Effect of physician attire on patient perceptions of empathy in Japan: a quasi-randomized controlled trial in primary care

Most patients report that physician attire is important and associated with their satisfaction with care [1‐3]. Patient preference for physician attire is influenced by age, locale, setting, and context of care [1, 3‐8]. Reports from several countries suggest that patients prefer primary care physicians to wear white coats [4‐6]. However, in some countries, most patients who visit a family physician (FP) no longer consider white coats a powerful symbol [8]. Previous research in Japan shows that most patients prefer physicians to wear white coats in a primary care setting [1]. However, one study found that some family medicine specialists certified by the Japan Primary Care Association (JPCA), which was established in 2010 following the merger of three primary care academic societies, preferred non-white coat attire, because they felt that casual attire allowed more empathetic interactions with patients [9]. However, there are no studies on whether FP attire influences relational empathy as perceived by patients in primary care settings.

Empathy contributes to effective general practice consultations [10] and has many beneficial effects in terms of health care, such as improved patient satisfaction, better medication adherence, greater patient enablement, and better clinical outcomes [11‐15]. The identification of specific nonverbal behaviors that enhance patient-perceived relational empathy may be important for building efficient therapeutic relationships and optimizing patient health outcomes [15‐18].

One study in a traditional medical clinic in Korea showed that patient-perceived empathy was substantially higher when physicians wore white coats and traditional dress than when they wore casual attire and suits [19]. However, a United States study of a large online sample in an analog medical context that manipulated physician nonverbal behaviors showed that patient-perceived empathy was affected by nonverbal communication (e.g., eye contact), not by physician white coat attire [17]. There is also evidence that, compared with male participants, female participants perceive doctors who express brusque nonverbal behavior as having low empathy. Empathy is a complex, multidimensional phenomenon that includes several functional processes, such as emotion recognition, emotional contagion, and emotion priming [20]. Empathy is also context-sensitive in patient–physician relationships [21]. Japan has a very unique culture that relies heavily on nonverbal and implicit communication [1, 22, 23]; thus, physician attire may play a more important role in patient–physician relationships in Japan than in other countries. In Japan, most patients prefer physicians to wear a white coat because it is considered professional or hygienic [24]. Conversely, white coat attire, with its connotations of professionalism, can be a symbol of a doctor’s paternalism, which may negatively influence the “human” aspects of medical care [25, 26]. It remains to be established whether FP dress style is associated with the perception of empathy in patient–physician relationships in Japan.

In this study, we investigated the use of alternating dress styles (casual attire vs. white coats) in FP practice to compare patient-perceived empathy, assessed using the Consultation and Relational Empathy (CARE) Measure. In addition, we tested previous findings [17] of a difference in perceived empathy between male and female patients.

This is the first multicenter, quasi-randomized controlled trial to examine the effect of physician attire (a white coat or casual attire) on patient-perceived relational empathy. We found no difference in empathy between the White coat and Casual attire conditions overall. However, the wearing of a white coat during FP consultations had a significant negative effect on male patients’ perceived empathy.

Previous studies in Japan [1, 24] have shown that most patients prefer Japanese FPs to wear white coats. It is likely that patients whose expectations are met in terms of their physician’s attire will experience more empathy in the therapeutic relationship. However, we found no difference in patient-perceived relational empathy between the White coat and Casual attire conditions. There are several possible patient-related reasons for this result. First, previous research indicates that FP clothing is a nonverbal cue that is perceived less frequently by patients compared with tone of voice, eye contact, and facial expressions [33]. Therefore, FPs’ choice of dress did not contribute substantially to empathy as perceived by patients. Second, a previous study [34] found that more tenseness was reported by new patients in a White coat group than in a Casual group, which suggests that the use of non-white coat attire in patient consultations may help to establish smoother patient–physician relationships. Third, modern patients have become more accustomed to physicians not wearing white coats, as increasing numbers of doctors do not wear white coats owing to concerns about contamination [9, 35‐37]. From the physician’s perspective, a white coat confers professional identity at the expense of personhood, and so is not necessarily empathetic [35]. Our results differed from previous research in Korea which showed that patients’ perception of empathy was substantially higher when a traditional Korean medicine doctor wore a white coat or traditional attire than when they wore casual attire or suits [19]. Patient-perceived empathy may differ according to cultural differences and type of medical professional.

We also found that male patients were significantly more affected than female patients by perceived physician empathy when their physicians wore casual attire. There was a 2-point difference in the CARE Measure score, which is greater than the difference observed in previous studies with and without facemasks [28]. In a previous study that investigated gender differences in an emotion attribution task using functional magnetic resonance imaging, women and men relied on different strategies when assessing their own emotions in response to other people [38]. Previous research using the CARE Measure has also shown that female patients are more attuned than male patients to empathy signals such as lack of eye contact and unequal eye-levels [17]. Women are generally more sensitive than men to empathy and the feelings of others [21, 39]. Women are faster and more accurate at recognizing facial expressions than men [20], better at recognizing emotions, and express themselves more easily [39]. Female patients may be affected by features that are more salient than physician attire, such as tone of voice, eye contact, and facial expression [33]. However, the empathetic responses of male patients tend to be more influenced by contextual cues than those of female patients [20]. Men are also more responsive to threatening cues (dominant, violent, or aggressive cues) [39]. A white coat may be perceived as indicating medical paternalism [40], and so may affect the perceived empathy of male patients more than that of females. Although intriguing, further research is needed to explore such differences between male and female patients, as this was a secondary analysis in the present study.

This is the first multicenter, prospective controlled trial in primary care clinics to explore the differential effect of wearing a white coat or casual attire on empathy. One strength of the present study is that, to reduce information bias (and with the permission of our ethics committee), we explained to patients that the research was about empathy, but did not reveal that we were investigating the effect of physician attire. Our study has several limitations. First, for pragmatic reasons, patients were allocated on a weekly basis and there was no randomization. Second, the study design meant that the study was non-blind. Third, we did not reach the target sample size because we had to terminate the study early owing to the COVID-19 pandemic. This makes it difficult to draw firm conclusions from the findings. Fourth, a previous study identified a weak positive association between CARE score and consultation length, satisfaction with consultation length, and how well the patient knew the doctor [30]. We did not evaluate consultation length and satisfaction with the length, so we could not adjust the results. However, as our study targeted new patients, it was unlikely that the findings were affected by how well the patient knew the doctor. Fifth, the FPs in this study may not necessarily be representative of all Japanese FPs. Certification of FPs is changing in Japan. The JPCA began to certify FPs as “JPCA-certified family physicians” in 2010 [41] and the number of JPCA-certified FPs was only 900 as of September 30, 2020 [42]. From 2018, the certification changed to be a specialty based on the acquisition of general practitioner board certification [43, 44]. Therefore, most physicians currently working as FPs are not well-trained certified FPs and do not necessarily follow the global standard of primary care physicians [45]. For these reasons, statistical data for physicians working as FPs are not available. However, the participants of this study currently work as FPs, and we believe that they are fairly representative of FPs in Japan. Of the seven FP participants, six work in private clinics; this is close to the national situation, as more than 95% of medical clinics in Japan are private clinics [46]. Sixth, we did not regulate clothing worn under the white coat. This may have been a confounding variable, because a patient’s impression of a physician changes according to what the physician wears under his or her white coat [1]. Seventh, the effect of physician sex was not assessed because only one female physician participated in this study (below our target number for females). As perceived changes in facial expression are affected by the gender of both the source and recipient [47], more research is needed on this topic.

The present results suggest that physicians should be advised that wearing a white coat or casual attire does not have an overall effect on the establishment of patient–physician relationships in Japan, but that casual attire may have a positive effect on male patients. Additionally, white coat attire is associated with several problems, such as white coat hypertension [48, 49] and bacterial dissemination [36, 37]. Given these findings, it is perhaps time for physicians to consider alternatives to white coat attire.

We found no difference overall in the effect of white coats and casual attire on patient-perceived relational empathy in primary care consultations, but male patients of physicians who wore casual attire reported higher physician empathy. Empathy is a complex, multidimensional phenomenon that is context-sensitive in patient–physician relationships and cannot be reduced to simple variables such as attire. White coats may have a negative effect on patients, depending on the context. FPs should choose their attire carefully.