The Jefferson Scale of Physician Empathy: A preliminary psychometric study and group comparisons in Korean physicians

Abstract

Background: Empathy is an important element of professionalism in medicine. Thus, evaluation and enhancement empathy in physicians is important, regardless of geographical boundaries.

Aim: This study was designed to evaluate the psychometrics of a Korean version of the Jefferson Scale of Physician Empathy (JSPE) among Korean physicians.

Methods: The Korean version of JSPE was completed by 229 physicians in Korea.

Results: Item-total score correlations were all positive and statistically significant. Cronbach's coefficient alpha was 0.84. The mean score was 98.2 (SD = 12.0), which was lower than that reported for American and Italian physicians. The emerged factor structure of the translated version was somewhat similar to that reported for American physicians, although the order was different. Significant differences in the mean empathy scores were observed between men and women and among physicians in different specialties.

Conclusion: Our findings provide evidence in support of reliability and construct validity of the Korean version of JSPE for assessing empathy among Korean physicians. The disparity between Korean physicians and physicians from other countries may be explained by differences in the culture of medical education and medical practice. It suggests an exploration of cross-cultural differences in physician empathy.

Introduction

Definition of empathy

Empathy is an ability to understand other people's experience, emotion, and feelings. Also it involves an ability to transfer this understanding to others (Hojat et al. 2001; Campbell 2004). The cognitive feature of empathy is a crucial factor to distinguish empathy from sympathy which is characterized as an affective or emotional attribute (Hojat 2007). Empathy is manifested consciously, in a relationship, and involves validation of experiences for accuracy of understanding and is free from judgmental bias (Forsyth 1980). On the contrary, sympathy is predominantly an emotional state that is spontaneous. The difference between empathy and sympathy is important in the context of patient care because they have different clinical outcomes (Hojat 2007).

Why empathy is important?

According to some studies, empathy declines during medical education (Newton et al. 2008; Hojat et al. 2009). Overreliance on technology in medical practice may be among the reasons for decline of empathy. In an overly technology-oriented environment, physicians often tend to focus on “test results” rather than “clinical signs” and on “diseases” rather than “patients” (Spiro 1992). This implies that physicians are in need for developing empathic skills in such a dehumanizing clinical environment (Suchman et al. 1997).

Table

Practice points

JSPE affords good reliability and constructs validity for Korean physicians. Lower empathy scores in Korean physicians than in physicians from other countries may be explained by differences in the culture of medical education and medical practice. Cross-cultural differences in physician empathy are evident.

The supporting pillar of empathic physician–patient relationship is built on trust (Bertakis et al. 1991; Levinson & Roter 1995) which in turn increases satisfaction of physicians as well as patients (Nightingale et al. 1991; Suchman et al. 1993) and facilitates history taking and physical examination. As a result, better clinical outcomes will be achieved (Hojat et al. 2002a). In other words, the ability to express empathy is of paramount importance in improving clinical outcomes. It has been reported that over 80% of growing malpractice suits result from miscommunication or lack of empathic engagement, between physicians and their patients. A recent study showed a significant link between physician empathy and optimal clinical outcomes in diabetic patients (Hojat et al. 2011).

Measurement of empathy

Despite the importance of empathy in patient care, empirical research on empathy is scarce, partly because of a lack of a measuring instrument with satisfactory validity and reliability. In response to the need for a psychometrically sound instrument, the Jefferson Scale of Physician Empathy (JSPE) was developed a decade ago (Hojat 2007). One version of the instrument was developed for administration to medical students (S-Version) (Hojat et al. 2001), another version was developed for administration to physicians and other health professionals (HP-Version) (Hojat et al. 2002b). A third version of the scale was recently developed for administration to students in the health professions schools other than medicine (e.g., nursing, pharmacy, dentistry, etc). The JSPE is a brief scale that includes 20 items, each answered on a 7-point Likert scale (1 = Strongly Disagree, 7 = Strongly Agree) that can be answered in about 10 minutes. A higher score indicates a higher degree of empathic orientation. Evidence in support of validity and reliability of the scale among American medical students and physicians has been reported (Hojat et al. 2002b). Previous studies showed the support of psychometrics of JSPE in the United States, Mexico, Poland, Italy, Korea, and Japan (Hojat et al. 2002b; Alcorta-Garza et al. 2005; Kliszcz et al. 2006; Di Lillo et al. 2009).

Recently, an empirical study with a sample of Korean medical students provided support for the validity and reliability of Korean translation of the S-Version of the JSPE (Roh et al. 2010). It is interesting to compare empathy scores of physicians and medical students when the scale is validated for both groups. We compared the JSPE score between medical students and physicians in this study. It is important to study empathy in patient care in different cultures to examine the cross-cultural similarities and differences due to medical education curriculum and social-cultural milieu in different cultures.

Study purpose

We designed this study to examine the psychometrics of the Korean translation of the HP-Version of the JSPE among Korean physicians. We also compared the empathy scores between men and women and among physicians in diverse specialties. And we also compared the results between Korean physicians and Korean medical students from a previous study (Roh et al. 2010).

Methods

We administered Korean version of JSPE to 229 physicians (126 men, 103 women, mean age 28.1 years, SD = 3.28 years) in Seoul National University Hospital. These physicians were practicing in different specialties. After mailing them the Korean translation of the JSPE, we called each of them personally to explain the purpose of the study and asked them to cooperate with us by completing and returning the JSPE. They were not compensated for their participation.

A Korean translation of the JSPE (HP-Version) was used in this study. The standard back-translation procedure was used. The JSPE (HP-Version) was first translated into Korean and then back-translated into English by two bilingual physicians to assure the accuracy of translation.

Subsequent to the approval of the research ethics committee of the Seoul National University Hospital, the Korean version of JSPE was distributed to physicians in that hospital. The survey was anonymous. Responders were asked to specify their age, gender, and specialty.

Statistical analyses

We used principal component factor analysis (varimax rotation) to examine underlying constructs of the JSPE among Korean physicians. We also used correlations, t-test, and analysis of variance to examine correlates and group differences on the scale's scores.

Results

All of the 229 physicians who received the JSPE, completed the survey (response rate = 100%). We attribute this perfect response rate to our efforts in calling and encouraging physicians in a friendly manner to cooperate in the study.

Descriptive statistics

The mean, SD, 25th, 50th, and 75th percentiles range and Cronbach's coefficient alpha reliability are reported in Table 1.

Table 1.  Descriptive statistics and reliability coefficient for the Korean translation of the JSPE for 229 Korean physicians

Statistics
Mean score	98.2
SD	12.0
Percentile
25th	90
50th (Median)	99
75th	107
Possible range	20–140
Actual range	59–135
Cronbach's coefficient alpha	0.84

As shown in Table 1, the scores ranged from 59 to 135 (possible range 20–140). The median score was 99, and the mean score was 98.2 (SD = 12.0). The 25th and 75th percentiles were 90 and 107, respectively. The mean score obtained by Korean physicians is smaller than that (M = 120) reported for the US physicians (Hojat et al., 2002b). The item mean scores ranged from a low of 4.1 (for this item: “I consider understanding my patients’ body language as important as verbal communication in care-giver-patient relationship”) to a high of 5.7 (for this item: “My patients feel better when I understand their feelings.”). Item-total score correlations ranged from 0.20 (p < 0.01) (for this item: “It is difficult for me to view things from my patients’ perspective”) (a reverse score item) to 0.68 (p < 0.01) (for this item: “I believe that empathy is an important therapeutic factor in medical or surgical treatment”). The median item-total score correlation was 0.54.

Underlying construct of the scale

Factor analysis of the item scores resulted in six factors, each with an eigenvalue greater than one (5.4, 2.1, 1.3, 1.2, 1.1, 1.0). However the screen test of the plot of eigenvalues showed that the magnitude of eigenvalues leveled off after the extraction of the third factor, suggesting that there are only three reliable components in the scale. The first factor, accounting for 27% of the variance was similar to the factor of “compassionate care” that emerged as the second factor in the US physicians (Hojat et al. 2002b). Seven items (out of 10 items under this factor in the US physicians) had a factor coefficient greater than 0.30 on this factor. The item with the largest factor coefficient (0.67) was: “Attentiveness to my patients’ personal experiences does not influence treatment outcomes” (a reverse score item). The second factor, accounting for 10% of the variance was similar to the first factor of “perspective taking” that emerged in the US physicians (Hojat et al. 2002b). Five items (out of eight items under this factor in the US physicians) had a factor coefficient greater than 0.30 on this factor. The item with the largest factor coefficient (0.68) was: “I try to imaging myself in my patients’ shoes when providing care to them.” The third factor was a residual factor accounting for 7% of the variance.

Reliability

The internal consistency reliability of the scale as indicated by the coefficient alpha of 0.84 is quite satisfactory, and in acceptable range for educational and psychological testing suggested by professional testing organizations (AERA et al. 1999).

Group comparisons by gender and specialty

Means and SDs for Korean physicians by gender and specialty are reported in Table 2. As shown in the table, women scored higher than men by 3.8 points, and the difference was statistically significant (t₍₂₂₇₎ = 2.35, p < 0.05).

Table 2.  Group comparisons on scores of the Korean translation of the JSPE by gender and specialty

Groups	n	Mean	SD
Gender^a
Men	126	96.5	12.0
Women	103	100.3	11.7
Specialty^b
Dermatology	55	101.9	9.8
Internal Medicine	11	100.0	13.9
Rehabilitation Medicine	11	99.8	10.7
Internship	85	99.4	11.6
Ophthalmology	12	97.9	10.5
General Practice	21	90.0	9.2
Radiology	11	88.5	10.8
Other	23	95.8	13.1

Notes: ^at₍₂₂₇₎ = 2.35, p < 0.05. ^bSpecialties with few physicians were grouped into “Other” category. F_(7,221) = 3.84, p < 0.01. Dermatology = Internal Med = Rehab Med = Internship > Gen Practice = Radiology = Other.

Mean scores for physicians practicing dermatology, internal medicine, and rehabilitation medicine were significantly higher than those in general practice, radiology, and other specialties (F_(7,221) = 3.84, p < 0.01). No significant difference was observed among physicians in dermatology, internal medicine, and rehabilitation medicine; and those in ophthalmology and internship were not significantly different from any other group.

Discussion

Findings of this study generally suggest that the Korean translation of the HP-Version of the JSPE is a psychometrically sound instrument for measuring orientation toward empathy in patient care in Korean physicians. Somewhat slightly different pattern of factor structure was found among Korean medical students by using the Korean translation of the S-Version of the JSPE (Roh et al. 2010). In Korean medical students, “perspective taking” emerged as the first factor, and “compassionate care” as the second factor. The findings of factor analysis in both samples of Korean physicians and medical students showed a three-factor solution that was somewhat similar to the pattern emerged in American and Italian physicians, which provides support for the construct validity of a Korean translation of the JSPE. Factor analytic results suggest that empathy as measured by the JSPE is a multidimensional concept with similar components in different cultures. The reliability coefficient we found in Korean physicians (r = 0.84) was as high as those reported for American physicians (r = 0.81) and Italian physicians (r = 0.85) (Hojat et al. 2002b; Di Lillo et al. 2009).

We found that the mean score for Korean physicians (mean = 98.2, SD = 12.0) was lower than that reported for American (mean = 120, SD = 12) (Hojat et al. 2002b), and Italian physicians (mean = 115.1, SD = 15.55) (Di Lillo et al. 2009). This disparity among physicians from different countries may be explained by differences in the culture of medical education and medical practice.

For example, Childress and Siegler (2001) examined five models for the physician–patient relationship; (1) paternalism (the physicians as caring parent, the patient as child); (2) partnership (both parties as collaborating in pursuit of the shared goal of the patient's health); (3) contract (physicians and patients as related to each other by specific contracts, detailing their obligations and rights); (4) friendship (physicians and patient as intimately related due to the highly personal nature of health); and (5) technical assistance (the physician as technician, the patient as customer).

In East Asian cultures, physician–patient relationship is vertical relationship (e.g., lord–servant). This notion is mainly derived from Confucian philosophy which places an emphasis on fidelity. Words from physician are considered as compulsory orders which have to be obeyed. This relationship resembles more to the “paternalism” category according to the above-mentioned classification. Physicians give orders to patients to follow their instructions without any question, somewhat similar to parent–child interaction in a disciplinary matter. In this kind of interaction, physician's apparent expression of empathy is not as important, because the physician is supposed to be the “captain of the ship” without requesting the patient's opinion. Asking a patient's opinion will be considered as an indication of physician's uncertainty, lack of competence, and weakness (Hojat 2007); thus, this will be reflected in physician–patient relationship, which could in turn reflect in the physician's responses to the JSPE. Substantial changes have occurred in Korean society toward economic growth and democratic values which is gradually influencing the vertical physician–patient relationships by shifting it into a horizontal one. Therefore, it appears that the relationship is gradually shifting from the “paternalism” to “partnership” pattern. However, the residuals of traditional relationship can still play a role. Also in Korean society, suppression of one's true feeling is regarded as a virtue. In this cultural atmosphere, the attitude of respondents who participated in the survey can be different from western society. This cultural peculiarity may partially explain why the mean of empathy in Korean physicians was found to be lower than their counterparts in the western society, in which a horizontal pattern of physician–patient relationship is prevalent. The aforementioned argument suggests that cross-cultural comparisons of empathy scores without taking into consideration these cultural differences may not be meaningful.

However, comparisons of physicians within a particular culture will still be of value. Similar to previous study with American samples (Hojat et al. 2002b), the empathy mean score for women was significantly higher than that for men, however, in a study with Italian physicians no gender difference on empathy was found (Di Lillo et al. 2009). The consistent findings in several studies, including our study that women score higher than men on empathy can be explained by female's higher sensitivity in interpersonal relationship, and their better understanding of patient's emotional signals (Hojat 2007) which can lead to a better empathic engagement in patient care (Kunst-Wilson et al. 1981; Newton et al. 2008).

We found that the empathy mean score for Korean physicians (mean = 98.2, SD = 12.0) was lower than that reported for Korean medical students (mean = 103.1, SD = 12.5) (Roh et al. 2010). As medical students become physicians, they are exposed to a more technical medical environment with overreliance on biotechnology that can erode their empathy toward patients. Moreover residents do not have sufficient time to spend with their patients to form an empathic engagement due to being overburdened in the hospital. It is reasonable to assume that these factors contribute to the decline of empathy from medical school to the practice of medicine. However, confirmation of this assumption requires a longitudinal study design. Also we found that the factor structure was slightly different between two groups. In students “perspective taking” is the first factor. However, “compassionate care” is the first one in physicians. It is interesting that the priority is changed from students to physicians. Probably this change is due to practical problems in Korean health care system (e.g., insufficient time to treat a patient, numerous responsibilities of physicians in medicolegal problems).

Findings on comparing empathy scores of physicians in different specialties were in more agreement with those reported in American physicians by Hojat et al. (2002b), than those in Italian physicians reported by Di Lillo et al. (2009). Similar to findings of Hojat and colleagues (2002b), we found that physicians in the so-called “people-oriented” specialists such as dermatologists, internal medicine, and rehabilitation medicine tend to obtain higher empathy scores than their counterparts in the “procedure-oriented” specialties such as radiology. However our findings of lower empathy mean score in general practitioners is not consistent with those reported by Hojat and colleagues (2002b). It is reasonable to assume that physicians in “people-oriented” specialties require more direct contact and talk with patients than others in “technology-oriented” specialties. The difference can also be explained by differences in academic preparation training of physicians in various specialties (Harsch 1989).

One limitation of this study is that participants were from only one hospital that may not represent the broader population of physicians in Korea. The Seoul National University Hospital ranks very high among Korean medical health care systems, and probably clinical environment may tend to be more authoritative than other hospitals in Korea. Also the extent of experience and small sample size of physicians in different specialties could jeopardize the statistical power of the findings. Thus, caution should be exercised in generalizing the findings to a broader population of Korean physicians.

Despite these limitations, psychometric findings are encouraging, and can add to the confidence of Korean investigators who want to use a psychometrically sound instrument for measuring empathy among physicians and other practicing health professionals in Korea.

Conclusion

Our findings provide evidence in support of reliability and construct validity of the Korean version of JSPE for physicians, and suggest that the JSPE can be used with confidence for assessing empathy among Korean physicians. The disparity between Korean physicians and physicians from other countries may be explained by differences in the culture of medical education and medical practice. It suggests us to explore cross-cultural differences in physician empathy.

Ethical approval: This study was approved as exempt research by the Institutional Review Board of Seoul National University Hospital. The work was carried out in accordance with the Declaration of Helsinki. There was no potential harm to participants and anonymity was maintained.

Declaration of interest: The authors report no conflicts of interest.