Satisfaction with work and quality of life among British Columbia’s physicians: A review of the literature

Introduction

The retention of primary care physicians in rural areas of Canada remains one of the most persistent problems confronting Canadian health care.[1-3] British Columbia is no exception, with many rural areas remaining under-doctored.[4] Typical retention rates for British Columbia’s rural communities of fewer than 7000 people are between 60% and 80%.[5] If a community has 10 physicians, this means each year two to four of these physicians will leave that community. Typical retention rates for communities with 7000 or more people are higher, between 85% to 90%. A recent decrease in absolute numbers of rural physicians in British Columbia’s rural communities is a worrisome finding, as it implies physician gain (recruitment) is not keeping up with physician loss.

The finding of high rural physician turnover rates[5] and the recent escalation of job actions including mass resignations of hospital privileges[6,7] would lead one to predict British Columbia’s rural physicians have lower job satisfaction compared to physicians working in urban areas of the province.[8] This prediction is based on studies that have shown that physician dissatisfaction with clinical workload is associated with increased likelihood of a physician leaving his or her practice[9-11] and increased likelihood of becoming involved in job action.[12,13]

This paper reviews the literature on British Columbia physician satisfaction in an attempt to find out whether rural primary care physicians are less satisfied with work and life compared to their urban counterparts.[14-19] Trends in rural physician job satisfaction are examined, as well as reasons for physician job dissatisfaction. 
Theoretically, if one understands the reasons for low job and low life satisfaction, one can institute policies that will lead to improved job satisfaction, improved physician retention, more equitable distribution of physician services, and ultimately, better care for the people living in under-serviced rural Canada.[20-23]

 Overview of BC satisfaction studies

Over the past 10 years or so, there have been four surveys in British Columbia looking at various issues of rural physician professional and personal satisfaction.[14-19]

In November 1989, Kazanjian and colleagues sent questionnaires to 702 physicians living in rural communities (population =10 000) in British Columbia, and 668 physicians living in urban communities (population >10 000). Both family physicians and specialists were surveyed (Table 1, Table 2 ).[14,15]

Van der Weyde and colleagues sent questionnaires to 60 primary care physicians living in isolated rural communities and 60 primary care physicians working in urban communities in January 1997.[16] The rural isolated group of physicians resided in communities considered to be isolated by the Medical Services Commission (and hence qualified for the Northern Isolation Allowance).[24] A small proportion (<10%) would have provided some specialty services.[14] The urban group of physicians resided in Vancouver, Victoria, their associated suburbs, and any other British Columbia area with a population exceeding 35 000, except for Prince George, which was considered separately.

In July 1998, the British Columbia Medical Association sent questionnaires to 380 physicians (87% general practitioners; 13% specialists) working, living, and taking call in rural isolated communities and to 388 physicians (87% general practitioners; 13% specialists) working in urban areas. Urban areas were considered to be the major provincial referral centres, including Greater Vancouver, the Capital Health Region, as well as Courtenay/Comox, Cranbrook, Kamloops, Kelowna, Nanaimo, Penticton, Prince George, Trail, and Vernon.[17]

Thommasen and colleagues sent questionnaires to 198 primary care physicians living in isolated rural communities in October 1998. Urban physicians were not surveyed. In addition to most of the same survey questions included in the van der Weyde survey, these physicians were also asked to complete questionnaires pertaining to depression and burnout.[18,19]

In all surveys, physicians were asked to rate their satisfaction with issues pertaining to quality of life, overall job satisfaction, work-week activities, and work-week job satisfaction. Quality-of-life issues include things such as satisfaction with health, housing, finances, relationships, recreation, and self-esteem. Likert scales were used to grade level of satisfaction in all studies. In two of these surveys a five-point Likert scale, ranging from very unsatisfied (score of 1) to very satisfied (score of 5) was used to grade the level of satisfaction. In the other two studies a seven-point Likert scale, ranging from very dissatisfied (score of 1) to very satisfied (score of 7) was used to grade the level of satisfaction. To account for the differences, scores reported in these studies are expressed as a percent of the maximal Likert Scale. For example, a Likert score of 3.5/5 = 70%, whereas a score of 3.5/7 = 50%. The test-retest reliability and validity of using Likert scales to measure job satisfaction has been demonstrated in previous studies.[25-27]

For the most part, the surveys are quite comparable. There are, however, differences in methodologies and in questions asked. The rural physician population in the November 1989 survey came from communities with fewer than 10 000 people and includes physicians working in both rural isolated communities and rural non-isolated communities. The rural physician population in the latter three surveys came from rural isolated communities, some of which have a population greater than 10 000 people (e.g., Creston, Dawson Creek) and so would have been excluded from the November 1989 survey. According to our calculations, 244 NIA physicians live in communities of fewer than 10 000 people, which means at least half (244 of 490) of the physicians surveyed in the November 1989 survey were probably physicians from rural isolated communities.[6,17,28]

Some questions were surprisingly similar across all surveys. For example, with the satisfaction with on-call time question, in the November 1989 survey physicians were asked to indicate their level of satisfaction with “length of working hours (on call).” In the July 1998 survey physicians were asked to indicate their satisfaction with “on-call schedule.” In the January 1997 and October 1998 surveys physicians were asked to rate their satisfaction with “number of on-call shifts/month.”

Another example is specialist availability. In the November 1989 survey physicians were asked to indicate their level of satisfaction with “access to specialist expertise.” In the July 1998 survey physicians were asked to indicate their satisfaction with “access to specialists.” In the January 1997 and October 1998 surveys physicians were asked to rate their satisfaction with “ease of obtaining specialist backup.” These questions were judged to be sufficiently similar that the answers could be categorized in our table under the heading “job satisfaction with specialist availability.”

There were sometimes subtle differences in how questions were asked in different surveys. Take global job/career satisfaction, for example. In the November 1989 survey physicians were asked to rate the statement, “When I think of my professional career I am quite satisfied with it and there is very little I would like to change.” In the July 1998 survey physicians were asked to rate the statement, “How satisfied are you with your practice.” In the January 1997 and October 1998 surveys, physicians were asked to rate their satisfaction with “Job.” These questions were judged to be sufficiently similar to be classified under global job satisfaction and the answers added to the summary tables.

Occasionally, the questions asked on a similar topic were so different that answers were not comparable. Take relocation plans, for example. All surveys had at least one question on relocation plans. The January 1997 survey and the October 1998 survey asked, “Do you currently wish to relocate?” In the 1989 survey, physicians were asked “How long do you plan to continue practising in this area?,” and in the July 1998 survey physicians were asked “Within the next year, do you intend to move your practice?” The questions are sufficiently different that one cannot compare results among the four studies, though we did think it was fair to compare the answers of the November 1989 survey and the July 1998 survey to one another, and to compare the answers of the January 1997 and October 1998 surveys.

 Rural physician satisfaction

Table 1 summarizes the British Columbia rural physician satisfaction survey data. In general, rural physicians appear satisfied with their overall quality of life. All aspects of quality of life reported scores of 64% or greater (Table 1), and except for education, rural physicians are not less satisfied than their urban counterparts (Table 2).

According to the October 1998 survey, overall quality of life correlates with days of vacation per year, age, and years in practice. In their study of clinical and academic specialists, Linn and colleagues reported that the more satisfied physicians in their study were also more likely to be older.[20]

According to the October 1998 survey, overall job satisfaction does not correlate with age, number of on-call shifts, days off per month, or days of vacation per year. This is consistent with other studies that show that it is perceived workload that seems to be important, not actual number of hours per week or days per month worked.[9,15,20] A logical policy that comes out of this finding would be to simply allow rural physicians to decide for themselves how much work he or she is willing to provide for a community. Issues such as call and continuous 24-hour patient coverage would then be concerns of the regional health boards, not individual physicians.[29]

Table 1 and Table 2 show that rural physicians are least satisfied (mean score <50%) with time for reading, number of on-call shifts per month, ease of obtaining vacation (locum) relief, clinic workload, work-related interruptions in sleep, continuing medical education (CME) availability, and days of work per month. Rural physicians are most satisfied with level of training, support of colleagues, and challenge of practice. Trends over time suggest that British Columbia’s rural physicians are becomingly less satisfied with many of these job-related factors. Other surveys have also identified relatively long work hours, lack of time for self, family, and CME, professional isolation, and income as important issues for North American rural physicians.[9,10,30-33]

Dissatisfaction with CME probably reflects the fact that rural physicians must travel greater distances, spend more money, and make a greater time commitment than their urban colleagues whenever they attend an educational meeting in a larger centre. Paradoxically, rural physicians often need more medical education training than their urban colleagues because rural communities typically have less specialist backup—so rural physicians are often called upon to deal with more complex medical situations by themselves.[34-36]

Three of the four British Columbia rural physician surveys had an urban control group. Overall job satisfaction does not differ between rural and urban physicians, but there are significant differences when comparing satisfaction with various aspects of their work (Table 2). For most of the job-satisfaction issues, urban physicians are significantly more satisfied. Interestingly, this table also suggests that urban physician satisfaction with vacation time, locum relief, income, and support of colleagues has declined over the past decade.

 Rural physician relocation plans

In the January 1997 survey, 15% of physicians in the rural isolated group said “yes” when asked “Do you currently wish to relocate?” and a further 28% answered “maybe.” In contrast, 2.5% of urban physicians answered “yes” and a further 12.5% said “maybe” respectively. Two years later rural physicians were asked the same question. Compared to January 1997, almost twice as many rural physicians (33%) said they wished to relocate, fewer said “maybe” (18%), and fewer had no plans to relocate.

The other two surveys also suggest that the number of rural physicians wishing to relocate has increased over time. In the November 1989 survey, 10% of rural physicians and 4% of urban physicians indicated that they would no longer be practising in 1 year. In the July 1998 survey, 24.5% of rural physicians and only 4% of urban physicians were planning to move their practices within the next year. The vast majority of rural physicians who were planning to move (83%) were planning to move to communities with more than 10 000 people. The 24.5% number is similar to that reported in a study looking at rural physician directory information, and is similar to two US studies, where 20% and 25% of rural physicians indicated that they were considering relocation.[9,10,37]

Work-related factors appear to be increasingly important in the decision to relocate. In the November 1989 survey, physicians were asked “If you intend to move, to what extent is the main reason related to…?”: personal/family (79%), professional (66%), community (51%), and income (44%) were listed in order of decreasing importance. In the July 1998 survey, physicians were asked “If you intend to move, what are key motivating factors…?” and professional reasons (e.g., on-call responsibilities) (81%), daily workload issues (70%), income reasons (e.g., on-call remuneration) (66%), were listed ahead of personal/family reasons (e.g., cultural opportunities) (51%), educational opportunities for children (45%), and spousal opportunities (43%) in this latter study. Workload has been shown in other studies to be important in the retention of primary care physicians in rural areas.[38,39] Moreover, a West Virginia study showed that one of the most important factors necessary to attract and retain family doctors in rural practice was the availability of other family physicians to share call.[40]

Given that satisfaction with certain aspects of work is lower among British Columbia’s rural physicians compared with urban physicians, and given that there is an association between work-related irritations (e.g., too much call, too many job-related sleep interruptions), and intention to relocate, shouldn’t there be a relationship between physician satisfaction level and community physician retention rates? In fact, physicians living in rural isolated communities with high long-term physician turnover rates (low retention) do not have higher burnout scores or lower job-satisfaction scores than physicians living in rural isolated communities with low long-term physician turnover rates (high retention).[19]

Presumably, an inability to demonstrate a relationship between physician job satisfaction and community physician retention rates means that other factors—such as personal/family ones or community-related ones—also play an important role in rural physicians’ decisions to relocate from a rural community.

With respect to personal/family issues, in their study of clinical and academic specialists Linn and colleagues also found that greater quality of life (life satisfaction) correlated with being married, engaging in sexual intercourse more often, arguing with or emotionally withdrawing from family or friends less often, having fewer health problems, less anxiety and depression, and experiencing less job stress.[20] They pointed out that when it comes to understanding and assessing quality of life among physicians, family life and social relations appear to be at least as important as work-related variables.

Physicians’ spouse/partner satisfaction also has an important role in rural physician retention. Krazanjian and colleagues also surveyed physician spouses/partners in their November 1989 survey. They found that the majority of spouses believed that their input was very important in the decision to stay in their partner’s current practice location. Rural physicians were asked, “How much were you influenced by the following people or events in the choice of location of your current practice?” (1 = not at all; 5 = very much). Practising physicians ranked spouse influence to be the most important (3.19 ± 1.4 [SD]) when making their practice location decision, followed by “desire to live/raise family in similar environment to the one I grew up in” (2.54 ± 1.60), “peers/friends” (2.41 ± 1.43), “postgraduate rural experience” (2.37 ± 1.55), “locum experience” (2.36 ± 1.65), “undergraduate rural experience” (2.14 ± 1.42), “location of internship” (1.69 ± 1.19), “closeness to parents/family” (1.69 ± 1.13), “professor/mentor” (1.53 ± 1.07), and “location of residency” (1.51 ± 1.05).

Kazanjian and Pagliccia found, using multivariate analysis of responses to 44 satisfaction items, that the decision regarding practice location is not based solely on professional/practice considerations, but also on personal, family, and community considerations. For rural physicians, community-related, challenge-of-practice, and professional support considerations ranked ahead of workload/time. Workload/time considerations accounted for only 11% of the variance explained. In contrast, community-related considerations, challenge-of-practice considerations, professional support, and financial security accounted for 18%, 16%, 13%, and 7% (respectively) of the variance explained.[15]

The January 1997 and October 1998 surveys asked rural physicians to rank factors that might improve their current living situation. More community services, better schools, better climate, and less isolation were the most frequently listed things that would improve the rural physician living situation (Table 1).

Unlike the situation with burnout and job satisfaction, plotting community factors (population, quality of school, recreational opportunities, climate, isolation, and cultural milieu) against long-term family physician retention rates reveals some interesting findings.[41] Long-term family physician retention increases with increasing community population, with increasing number of specialists, with increasing Fraser Institute school rating scores, with proximity to ski hills, with a more southerly location, with decreasing isolation (NIA) score, and with decreasing percent of aboriginal population. Multivariate analysis reveals that when community population is taken into account, latitude, percentage aboriginal population, and Fraser Institute school ratings do have additional predictive information. Ski-hill proximity and number of specialists are not independent of community size.

 Limitations

The rural physician populations surveyed were not strictly comparable. First, the November 1989 survey defined rural physicians as those physicians who worked in communities of less than 10 000 people. The other three studies surveyed physicians working in communities that qualified for the Northern and Isolated Allowance, some of which have more than 10 000 people. There is, however, quite a bit of overlap between these two kinds of rural populations. About 84% of British Columbia’s physicians working in communities of fewer than 10 000 people work in NIA communities.[18] Second, the studies differ in the proportion of general practitioners and specialists surveyed (Table 1, Table 2). Future studies will be needed to determine whether there are significant differences between specialist and general practitioner satisfaction levels at any given time.

 Conclusion

Low job satisfaction is associated with both intention to relocate from rural family practice and/or undertaking job action.[9,12] In terms of overall quality of life and overall job satisfaction, there are no obvious differences between British Columbia’s rural and urban physicians. These data suggest that rural physicians as a whole are not as unhappy as is generally believed.[6,8] There are, however, certain aspects of work (e.g., on-call responsibilities, daily workload issues, on-call remuneration) and profession (e.g., lack of time for CME) where rural physicians are clearly more dissatisfied than urban physicians. Physician dissatisfaction with these items also appears to have increased over the past decade or so. Professional and work-related dissatisfaction is associated with intention to relocate, but it does not appear to be the only factor associated with physician retention. Non-work factors including personal/family (e.g., educational opportunities for children), community (e.g., cultural and recreational opportunities), and environmental (e.g., climate) also play a role in a physician’s decision to relocate. Sorting out the relative importance of these factors is the next logical step of any future British Columbia physician satisfaction and retention study.

The available data suggest that if health care planners are interested in improving physician job satisfaction they should probably concentrate their efforts on improving locum coverage, CME opportunities, and ensuring prompt tertiary medical centre support.[42,43] Allowing physicians to individualize their day-to-day workload and on-call schedule is an option worth exploring.[29,44] Admitting students into medical school who have rural backgrounds makes sense as these physicians would more likely be comfortable and satisfied with those things that are facts of life for those who live in British Columbia’s rural communities; namely, the lack of community services, a relative lack of educational opportunities for children, the occasional harsh climate, and isolation.[41]

 Competing interests

None declared.

 Table 1. British Columbia’s rural physician satisfaction surveys.

Table 1. (Continued)

Table 2. British Columbia’s satisfaction surveys: Rural vs urban physicians.

Harvey V. Thommasen, MD, CCFP, Marlene P. van der Weyde, MD, CCFP, Alex C. Michalos, PhD, FRSC, Bruno D. Zumbo, PhD, and Catherine A Hagn, MD, CCFP

Dr Thommasen is professor and chair of community health in the Faculty of Health and Human Sciences at the University of Northern British Columbia (UNBC) in Prince George. Dr van der Weyde is a family physician in Lumby, British Columbia. Dr Michalos is director of the Institute for Social Research and Evaluation at UNBC. Dr Zumbo is a professor of measurement, evaluation, and research methodology and an associate member of the Department of Statistics at the University of British Columbia. Dr Hagn is a clinical instructor at the UBC Department of Family Practice, Prince George site and an emergency room physician at Prince George Regional Hospital.