Mean systolic BP was significantly lower in patients under the care of physicians with low burnout for both the overall population (p = 0.001) and patients with hypertension (p = 0.011). In other words, burnout among family physicians was significantly associated with BP levels. In the case of empathy, we found evidence of better BP control for hypertensive patients under the care of more empathic physicians. In the general population, BP levels were lower in patients under the care of less empathic professionals but the association was not statistically significant. To our knowledge, this is the first study to investigate how levels of burnout and empathy among primary care providers in our health care district might affect clinical outcomes.
Although PRESCAP 2010, a cross-sectional multicenter study conducted in Spain, investigated how primary care physicians manage hypertension in routine practice, it did not consider factors related to either empathy or burnout [28]. In our study population, high empathy was significantly associated with low burnout, supporting the theory that actions designed to improve communication skills and empathic tendencies among health care professionals could help to mitigate burnout.
Although we found a positive association between high empathy/low burnout and better BP control and management, there are no similar studies with which to compare our results, apart from that by Hojat et al.,[14] which detected a significant association between physician empathy and metabolic control in diabetes patients. Indeed, a recent systematic review and meta-analysis of the influence of the patient-clinician relationship on health care outcomes found that very few studies had detected a significant association between empathy and clinical benefits [29].
Empathy, by contrast, has been shown to have a positive impact on the doctor-patient relationship in terms of immediate outcomes, such as greater patient satisfaction [30] and treatment adherence [31]. Adherence to treatment and physician recommendations is a key component of good BP control, particularly in the early stages of management, where lifestyle modifications can have an impact. It has been proposed that empathically engaged physicians communicate more openly with their patients, fostering a climate of conversation in which patients are encouraged to talk about their symptoms and their fears, and in which physicians explain the nature of patients’ disease and the different aspects of treatment [32]. In 2007, a German team found physician empathy to be associated with improvements in several long-term outcomes reported by patients with cancer, including depression and pain perception [33].
In our series, family physicians with high empathy scores were significantly more likely to achieve good BP control, as the patients under their care, including those with hypertension, had a lower mean systolic BP than those under the care of physicians with lower scores on the JSPE. They also had a higher proportion of patients with a systolic BP of <140 mmHg. Although the above differences have many potential explanations, it is possible that patients seen by more empathic physicians are more inclined to follow treatment and lifestyle advice.
Our results for burnout were similar in that physicians with low burnout had significantly better BP control results than those with high burnout (more patients with a systolic BP <140 mmHg and a significantly lower mean systolic BP in both the overall and hypertensive populations).
Patients under the care of family nurses with high empathy scores also had better BP control rates than those seen by nurses with low empathy. However, contrasting with the situation observed for nurses with low burnout, more empathically engaged nurses performed fewer BP tests. We observed both a lower mean systolic BP and a higher number of patients with a systolic BP <140 in both the overall and hypertensive populations, suggesting perhaps that empathic nurses, rather than simply measuring blood pressure, take an active role in encouraging actions than can help to reduce hypertension. While we also observed better BP results in the overall population under the care of nurses with low burnout, we were surprised to find that the opposite was true for the hypertensive population.
Using indicators such as good verbal and non-verbal communication and time spent with patients, several studies have suggested that empathic engagement by physicians can lead to increased patient satisfaction [34, 35] and treatment adherence. Our results suggest that this might also be true in the primary care setting, as better control of hypertension, a condition in which both pharmacologic and non-pharmacologic treatments are important, was achieved by more empathic primary care providers. In future studies, it would be interesting to analyze treatment adherence among hypertensive patients according to levels of physician empathy or burnout.
Our study has both strengths and limitations. Strengths include the high response rate and large sample size, as we analyzed BP data for over 300,000 patients under the care of 60% of all primary care providers in our health district. One of the main limitations of our study is that although we observed statistically significant differences between levels of BP control according to levels of burnout and empathy, we cannot know whether these differences resulted in actual clinical improvements. Because our sample size was so large, it is possible that some associations might have had statistical but not clinical significance. The possibility of a white-coat effect should also be considered, although the effect would have been similar for patients with or without hypertension.
Another limitation of our study is that the physician and nurse burnout and empathy levels were based on questionnaire responses, and there is an obvious risk that the respondents may sometimes have answered what they thought was “expected”, rather than what they truly believed. While this risk is greater in the case of socially sensitive subjects such as empathy and burnout, we believe it was minimized by the use of two widely used and validated questionnaires: the MBI and the JSPE. The interpretation of our findings is also limited by a lack of comparative data.
We believe that the limitations associated with the empathy and burnout questionnaires are acceptable as these tests have been amply validated and widely used to analyze empathy and burnout in other studies. In our study, we first analyzed empathy and burnout levels and then looked at BP results for patients under their care. We could not have done this the opposite way around as this would have been a breach of confidentiality, although such an approach would certainly have improved the power of the study.
Finally, we did not control for the multiple factors that could influence BP control, such as treatment adherence, age, and concomitant disease. Moreover, the BP measurement heterogeneity can be considered as a limitation because we cannot assure that all the professionals measure BP in the same way (, patient standing, sitting, laying, with or without a rest of different durations).
Our primary aim was to investigate a possible association between levels of empathy and burnout among family physicians and nurses and different markers of BP management. We did not study other potentially spurious variables. We are well aware that multiple factors can influence empathy, burnout, and BP control, such as patient age, number of years working at the same centers, organization of the medical team, etc. Information on these factors, however, was not available in the database, so we simply analyzed the association, without controlling for other factors. It would also be interesting in future studies to investigate different factors that can influence burnout and empathy, such as working environment and personal, family, and social factors.