Quality of Work Life of Independent vs Employed Family Physicians in Wisconsin: A WReN Study

Study Groups

For analysis, we grouped the physicians who were solo family physicians and members of smaller independent groups (groups not owned by a larger parent organization) into a single group—the independent group. In general, members of this group ranged from solo family physicians (very few) to independent groups of up to roughly 20 physicians. We then grouped those physicians who reported having a parent organization, such as a managed care organization, a large multispecialty group, or a hospital, into a single group—the HCO-employed group. This group included family physician–only groups and single physician offices that were owned and managed by larger HCOs. Physicians working in academic teaching practices were included in the HCO group as well. We defined the work group as “the group of physicians you work with on a daily basis—generally, that is the people with whom you share call.”

Questionnaire

The questionnaire was developed using questions from the literature and from telephone focus groups of both family physicians and their clinical assistants. The questionnaire included a total of 77 content questions (for which responses were given on Likert scales with 5 points, excepting the scale for intention to leave, which had 7 points) plus 12 demographic questions. Space was provided for open-ended comments.

On the basis of the literature and concerns of the focus groups, we selected 13 questions in 3 categories for analysis.

Four questions addressed work satisfaction issues:

1. Satisfaction with income: “How satisfied are you with your current income?” (0 = not satisfied, 4 = very satisfied.)⁴

2. Amount of family time: “How satisfied are you with the amount of family time you have?” (0 = not satisfied, 4 = very satisfied.)⁵

3. Quality of relationships within the work group: “How would you rate the quality of the working relationships among the physicians in your work group?” (0 = poor, 4 = excellent.)⁶

4. Continuity of care: “How satisfied are you with your ability to provide continuity of care?” (0 = not satisfied, 4 = very satisfied.)⁷

The following 5 questions addressed practice issues:

1. Often work under time pressure: “How often do you work under time pressure?” (0 = never, 4 = always.)⁸

2. Amount of paperwork is reasonable: “Do you agree or disagree that the amount of paperwork you process is reasonable?” (0 = strongly disagree, 4 = strongly agree.)⁹

3. Influence over management decisions: “How much influence do you have over management decisions that affect your practice?” (0 = very little, 4 = very much.)⁷

4. Ability to match time to complexity of patient: “How often are you able to match the amount of time you have to spend with patients to the level of complexity of each patient’s case?” (0 = never, 4 = always.)¹⁰

5. Opportunity to fully use skills: “How satisfied are you with your opportunities to fully utilize your skills in your practice situation?” (0 = not satisfied, 4 = very satisfied.) (New item from focus group.)

The following 4 questions addressed outcomes:

1. Satisfaction with being a physician: “How satisfied are you with being a physician?” (0 = not satisfied, 4 = very satisfied.) (New item from focus group.)

2. Perceived quality of care: “Given your work situation in total, how would you rate the overall quality of the medical care you are able to provide?” (0 = poor, 4 = excellent.)¹¹

3. Ability to achieve professional goals: “To what extent are you able to achieve your overall professional goals within your current practice situation?” (0 = not at all, 4 = very much.)¹²

4. Intention to leave the practice: “I plan to leave my work group in the near future.” (1 = strongly disagree, 7 = strongly agree.) (New item from focus group.)

Statistical Analyses

We conducted statistical analysis with SPSS software version 11.5 (SPSS Inc, Chicago, Ill). To compare the 2 groups on demographic variables, we used the □χ ² test for dichotomous variables (eg, sex), and we used the independent sample t test with equal variance not assumed to compare means.

Three separate multivariate analyses of covariance (MANCOVAs) were used to examine whether independent physicians differed from physicians working for HCOs on the 13 study variables. The single independent variable for the 3 MANCOVAs was health care organization, which had 2 levels: independent or HCO. The 5 covariates in each of the 3 equations were age, sex, ethnicity, hours worked per week, and years at current practice. Although the groups differed significantly in years in practice overall, this measure was highly correlated with years in current practice, so only the latter was retained. The 3 groups of dependent QOWL variables were work satisfaction issues (4 variables), practice issues (5 variables), and outcomes (4 variables). The Box M test was used to assess homogeneity of covariance matrices. To further protect against type I error, if the omnibus F test for the MANCOVA was significant, it was followed by univariate F tests of the individual dependent variables with Bonferroni correction, after correction for the covariates.

The total N of 584 classified respondents (78 independents, 506 HCOs) was reduced in each of the 3 MANCOVAs (work satisfaction issues, practice issues, and outcomes) to 442 (55/387), 468 (63/405), and 451 (58/393), respectively, because of missing data. Multivariate normality was not directly assessed, as the sample size in both groups far exceeded 20, the minimum typically required for robust MANOVA in the case of violations of multivariate normality. The Box M test indicated violations of homogeneity of variance-covariance matrices for the work satisfaction and outcomes dependent variables, but not for the practice issue dependent variables. Inspection of the variances between the HCO-employed and independent groups for each dependent variable, however, showed that no ratios exceeded even 2:1, so the use of MAN-COVA was not invalidated and the analysis proceeded as planned.¹³

Comparison of Independent and HCO-Employed Respondents

Table 1⇓ shows a comparison of the independent and HCO-employed physician groups. There were non-significant differences between groups for sex and age. There were significant differences for number of hours worked per week, number of individuals in the work group, years in practice, and years in the present practice. The independents worked more hours per week in smaller work groups and had been in practice longer and in their present practice longer.

Comparison of the QOWL Dependent Variables Between Independent and HCO-Employed Respondents

Table 2⇓ shows a comparison of the groups with respect to the 13 QOWL dependent variables; the values given for these variables in the text that follows are expressed in points as rated by respondents on the previously described Likert scales. Status as an independent vs HCO-employed physician was the independent variable.

Comments by Independent and HCO-Employed Respondents

A total of 120 physicians who could be categorized as either independent or HCO-employed physicians used the “Comments” section of the questionnaire. These physicians did not differ significantly from those who did not make comments in level of satisfaction with being a physician (3.3 vs 3.2, P = .14), perceived quality of care (3.1 vs 3.0, P = .64), and ability to achieve professional goals (2.5 vs 2.4, P = .23), but those providing comments were more likely to agree with the statement “I intend to leave my work group” (2.6 vs 2.1, P = .01, where higher values reflect greater intention to leave).

Physicians often commented on satisfaction being related to having autonomy over their work life and being able to care for patients as they were trained to. Meaningful relationships with colleagues and patients added to satisfaction. Dissatisfaction was expressed over decreasing compensation despite seeing more patients, loss of autonomy over work life decisions, emphasis on productivity, overwhelming paperwork, documentation requirements, patient care decisions being made by business people, and time away from family. The number of comments expressing dissatisfaction greatly outnumbered those expressing satisfaction.

Table 3⇓ contains some quotes from these doctors as they relate specifically to their independent or HCO-employed status. Comments relating to paperwork, payment problems, and other general issues, which did not differ substantively between independent and HCO-employed physicians, are not quoted here.