables with r>0.1).

All data analyses were weighted to adjust for response rates and sampling probabilities using a software package (STATA, version 5.0; Stata Corp, College Station, Tex). This adjusts for sampling design by using Taylor series linearization. In all analyses involving the 3 specialties in which pairwise differences in means or proportions were considered, P values are given after adjusting for multiple comparisons. This was done by finding the distribution of the maximum absolute test statistic for the 3 pairwise comparisons under the null hypothesis of no differences and comparing the t statistics for the individual hypothesis tests with the percentiles of this distribution.

Mean differences between specialties were examined for the 10 facets of satisfaction, 3 areas of global satisfaction, and intermediate outcomes of job stability, stress, burnout, and specialty recommendation to medical students. Burnout was assessed with a single-item measure of the degree of burnout. Job stability was assessed by asking intention to leave the job or practice of medicine. For general internists, satisfaction and recommendation analyses were also performed comparing gender and age groups (<45 years and ≥45 years). Mean differences between specialties were evaluated for clinical training, patient case mix, work hours, and percentage of time spent in different activities. Time pressure was a variable representing respondents' needing more time in an office visit than allotted. Worklife control was measured with 13 single-item questions asking about control over multiple work-related issues. Factor analysis determined 3 work control factors: control of workplace, medical decisions, and hassles. Health maintenance organization physicians were defined as those practicing in staff- or group-model health maintenance organizations with greater than 50% capitation or managed care patients. Full-time physicians were defined as those practicing more than 40 hours per week. Logistic regression analyses were used to model predictors of general internist satisfaction for the facets and global scales, controlling for age, gender, health maintenance organization practice setting, percentage of patients with complex psychosocial problems, income, work hours, and the time pressure factor. Logistic regression was also used to model predictors of general internists' desire to recommend their specialty to medical students, controlling for global specialty satisfaction, age, gender, and variables specifying control of medical decision making, hassles, and workplace control.²⁵

After accounting for inaccurate addresses, ineligible physicians, and returned envelopes, the adjusted response rate was 52% (n = 2326). Adjusted response rates by specialty were 45% for general internists, 52% for FPs, and 48% for IMSSs (pediatricians had substantially higher return rates). Of the 2326 respondents, 19% were general internists; 22%, FPs; and 19%, IMSSs (Table 1). Respondents were mostly men, white, and married; their average age was 47 years. There were few differences in demographics or practice settings (Table 1).

For the 10 satisfaction facets, general internists were most satisfied with relationships with patients and least satisfied with income, personal time, and administrative duties (Table 2). When compared with other specialties, general internists were less satisfied than were IMSSs in their relationships with colleagues and in patient care issues (P<.01 for both) and less satisfied than were FPs with community ties and income (P = .001 and P<.05, respectively). For global satisfaction measurements, general internists were the least satisfied with their job, career, and specialty (Table 2). Most of these differences in the global satisfaction measures (0.2-0.5 on a 5-point Likert scale) were believed to be clinically and statistically meaningful. There were no gender differences in global satisfaction for general internists. However, analysis by gender and age showed that older male general internists were more satisfied than younger male general internists with their jobs and career (Table 3). Interestingly, the opposite was seen for women: younger female general internists were more satisfied with their jobs than older female general internists.

General internists reported significantly less training in primary care, psychosocial care, and practice management than did FPs (P<.001 for all) (Table 4). Exposure to utilization review and managed care training was minimal for all 3 specialties. Remarkably, general internists reported significantly (P = .001) less training in psychosocial care than did IMSSs. An analysis of case mix by specialty (Table 5) showed that general internists reported that 46% of their patients were elderly persons, nearly double that reported by FPs. General internists also reported significantly more patients with medically and psychosocially complex problems than FPs. For general internists working full time, just more than half of their time (56%) was spent in outpatient practice, 20% in hospital care of patients, and 25% on patient and other work-related activities (percentages do not total 100 because of rounding) (Table 5). General internists spent proportionately more time in the hospital than did FPs and more time in the office than did IMSSs (P<.001 for both).

After controlling for age, gender, practice setting, work hours, income, time pressure, and percentage of patients with complex psychosocial problems, older general internists were more likely to be satisfied than were younger general internists in 7 of 10 facets of satisfaction and 2 global satisfaction scales (Table 6). Women had higher job and career satisfaction compared with men. More weekly work hours were associated with decreased satisfaction in multiple areas, including autonomy, relationships with patients, patient care issues, personal time, and administrative duties. Increased time pressure during office visits was associated with decreased satisfaction in many areas, including autonomy, patient care, and global job satisfaction. More weekly work hours were also associated with increased stress levels among general internists. Interestingly, stress also increased for general internists who had more patients with psychosocially complex problems. A separate regression analysis found no relation between the 5 training areas and the facets of satisfaction or global satisfaction (data not shown).

Despite the lower satisfaction among general internists, there were no differences between the 3 specialties in stress, burnout, or intent to leave the job (data not shown). However, general internists were substantially less likely to recommend their specialty as a career to students compared with FPs (3.26 vs 3.85 on a 5-point Likert scale; P<.001). Women general internists were more likely to recommend their specialty than were men (Table 3), with younger women internists the most likely to recommend the specialty. Regression analysis of factors predicting recommendation of specialty for general internists revealed that global specialty satisfaction (β = .84, P<.001), female gender (β = .26, P=.01), and control of hassles (β = .13, P=.04) were significant predictors and explained 70% of the variance.

In this sample of 1390 physicians from the PWS, we have shown that general internists are less satisfied with their jobs, careers, and specialty than are FPs and IMSSs. General internists are also less satisfied with their relationships with colleagues and with patient care issues than are IMSSs and less satisfied with community ties than are FPs. We found that general internists have less training in primary care, psychosocial care, and managed care and that they have more patients with psychosocially complex problems (which were also a source of increasing job stress) and more patients with medically complex problems than FPs. More than half of a general internist's 59-hour work week is spent in outpatient care, while 20% is spent in the hospital. General internists' satisfaction was associated with older age, female gender, fewer weekly work hours, and less time pressure during office visits. While general internists were substantially less likely than FPs to recommend their specialty to medical students, young women in GIM displayed higher enthusiasm for recommending the specialty than other general internists.

Our results show that general internists' satisfaction continues to be of concern. Our findings of specific areas of satisfaction and dissatisfaction are consistent with and amplify those of prior studies.^2,6 Lower satisfaction with relationships with colleagues and patient care issues than IMSSs is a new finding and may reflect dissatisfaction with "second-class citizenship" compared with procedurally oriented colleagues and the increasing challenges of the gatekeeper role found by Wahls et al.⁶ Factors addressed in the patient care issue questions included the gatekeeper role (which may increase adversarial relationships with physicians) and increasing time pressure (which may overwhelm primary care physicians). In this regard, FPs had comparable low levels of satisfaction with patient care issues. General internists in our study were somewhat less satisfied with income than were FPs. Because income has been shown to be important to medical students who do not choose GIM but less so for those who enter,^9,26 fair compensation across the specialties may increase general internist satisfaction and attractiveness.

The Robert Wood Johnson Foundation's Survey of Young Physicians in 1991²⁷ found low levels of training in the business aspects of medicine. However, those trained in family medicine reported better preparedness over internal medicine–trained physicians in preventive care, psychosocial care, and practice management. Merkel et al²⁸ surveyed internal medicine residency programs in 1985; less than half had mandatory training in psychosocial care. Our findings confirm low levels of training among the specialties in the business aspects of medicine and show that general internists have less training in primary care and psychosocial care than do FPs. Training did not relate to satisfaction, although this lack of association may be due to the consistently low degree of preparedness in these areas.

Case mix has become an area of increasing interest, especially considering the relationship between time allotted for appointments and physician job satisfaction.¹⁶ Our study reinforces the findings of Kravitz et al,²⁹ with general internists reporting twice as many elderly patients and patients with complex medical problems than FPs. In addition, we found that general internists have more patients with complex psychosocial problems than FPs. Zinn et al⁸ found that academic general internists experienced less satisfaction from taking care of patients with psychosocially complex problems than did FPs. Our study showed low levels of general internist training in psychosocial aspects of care and general internists had an association between patients with complex psychosocial problems and decreased resource satisfaction and higher stress. This suggests an important but remediable deficiency in internal medicine residency program training.

More work hours per week for general internists were associated with less satisfaction with autonomy, patient care, administration, and personal time and with higher stress levels. Full-time general internists in our study worked 59-hour work weeks and spent proportionately more hours per week in the hospital than did FPs, attesting to the role of general internists in dealing with patients with complex medical problems. This difference becomes important given the evolution of the hospitalist movement in internal medicine.³⁰ The full impact of this new field on the practice of GIM is not yet known.³¹ Designated hospitalists incorporated into a group practice's call schedule may better define inpatient and outpatient roles, decrease work hours, and increase personal time and satisfaction.

For general internists, time pressure during office visits was related to lower satisfaction with autonomy and patient care and to lower global job satisfaction. A previous study¹⁶ from the PWS showed that IMSSs reported having more time for new consultations and physical examinations (51 minutes) than did general internists (39 minutes) and FPs (34 minutes). Discrepancies in visit length between general internists and FPs may be explained by differences in case mix, because general internists have more elderly patients and patients with complex problems. However, when compared with IMSSs, generalists typically need to provide routine preventive care and screening and coordinate all of a patient's medical issues in 39 minutes, while IMSSs often focus on problems within one organ system and are allotted 51 minutes. Research has shown that time pressure may also have an adverse effect on patients because the physician has less time to address psychosocial concerns³² and alcohol use,³³ and this may lead to inappropriate³⁴ or increased³⁵ medication prescribing. Generalist satisfaction may increase in systems that take into account case mix and perceived time pressure when creating physician schedules and panel size.

The future of GIM relies on medical students choosing this specialty as a career. It is, therefore, disturbing that general internists reported lower rates of encouraging students to select their specialty than did FPs, confirming and expanding on an earlier study⁸ of academic general internists and FPs. We did find that young women general internists were the most enthusiastic about recommending GIM as a career and that satisfaction with internal medicine and female gender predicted general internists' recommendation of their specialty. Our findings will hopefully trigger additional research to detect why general internists are dissatisfied and not recruiting medical students.

The higher satisfaction levels of older general internists in our study echoes previous research. Haas et al,¹ in a study of academically affiliated general internists in Boston, found that younger physicians (those aged <50 years) had lower overall satisfaction than did older physicians. This may represent selection bias because unhappy internists may have left the specialty or practice of medicine, leaving mainly satisfied physicians. We found a potential gender difference in older general internist satisfaction, however, with older women having lower global job satisfaction than younger women. Our overall finding, then, of older general internist satisfaction compared with young physicians may be more reflective of the male preponderance in medicine. Interestingly, female gender predicted higher job and career satisfaction, findings not seen in a study in 1990.¹ The gender and age satisfaction differences, along with young women internists' enthusiasm about student recruitment, may impact favorably on the future of GIM because the percentage of women in this specialty is expected to reach 32% by 2010.³⁶ Further research should be directed at identifying factors important in young physician work life and satisfaction—especially addressing gender differences, personal time, community ties, patient care issues, resources, and relationships with patients and colleagues.

This study has several limitations. First, the response rate of 52% was suboptimal but typical of physician survey responses of 54%³⁷; fortunately, our assessment of late or nonresponse bias suggests that little such bias existed. Second, the self-report nature of the survey could lead to inaccuracy. Third, some physicians claimed to be practicing another specialty other than the specialty in which they were sampled from the American Medical Association's Masterfile. These discrepancies may result from physician preference for categorizing their limited practice³⁸ or physician switching of their practice from GIM to a medicine subspecialty or vice versa.³⁹ Last, none of the factors we explored predicted the relatively low specialty satisfaction among general internists.

Our study clarifies the role of general internists as providers of primary care to adults with complex medical and psychosocial problems in inpatient and outpatient settings. However, general internists' satisfaction with their job and specialty is moderate and lower than that of their 2 closest related specialties. Also, many general internists are not recommending their specialty to medical students with the vigor of physicians in other specialties. Fortunately, many factors that predict general internist dissatisfaction are modifiable. Attention to such factors as training in primary care and psychosocial care, long work hours, case mix adjustment (including psychosocial patient complexity) when devising panel sizes, and monitoring and minimizing time pressure during office visits may serve to increase the satisfaction of general internists and reaffirm the future of this primary care specialty. Higher levels of satisfaction among young women general internists and their willingness to recruit medical students may also strengthen the future of GIM.

Accepted for publication July 30, 2001.

Dr Douglas is now with the Department of Statistics, University of Illinois at Champaign-Urbana.

This study was supported by grant 27069 from the Robert Wood Johnson Foundation, Princeton, NJ.

This study was presented in part at the Midwest regional meeting of the Society of General Internal Medicine, Chicago, Ill, September 11, 1999; and the annual meeting of the Society of General Internal Medicine, San Diego, Calif, May 5, 2001.

Society of General Internal Medicine Career Satisfaction Study Group

University of Wisconsin, Madison: Mark Linzer, MD; Julia E. McMurray, MD; Jeffrey Douglas, PhD; John Frey, MD; William E. Sheckler, MD. University of Alabama, Tuscaloosa: Eric S. Williams, PhD. University of North Carolina at Chapel Hill: Donald E. Pathman, MD, MPH; Thomas R. Konrad, PhD. University of Alabama, Birmingham: Kathleen Nelson, MD. University of Washington, Seattle: Richard Shugerman, MD. New York University, New York: Mark D. Schwartz, MD. Brigham and Women's Hospital and Harvard Medical School, Boston, Mass: JudyAnn Bigby, MD. Oregon Health Sciences University, Portland: Martha S. Gerrity, MD. Society of General Internal Medicne, Washington, DC: David Karlson, PhD; Elnora Rhodes.

Corresponding author and reprints: Tosha B. Wetterneck, MD, Department of Medicine, University of Wisconsin Hospital and Clinics, 600 W Highland Ave, J5/214 Clinical Science Center, MC 2454, Madison, WI 53792-2454 (e-mail: tbw@medicine.wisc.edu).