Physician Characteristics and Beliefs in Sharing
Seventy percent of the study physicians were men, and 71% were white (the largest group of nonwhites were Asian/Pacific Islanders [n=6]). Their mean age was 43.9 years, with a median of 13 years since graduation from medical school. They had been affiliated with their current system for a mean of 8.3 years. Ninety-six percent of the physicians were board certified in their primary specialty, and they spent a mean of 39.1 hours in patient care weekly.
Although the physician sample was limited in size, we also explored the relationship of physicians’ PPOS scores to their demographic and personal characteristics. In contrast to the patients, the scores of men and women were very similar (4.52 and 4.45, respectively), and no significant differences were found according to ethnicity, specialty, time spent in patient care, or workplace satisfaction. Also, beliefs in power and information sharing did not differ according to experience, either by splitting physicians at the median on age or the time since their graduation from medical school.
Attitudes Toward Sharing and Patient Evaluations
Patients’ evaluations of their physicians and their visits were measured in 3 different ways: trust (pre-visit), visit satisfaction, and endorsement of physician (both post-visit). Although these measures were themselves highly intercorrelated (between 0.45 and 0.48), separate GEE analyses were performed for each. For each measure, the analysis was run 3 times, each using a different predictor. First, we entered the patients’ PPOS scores as the only predictor, then the physicians’ PPOS scores, and then the difference between the patients’ and physicians’ scores (patient minus physician). As indicated in Table 2, visit satisfaction was not significantly related to any of the predictors. However, patient-centered patients and those whose attitudes were discrepant from their physicians, were both significantly less trusting and less likely to endorse their physicians. Physicians who were patient-centered were marginally more likely to be trusted (P=.09).
Since patients’ PPOS scores were related to several other variables, the GEE analyses for those variables showing significant (P <.05) associations between beliefs and belief congruence as predictors and patient evaluations as outcomes were run a second time, controlling this time for sex, age, education, income, and ethnicity. The results of these analyses did not weaken any of the relationships. Patient PPOS and degree of congruity were each found to be stronger independent predictors of the trust and endorsement than any of the potentially confounding variables.
Discussion
The results of our study provide us with information about where patient-centered beliefs reside. Among patients, a belief that power and information should be shared appears to be a cultural phenomenon; younger age, female sex, white ethnicity, higher income, and more education were all closely associated with a desire for sharing. Yet we found a somewhat unexpected pattern for age: There was little difference among the 3 youngest categories (18-39, 40-49, and 50-59 years), and then relatively sharp drops among those in their 60s and older. These findings suggest that if there is a generation gap in patients’ beliefs about empowerment, it exists not so much between younger and middle-aged patients as it does between those older than 60 years and those younger than 60 years.
Although the comparable physician data have to be interpreted with extreme caution because of the small sample size, we found that physicians are apparently less affected by those same societal factors that shape patients’ attitudes about sharing of power and information. Consistent with previous administrations of the scale to other physician samples,24 male and female physicians did not differ in their patient-centered beliefs, nor did we find significant relationships between patient-centeredness and physician experience. Contrary to the stereotype that older physicians take a more authoritarian orientation toward patient relationships, the data suggest that patients seeking a physician who values information and power sharing are likely to be disappointed if they merely use physician age as a proxy for patient-centeredness.
Perhaps the most significant finding of this study was that the degree to which patients and physicians held similar orientations was a strong predictor of 2 of the 3 patient evaluation measures. Patients whose beliefs were congruent with their physicians’ beliefs trusted them more, as indicated before they completed the target visit. After the visit, they were also more likely to recommend to others, follow the advice of, and make a special effort to see their physicians (the 3 components of the endorsement index).
Limitations
Generalizations from our data are limited not only by the small sample size of physicians, and by the fact that the patients and physicians all came from managed care systems in one region of the country. Another limiting factor may be that the visits studied represented a targeted subsample of patients who had an ongoing or worsening problem that concerned them. Nonetheless, the most surprising finding was that physicians who held patient-centered beliefs about power and information sharing were rated no more positively on measures of satisfaction, trust, and endorsement. One possible explanation for this may have to do with the study sample of patients, all of whom had a significant problem or concern. Previous research15,16 has indicated that physicians act differently toward their patients who are more ill or more emotionally distressed, showing greater signs of conflict or tension. It is therefore possible that the power-sharing beliefs of the patient-centered physicians were not translated as directly into action in the course of treating these patients. A second possible explanation is that patients who have strong health concerns may actually want their physicians to revert to ways that are more authoritarian and to take greater control during the course of the visit.