The Association Between Well-Being and Empathy in Medical Residents: A Cross-Sectional Survey

Abstract

Objective:

To evaluate the extent to which personal well-being may be associated with empathy, while controlling for potential confounders.

Settings/Location:

Residency programs throughout the United States.

Subjects:

A total of 407 medical residents from residencies including general medicine, surgery, specialized and diagnostic medicine participated in this study.

Outcome Measures:

Well-being was measured using the modified existential well-being subscale of the spiritual well-being scale. Empathy was measured using the Jefferson Scale of Empathy.

Results:

Well-being was found to be positively correlated with empathy when adjusted for possible confounders (p < 0.001). In addition to well-being, other factors noted to be statistically significant contributors to higher empathy scores while controlling for the others included age, gender, year in residency, specialty, and work-hours (p < 0.05 for each). After controlling for these factors, a resident's year in residency was not found to be a statistically significant contributor to empathy score.

Conclusions:

In this study, well-being was associated with empathy in medical and surgical residents. Empathy is a fundamental component of physician competency, and its development is an essential aspect of medical training. These findings suggest that efforts to increase well-being may promote empathy among medical residents.

Introduction

As the rates of burnout, mental health issues, and suicide among physicians are at an all-time high, there is an urgent need to protect physician well-being.¹ Studies show that psychologically distressed physicians are more prone to making medical errors, experiencing professional lapses, and leaving medical training altogether.^2–4 Residency is a particularly unique and stressful period in medical training, as newly minted physicians face a multitude of formidable challenges that frequently lead to personal and psychological distress.^5,6 These challenges include assuming primary responsibility for patients, working long hours in the hospital, managing financial pressures of student debt, and/or moving to a new city without well-developed support systems.^7,8

Residents are thus assuming their professional identities while facing a multitude of challenges, and studies suggest that the timing of these stressors contributes to the erosion of empathy and the decline in quality of patient care that residents provide.^9–11 A decline in resident empathy is a particular concern, as research suggests that empathy is an essential component of physicianhood.¹²

Studies have also found that physicians with higher empathy scores are more likely to have patients with significantly better control of their chronic diseases and improved quality of life.^13,14 This may be because empathy enhances a physician's diagnostic accuracy, increases patient compliance to treatment recommendations, and improves patient satisfaction.^12,15,16 For these reasons, empathy is a well-established component of clinical competency in medical education.¹⁷

Although previous studies have looked at well-being and empathy in medical students and other health care providers, less is known about residents.^4,18 Previous studies looking at empathy in residents have been small and focused on one medical specialty.^5,11 Less is known about the potential association between well-being and empathy among residents in different medical specialties and among residents as a group.

The purpose of this study was to survey residents from various specialties across the United States to understand the association between well-being and empathy. An exploratory analysis was conducted to assess whether well-being is associated with empathy, when controlling for potentially confounding factors including age, gender, year in residency, specialty, and work-hours. This study adheres to The Strengthening the Reporting of Observational Studies in Epidemiology Guidelines from the Equator Network (Supplementary Fig. S1).

Methods

Study design and participants

A cross-sectional survey study was designed to measure well-being and empathy scores among medical residents in the United States. Only residents in the United States who were postgraduate year (PGY) 1 through PGY 6 and self-identified as men or women were eligible to participate. Only fully completed surveys were analyzed. The study was certified as exempt from institutional review board (IRB) oversight requirements on the basis of category (d)(2) under 45 CFR 46.104 by the Vanderbilt University School of Medicine Institutional Review Board (IRB #211544). The IRB waived the requirement of written consent documentation; consent was assumed based on willingness to complete the survey.

Survey design and distribution

The Qualtrics-based study consisted of six blocks of questions that assessed demographics, residency program characteristics, work–life balance, empathy, well-being, and flow (Supplementary Fig. S2).

Well-being was assessed using the 10-item existential well-being (EWB) subscale of the spiritual well-being scale modified to a 5-point Likert scale.^19–21 The EWB measures positivity, satisfaction, and life fulfillment in a nonreligious language.^22,23 This scale has been translated in >10 languages and used in hundreds of studies and dissertations. This application of the EWB subscale is consistent with a definition used by the Center for Disease Control and Prevention (CDC), in which well-being is defined as the state of positivity, satisfaction with life, and fulfillment.²⁴

Participants responded to each item on a 5-point Likert scale from 1 to 5 with response options labeled not at all, a little bit, somewhat, quite a bit, and a lot. Scores on 5 negatively worded items were reversed scored, and all 10 items were summed. Possible scores ranged from 10 to 50, with higher scores representing greater well-being.

Empathy scores were evaluated using the Jefferson Scale of Empathy (JSE).²⁵ The JSE is a 20-question validated survey commonly used to measure empathy in physicians and other health care professionals involved in clinical patient care. Participants responded to each item on a 7-point Likert scale from 1 to 7 with response options strongly disagree, disagree, somewhat disagree, neither agree nor disagree, somewhat agree, agree, and strongly agree.

Scores on 10 negative items were reversed scored, and all 20 items were summed. Possible scores range from 20 to 140, with higher scores suggesting more empathetic behavioral patterns. The JSE has been translated into >50 languages, has been used by >80 countries, and has been recognized as the most researched and widely used instrument in medical education research.²⁶

Additional baseline demographic information was collected including age, gender, and year in residency. Information regarding residency program characteristics, including specialty and average work-hours within the past month, was also obtained.

Survey distribution

The survey was distributed through the mailing lists of two medical societies, the Association of Academic Physiatrists and the American Medical Women's Association. Participation was voluntary and responses were collected anonymously. The consent page for the survey described the study as looking at “wellness, empathy, and flow” in medical residents. At the end of the survey, residents were given the opportunity to provide their email addresses to enter a random lottery to win an Amazon gift card.

Statistical analysis

All data used for analysis are available for review (Supplementary Fig. S3).

Respondent age was dichotomized to ≤29 or ≥30 years old based on the average age at which residents were projected to start residency, given that the average age of matriculating medical students is 23–25 years old.²⁷ This dichotomy reflects an assumption that the effects of age are unlikely to be strictly linear. Linearity is implicitly assumed when “raw” age is used in statistical models. Furthermore, the effect of age may be more categorical. Residents who are older are more likely to have taken time off during or before starting medical training, or they may have entered medicine as a career change. Such residents are likely to have different life experiences than those who went straight through undergraduate and medical school and directly into residency.

Information on specialty profiles from the Association of American Medical Colleges Specialty Profile was used to sort participants' medical specialties into four major categories as follows: general medicine, specialized medicine, diagnostic medicine, and surgery.²⁸ The general medicine category included internal medicine, pediatrics, emergency medicine, and family medicine. The specialized medicine category included nonoperative subspecialties within medicine that require additional advanced training after internship, such as cardiology, hematology, rheumatology, physical medicine and rehabilitation, neurology, and psychiatry.

The surgery category included all operative specialties such as general surgery, neurosurgery, orthopedic surgery, and plastic surgery. The diagnostic medicine category included pathology and radiology.

Participant work-hours were combined into two categories as follows: ≤59 or ≥60 h per week. Sixty hours per week was chosen as the cutoff, as the majority of practicing physicians work <60 h per week.²⁹

The potential association between well-being and empathy was assessed with a least-squares multiple-regression analysis including the five potential confounders of demographic and residency characteristics: age, gender, year in residency, specialty, and work-hours.

Results

Survey completion

The survey was conducted from September to December of 2021. During this time, a total of 435 residents from various specialties including general medicine, specialized medicine, diagnostic medicine, and surgery responded to the survey. A total of 407 (93.56%) participants provided complete responses; only complete responses were used in the analysis.

Participant demographics

A slight majority of participants were ≥30 years old (212 residents, or 52.09%), and a slight majority were women (211 residents, or 51.84%). Residents varied in their level of training from PGY 1 (interns) to PGY 6 residents, with most participants (250 residents) being in the PGY 2 to PGY 4 range. There were 70 PGY 1's (17.20%), 85 PGY 2's (20.88%), 83 PGY 3's (20.39%), 82 PGY 4's (20.15%), 58 PGY 5's (14.25%), and 29 PGY 6's (7.13%) who participated in the survey (Table 1).

Table 1.

Demographic Information of Residents Responding to the Survey

Characteristics	No.	(% of 407)
Age
≤29 years	195	47.91
≥30 years	212	52.09
Gender
Female	211	51.84
Male	196	48.16
Year in residency
PGY 1 (intern)	70	17.20
PGY 2	85	20.88
PGY 3	83	20.39
PGY 4	82	20.15
PGY 5	58	14.25
PGY 6	29	7.13

PGY, postgraduate year.

Of the respondents who fully completed the survey, 135 (33.17%) were residents in general medicine, 130 residents (31.94%) in specialized medicine, 39 residents (9.58%) in diagnostic medicine, and 103 residents (25.31%) in general surgery. Most (268 residents, or 65.85%) worked ≤59 h per week (Table 2).

Table 2.

Residency Characteristics of Residents Responding to the Survey

Characteristics	No.	(% of 407)
Specialty
General medicine	135	33.17
Specialized medicine	130	31.94
Diagnostic medicine	39	9.58
General surgery	103	25.31
Work-hours
≤59 h per week	268	65.85
≥60 h per week	139	34.15

Participant well-being and JSE scores

The mean (±SD) EWB well-being score was 38 ± 8. JSE empathy scores were found to be bimodally distributed, with peaks in the 80–85 and the 110–115 range. The median (IQR) empathy score in the sample was 110 (90–119).

Factors that contribute to empathy

A least-squares multiple regression model was used to evaluate the relation of well-being scores with empathy scores, while controlling for potentially confounding factors. After controlling for potential confounders, a linear relationship between well-being scores and empathy scores was found that was significant at p < 0.001. On average, each additional point of the well-being score was associated with a 1.1-point increase (95% confidence interval: 0.9–1.2) in empathy score. Considered factors found to be statistically significant contributors to empathy while simultaneously controlling for the other factors included age (p = 0.047), gender (p < 0.0001), specialty (p = 0.002), and work-hours (p < 0.0001).

Age >30 years, female gender, specialized medicine, and working >59 h per week were all associated with higher empathy scores. While adjusting for the other factors, being >30 years old increased empathy scores by an average of 1.42 points, being women by an average of 3.9 points, being in a specialized medicine residency by an average of 2.84 points, and working >59 h per week by an average of 3.01 points. Year in residency was not found to be a statistically significant contributor when controlling for the other factors.

Table 3 gives the coefficients for the overall model using the five factors to predict empathy scores. All factors in Table 3 were simultaneously considered as potential confounders on the association between well-being and empathy. The effect column shows the number of empathy scale points that are associated with each factor in relation to the reference, and the p-value column shows significance for a test of the null hypothesis that the estimate for this step or factor should be 0.

Table 3.

Overall Model Using Five Factors to Predict Empathy Scores

Variable	Effect (empathy points)	95% confidence interval	p
Age
≥30 years	Reference
≤29 years	−1.42	−2.83 to −0.017	0.047
Gender
Male	Reference
Female	3.90	2.57 to 5.23	<0.0001
Year in residency
PGY 1	Reference
PGY 1–PGY 2	−1.06	−5.29 to 3.17	0.62
PGY 2–PGY 3	−0.25	−4.24 to 3.75	0.90
PGY 3–PGY 4	−2.26	−6.24 to 1.71	0.26
PGY 4–PGY 5	0.21	−4.19 to 4.62	0.92
PGY 5–PGY 6	1.82	−4.04 to 7.68	0.54
Specialty
Surgery	Reference
Diagnostic medicine	−2.76	−6.04 to 0.52	0.099
General medicine	2.51	0.30 to 4.72	0.026
Specialized medicine	2.84	0.62 to 5.06	0.012
Work-hours
>59 h per week	Reference
≤59 h per week	−3.01	−4.50 to −1.52	<0.0001
Total well-being score	1.07	0.91 to 1.23	<0.0001

PGY, postgraduate year.

The default participant (characteristics selected by the software) is a resident >30 years old, man, PGY 1, and surgery resident working ≥60 h per week. To predict any given resident's empathy score, the value from the “estimate” column should be added or subtracted. The only continuous variable in the model was well-being score, and the listed estimate reflects additional empathy points for each scored point on the well-being instrument.

Discussion

This was an exploratory analysis of a survey on the association between well-being (as measured by the EWB subscale) and empathy (as measured by the JSE) in a cross-section of medical residents working in different parts of the United States. This pilot study demonstrated that there may be a nearly one-to-one association between well-being and empathy, when controlling for potentially confounding factors including age, gender, year in residency, specialty, and work-hours. The results of this study suggest that well-being and empathy are positively associated, even after adjusting for the effects of likely confounders.

Intuitively, the possible association between well-being and empathy is highly plausible. Empathy involves a cognitive ability to perceive the emotions of others, to resonate with them emotionally, and to understand their perspectives.³⁰ This ability is an essential trait for physicians who must tailor recommendations to the personal goals of each patient, making empathy an essential aspect of resident competency.³¹

One could postulate that having a positive outlook and sense of well-being gives a resident the capacity to think beyond oneself and consider another's perspective and needs. Indeed, there is increasing evidence to suggest that a physician's competency is linked to a physician's personal well-being.^9,11,32 This relationship has been studied within specialty-specific residency programs, and this study suggests that this is the case on a national level.

A variety of tools have been proposed to cultivate empathy, not only for use among physicians and health care workers but also for use among business leaders, designers, architects, and engineers. These empathy fostering tools include communication training, service to underserved communities, shared reflective exercises, narrative writing, clinical vignettes of culturally and linguistically diverse patients, and virtual reality simulations.^33,34 Some approaches to help physicians further develop empathy also incorporate promoting physician wellness through mindfulness meditation and other well-being initiatives.³⁵ As this study suggests that well-being is closely related to empathy in residents from different specialties, well-being initiatives should be incorporated into empathy training curricula and vice versa.

Data of this study reveal several interesting findings. In this study, age ≥30 years was found to be a statistically significant contributor to increased empathy. One could postulate that older age is associated with increased life experience, which might correlate with increased ability to understand different perspectives. However, evidence regarding the effects of age on compassion and empathy remains conflicted, with some studies suggesting a neural mechanism for increasing emotional empathy that correlates with increasing age and other studies showing no association at all.^36,37

Interestingly, the residents' year in residency did not show a consistent effect direction; although the effect sizes were not statistically significant, additional years of progress in residency sometimes added empathy points and sometimes subtracted them. This result lends some support to the authors' decision to dichotomize respondent age.

In addition, women had statistically significant higher empathy scores than men when controlled for other factors, which is consistent with many prior studies.^38–41 These prior studies postulate that a probable explanation for this finding could be the higher level of emotional receptivity in women than in men and a biological evolutionary basis for these differences, which may translate to higher empathy scores in women.⁴²

In terms of residency program characteristics, this study found no association between year in residency and level of empathy. Indeed, prior studies suggest that empathy is a relatively stable trait that remains unchanged during residency, which is consistent with the findings.^39,43 This study also found that there is a possible association between resident specialty choice and empathy scores, with lower empathy scores in residents pursuing diagnostic specialties. Results from other studies regarding the role of specialty preference on empathy remain mixed.

A minority of studies have suggested that empathy scores are lower among residents in specialties that involve less direct patient contact than among residents in specialties with more direct patient contact, whereas other studies have found no association between specialty preferences and empathy scores.^38,44 The implications of these results remain unclear, and further study is needed to evaluate whether an association indeed exists.

Lastly, working ≤59 h per week was a statistically significant contributor to lower empathy scores. This finding conflicts with prior research suggesting that long hours and difficult training contribute to the decline in empathy among medical residents.^45,46 This study seems to suggest that working more than the average number of hours per week is correlated with increased empathy scores. Though the implications remain unclear, perhaps more time spent working with patients helps foster empathy and meaning in the population of residents who participated in this study, thereby increasing their empathy scores.

Limitations

This study has several limitations. First, this was an exploratory study. The study was cross-sectional by design, and, therefore, any causal conclusions could not be drawn. Although findings suggest that resident physician well-being may be associated with empathy, the contribution remains uncertain. A prestudy power analysis was not conducted to determine the significance of year in residency, specialty, or other contributing factors. Although the number of participants in the analysis (n = 407) is fairly large, when considering that the number of medical residents in the United States is in the 100,000’s, this study captured only a small sample of the overall resident population.⁴⁷

Thus, the generalizability of this study is still unknown. The timing of the survey a few months into the new academic year was chosen to optimize resident participation. However, the study also took place during a global pandemic, a time in which well-being and/or empathy levels among health care professionals might be lower than expected. Indeed, the mean JSE score for this population was 110, which is slightly lower than the average of 116–120 found in other nationwide studies looking at JSE scores among health professionals.^48,49

Several different biases may also affect the study results. Participation bias could be a factor. Residents with lower levels of well-being or less empathetic behavior may be less likely to respond to this survey, given a lower level of perceived importance or competence in this subject matter. The well-being and empathy scales were additionally self-reported, and there may be concerns for reporting bias. Given that the nature of medicine is a human-centered field, residents may feel pressured to choose responses that demonstrate a higher level of empathy and well-being.

However, it could be argued that well-being is a person's self-reported outlook and use of a self-reported scale is thus appropriate. Finally, these scales are subject to a degree of cultural biases, as the determination of high well-being or highly empathetic behavior on which these scales are based is subject to social perceptions of what defines such a state.

Future directions

The results of this study suggest that well-being is associated with empathy when considering a group of medical residents. Thus, it may be beneficial for residency training programs to consider focusing on well-being in addition to empathy when designing their curricula.

Although empathy is a well-established core medical competency, how to foster this skill remains largely unknown. Numerous interventions to improve empathy have been developed, and initial evaluation of such programs suggests that empathy is a skill that can be acquired and refined.³⁴ However, no standard curriculum for fostering empathy has been established.

This study suggests that one approach to improving empathy may be to consider a greater emphasis on improving resident well-being. When developing residency training curriculums, a greater emphasis on promoting personal well-being may lead to increased empathy and excellent patient care. Future studies on the relationship between well-being and empathy are needed to further clarify and validate the demographic and residency program factors that affect empathy scores.

Conclusion

This study suggests that resident physician well-being may be associated with empathy. Residency training is a critical phase for development of lifelong skills and behaviors, including empathy. Empathy has been shown to affect clinical outcomes, and thus, empathy has become a core clinical competency in most residency training programs. Given the association between well-being and empathy, educational efforts to increase empathy in residents should also include efforts to increase well-being and vice versa.

Authors' Contributions

T.L.S. contributed to conceptualization, methodology, data collection, and writing—original draft. R.E. was involved in data analysis and interpretation of results, and writing—original draft. R.T.T. took charge of software and survey creation, data collection, and writing—original draft. T.R.P. carried out data analysis and interpretation of results, and writing—review and editing. D.J.K., M.R., M.O., D.L.N., and M.F. carried out conceptualization, and writing—review and editing.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.