Empathy in Healthcare Profession

Empathy is no doubt one of the most important central values for healthcare professionals (Bardes, 2006; Bishop & Rees, 2007; Norman, 2007; Shapiro, 2008). It can not only help establish a better patient-physician relationship, but also create the broader connections with the community. This good relationship would improve patients’ adherence, treatment compliance and enhance satisfaction with medical services (Burks & Kobus, 2012; Hall, Epstein, DeCiantis, & McNeil, 1993; M Robin DiMatteo et al., 1993; McGaghie et al., 2002; Ong, Dehae, Hoos, & Lammes, 1995).

Empathy may also help encourage patients to be more willing to report their symptoms cal proffffffffesesesesesessisssisisisisisisiononononononononalalaalalalalalals ssssss to the ototootototooheheheheheheher er r rr r rrasasasasasasaspespepepepepepepep ctcctctctcctct oooooooff ffffff me baaaaaaaadd dd d d d dd imimimimimimimimpapapapapapapapactctctctctctctct ooon hman & Orlander

in details, which would increase the accuracy of diagnosis (Chen et al., 2007;

Hemmerdinger et al., 2007; Neumann et al., 2011). Furthermore, a better communication between patients and medical professionals could even reduce the severity of symptoms and decrease blood pressure, blood glucose levels or anxiety levels of patients (Griffin et al., 2004; Neumann et al., 2007; M. A. Stewart, 1995; M.

Stewart et al., 2000).

Compared to physicians, therapists (e.g., clinical psychologists, physical therapists, occupational therapists, speech therapists) usually have longer and more frequent contact with patients or clients. For example, stroke patients in average spend half to one hour per day in their acute stage and at least one hour per week in their sub-acute or chronic stage in a rehab clinic with occupational therapist. In addition, empathy has been an essential core for healthcare professionals to apply ‘client-centered’ approaches in daily practices (Brown et al., 2010; Dalia Sachs, 1994; McAuliffe & Barnett, 2009;

McKenna, Adri-Anne Scholtes, Fleming, & Gilbert, 2001; Norman, 2007; Stachura &

Garven, 2007).

Although empathy is an important focus in the education of healthcare professions, several studies show that the levels of empathy of medical students declined as their medical education years increases. The declination is even more prominent when they are in their senior grades and start their clinical training (Bardes, 2006; Chen et al., 2007;

Neumann et al., 2011; Williams et al., 2012). Additional evidence also shows that medical students may not have more empathy than other students, and their empathy even decays when they contact clinical patients (Au & Man, 2006; Brown et al., 2010;

McAuliffe & Barnett, 2009; Stachura & Garven, 2007). Surprisingly, one cross-campus study even reported that business students show greater altruistic attitudes than medical and law students, suggesting that medical students do not necessarily

hen et t tt t tt alalalalalalalal.,,, 2 2222222200000000000 7;

hermmmmmmmoororooroo e,e,ee,e,e,e,e, a a a a a a ba b b bebbbbeteteeetetetteeeeeeerr rrrrr d even nnnnnnn rerererererereredudududududududucececececececece t t thettttthhhhhhh

levels or anxiety

exhibit more altruistic behavior than other students (Coulter et al., 2007).

In contrast, the declination trend of empathy in students of occupational therapy does not appear to be significant (Brown et al., 2010). The results, however, remain inconclusive in that the survey used in the study was designed for physicians (Jefferson Scale of Physician Empathy (JSPE)) and may not be suitable for evaluating empathy traits of occupational therapy students (Brown et al., 2010). Data from senior students, who were in their clinical training year, were also missing, which minimize the interpretation power of the study. Furthermore, like most studies, they used self-report assessments, which lack reliability and validity and could show social expectation bias effect (Fields et al., 2011; Hemmerdinger et al., 2007). . Previous research also suggested that these questionnaires have not enough sensitivity to detect the difference between individuals and could only be used as a screening tool to identify the low empathy one (Hemmerdinger et al., 2007). Hence, it is essential to use an indirect approach to assess students’ empathy with reduced interference from social expectation bias effect.

Several reasons could explain the declination of empathy observed in medical students: distress from heavy workload, sleep deprivation, poor interaction with patients and feeling of helplessness with difficult medical situations (Chen et al., 2007;

Maudsley, Williams, & Taylor, 2010; Neumann et al., 2011). Surprisingly, despite health caregivers trained to be promote health , there were high rates of psychiatric morbidity and stress within medical professions (Firth-Cozens & Payne, 1999; Wall et al., 1997). Further, when medical students faced with clinical reality such as painful illness and death, they might decrease their feelings toward patients to avoid emotional distress, protect themselves and tried to be more rational on their duties (Chen et al., 2007; Hojat et al., 2004). A recent ERP study investigated physicians’ response to

07).

cupatataatatatatatioioioioiooonanananananananall ll llll ththththththththhererererererererapapapapapapppyy yyyyy , howwwwwwwweveveveveveveveve erererererererer, rerererererereremamammamammamain ysicians (J( efferson

pictures representing body parts in contact with painful or non-painful objects, and demonstrated that, compared to the control group, physicians seemed to suppress their emotional reaction through early affective processing when they saw others in pain (Decety, Yang, & Cheng, 2010). Therefore, the observed common declination of empathy in physicians could be an emotional regulation mechanism to keep them away from work-related distress, burnout or fatigue and help them make rational clinical decisions. It is possible that empathy decline is a necessary and normal process to adjust to the clinical reality and medical surroundings when medical professionals pass through medical practice training. According to these findings, there seems to be some discrepancies between empathy regulation and helping behavior in healthcare professionals.

1.5 Aims of Current Research

Although several factors have been assumed to explain the declination of empathy in healthcare professionals, very few studies can provide the evidence that supports these assumptions. In terms of empathy evaluation, previous research also suffered from robust vulnerability to response bias due to the nature of explicit measurement by empathy questionnaires (Fields et al., 2011; Hemmerdinger et al., 2007). Furthermore, most of previous research focused only on the empathy declination, the relationship between empathy level and prosocial behavior patterns in healthcare professionals remains unclear. Therefore, there are two aims for the current research:

1. To develop experimental paradigms to implicitly evaluate both the level of empathy and prosocial behavior patterns.

2. To investigate the chronological change of the relationship between empathy levels and prosocial behavior patterns in healthcare professionals.

ainful obobobobobobobobjejejejejejejejectctctctctctctcts,s,s,s,s,s,s,s, a a aaaaaandd d tooooooo s s ssupupupupupupupprpprprprprprprpresesesesesesess ss s sss ss s thtththththththeeeiriririririirir

aw ototottttttheheheheheheheheh rsrsrsrsrsrsrsrs in iiiiiiin n nnnn papapapapapapapain mon decliination of

2. General Methods

2.1 Apparatus

A Win XP PC with Matlab installed and a 21” CRT monitor of up to 100 Hz were needed to present experimental paradigms and collecting behavioral data. All experiment scripts were written in Matlab using the Psychophysics Toolbox extensions (Brainard, 1997; Kleiner, Brainard, & Pelli, 2007; Pelli, 1997)

2.2 Participants

Table 1

Demographic Data

Gender portion (Female/ male)

Age (y/o)

Formal education year

Professional training year

Seniority

Professional

controls 19 / 11 27.27± 3.39 16.83± 1.80 4.60± 1.22 3.15± 1.97

Occupational

therapists 20 / 10 27.63± 4.30 16.23± .553 4.33±.71 4.37± 3.27

t - -0.367 1.74 1.03 -1.75

p - 0.715 0.087 0.305 0.086

Occupational

freshmen 22 / 8 18.73± .79 12.03± .183 0 0

Freshmen

controls 23 / 7 18.47± .51 12.00± .000 0 0

t - 1.56 1 -

-p - 0.124 0.321 -

-We recruited 120 healthy volunteers to participate the current study. Since it is important to conceal the true purpose of the present experiments before participants finish all tasks, all participants signed one informed consent before the start of experiments and signed another form which revealed true experimental aims after they completed the experiments. The experimental protocol was approved by Research Ethic Committee at National Taiwan University (201209HS004).

All participants were categorized as one of the following four groups with 30 participants for each group:

Naive Occupational Therapy students: 30 freshmen in the school of occupational therapy at National Taiwan University. (22 female, mean age ± S.D. = 18.7 ± 0.8 y/o).

Naive control students: 30 freshmen who do not major in healthcare or social work related undergraduate program at National Taiwan University (23 female, mean age ± S.D. = 18.5 ± 0.5).

Senior Occupational Therapists: 30 registered occupational therapists who have worked full-time for more than 2 years in Taiwan (19 female, mean age ± S.D. = 27.3 ± 3.4).

Senior professional controls: 30 adult participants who have a full-time job in other professions, excluding anything related to health care or social work, for more than 2 years in Taiwan (20 female, mean age ± S.D. = 27.6 ± 4.2).

Since the proportion of male occupational therapists and occupational students generally lies between 1/3 to 1/4, we tried to recruit similar gender proportion of participants for each group (Sylvia R Cruess & Cruess, 1997). The demographic information of all participants was listed in Table 1.

studydydydydydydydy. .. S S S S SinSSSininininininncecececececececec i i i i i i it tttt isisisisisiisis before eeeeee papapapapapapapartrtrtrtrtrtrtrticicicicicicicicipipipipipipipanpanaaanananants efore the start of

2.3 Experiment 1: Implicit Evaluation for Levels of Empathy Using Emotional Stroop Paradigm

The levels of empathy was traditionally evaluated by administering psychometric questionnaires such as Jefferson scale of physician empathy, Empathy quotient, Davis' interpersonal reactivity index, Empathy construct rating scale. However, it was not rare to observe response biases favoring empathetic decisions resulting from social expectation effect (i.e., people tend to respond in terms of what a society expect an empathetic person would do) (Neumann et al., 2011). To downgrade the interference from social expectation effect, we used a novel version of emotional Stroop paradigm to implicitly evaluate participants’ level of empathy. The traditional Stroop paradigm required participants to selectively respond to a given feature (e.g., colors) of stimuli (e.g., colored words of “colors”) while suppressing the distraction from another feature of the same stimuli (e.g., the semantic meaning of words) (Stroop, 1935). The Stroop effect occurs as participants respond with longer reaction times (RT) and prone to report more errors when the two features of a stimulus are incongruent (e.g., a “Red” written in green color) than when the two features are congruent (e.g., a “Red” written in red color). The size of the effect, usually measured as the RT difference between congruent vs. incongruent trials, represents the degrees of interference from distracting features.

The Emotional Stroop, modified from the traditional Stroop paradigm, combines emotional words and emotional facial expression pictures to create an emotional interference (see Figure 1) (Haas, Omura, Constable, & Canli, 2006; Hofelich &

Preston, 2012; Hu, Liu, Weng, & Northoff, 2012; Monk et al., 2003; Preston &

Hofelich, 2012; Preston & Stansfield, 2008) and the Emotional Stroop effect occurs

r

when the valence of the words is incongruent with the valence of the facial expression.

The intensity of the emotional interference is therefore measured by the RT differences between congruent and incongruent trials. The paradigm appears to be an efficient behavioral probe for implicitly evaluating emotional responses. For example, patients with major depressive disorder suffered greater interference from negative words than positive words in Emotional Stroop (Hu et al., 2012).

In the current experiment, we aimed to use the Emotional Stroop paradigm to implicitly evaluate participants’ empathizing capability. Specifically, we compared the Emotional Stroop Effect under three different contexts of empathy: Empathy Induction (see Figure 2), Cognitive Interference (see Figure 3), and Neutral Viewing conditions (see Figure 4). In the Empathy Induction (EI) condition, participants were required to look at a picture and made efforts to empathize the emotional contents in the picture before performing the Emotional Stroop tasks (see Figure 2). If a participant possesses a higher level of empathy, the picture will presumably induce higher emotional valence and subsequently amplify the emotional interference effects (i.e., larger RT differences between incongruent trials vs. congruent trials). In order to further investigate if there is any difference of modulation effects between medical and non-medical empathy contexts, the emotional content pictures were separated into non-medical contexts and medical contexts during analyses. The Cognitive Interference condition (CI) served as an empathy baseline condition in which we aimed to suppress top-down empathizing processing by introducing an additional working memory task (see Figure 3). Accordingly, the Emotional Stroop effect contrast between the EI and CI presumably reflects the modulation of empathy upon Emotional Stroop effects. It is possible that the Emotional Stroop effect in both EI and CI conditions, if any, could result from some bottom-up emotional responses by simply e facial l l l ll exexexexexexexxprprprprprprprpresesesesesesesessissisisissisision.

the RRRRRRRRT T T T T T Tdidididididididiffffffffffffffffererererererererrenenenenenenenenceeeeeees ssssss to be eeeee anananananananan e e e eeeeefffffffffffffffficicicicicicicicieieieieieieieient r example patients

viewing emotional contents. Therefore, the Neutral Viewing (NU) condition further served as a general emotion baseline measurement in which emotional pictures were replaced by non-meaningful pixel-shuffled pictures from the same set of pictures used in the EI condition (see Figure 4).

Figure 1

Three stimulus examples in the Emotional Stroop Experiment.

Here we show three examples of emotional face-word pairs for the Emotional Stroop paradigm. Picture (a) represents a typical emotionally congruent trial (Con) in which the facial expression (happy) matches the double-character descriptions (happy). In contrast, picture (b) and (c) show two typical emotionally incongruent trial in which the facial expression does not match the double-character descriptions. Picture (b) shows happy-word-sad-face (PWSF) incongruent trial, and picture (c) shows sad-word-happy-face (NWHF) incongruent trial.

) condididididididititititititiionononononononon f f f f f furfffurururururuururthttttttt er onal l ll l l ll pipipipipipipictctctctctctctcturuururururururesesesesesesesess w w w w w w w werrrrrrre eeeeee et of pipipipipipipipiicctctctctctctcturururururururureseseseseseseses u u u uuuuussssessss d

Figure 2

The sequence of a typical trial for the Empathy Induction Condition.

Empathy Induction condition (EI): In this condition (Figure 2), participants were required to empathize the person’s situation of the emotional context from a given picture. Specifically, participants were instructed to imagine what they feel if they undergo the same emotional incident and also imagine how the incident may influence their life in the future. There were four emotional context categories with six pictures in each context. Each context category (i.e., an experiment block) started with a 4 sec context description followed by a 2 sec task instruction. Subsequently, 6 trials of emotional Stroop were presented. Each trial started with a 4 sec emotional context picture followed by a picture containing a pair of double-character word and an emotional facial expression with a 1 sec latency. The next trial would not start until participants made a response.

onnnnnnnnnn......

Figure 3

The sequence of a typical trial for the Cognitive Interference Condition.

Cognitive Interference condition (CI): In this condition (Figure 3), a working memory task was introduced to presumably prevent participants’ automatic empathizing processes, if any, induced by emotional context pictures. The working memory task required participants to remember an 8 digit number presented before the emotional context picture. Following the picture presentation, a pair of memory test numbers was presented and participants needed to indicate which number matches the content of their working memory. Note that the distracting number and the correct number differed from each other with only one digit, maintaining a reasonable load for working memory. After the memory test, participants performed the regular Emotional Stroop task.

di di di d d d d d d di d d di dtititttttitooonononoononooooooon.. .

Figure 4

The sequence of a typical trial for the Neutral Viewing Condition.

Neutral Viewing Condition (NU): In this condition (see Figure 4), participants were required to look at neutral pictures naturally without any extra effort as if they were sitting on the sofa, reading a magazine and coming across to these images (Rameson, Morelli, & Lieberman, 2011). For each trial, a pixel-shuffled picture was presented for 4 sec and was followed by a picture containing a pair of double-character word and an emotional facial expression with a 1 sec latency. The next trial would not start until participants made a response.

.

2.3.1 Procedures

2.3.1.1 Stimuli and Materials

The emotional context pictures were selected from the International Affective Picture System (IAPS) (Van Lange, 2008) and the picture repertoire used in a previous study (Rameson et al., 2011). All the pictures were subdivided into 16 sets according to their contexts and reevaluated subjective empathy valance, arousal level and emotional valance in our study. The facial expression pictures were selected from 12 (6 men and 6 female) actors in Taiwan emotional standard stimuli database (Farnand et al., 2009) . One happy face and one sad face expression were chosen form each actor.

The 12 double-character emotional words (៿኷ǵඍ৹ǵࢲΚǵݒ഻ǵ៿഻ǵ౏זǵ ൿኌǵኁ໾ǵݪ഼ǵൿࠉǵधඊǵ໾Ј) were drawn from the emotional standard Chinese phrase database developed by Lee et al (Lee & Lee, 2011). There were no significant differences between arousal level, concreteness, familiarity, frequency and stoke count in chosen words.

2.3.1.2 Tasks

Emotional Stroop

A typical trial of Emotional Stroop started with an emotional (or neutral) context picture presented for 4 sec. One second after the offset of the emotional context picture, a picture of a face paired with a double-character word was presented and last until participants made responses. The two characters were displayed at the cheek level so that they would not cover the eyes of faces. The face could be either a happy or a sad facial expression, and the double-character word could describe either a positive or a negative emotion. The combination of face-word pairs was

natioionan l Affeectivve

pseudorandomly presented in three different kinds of trial types: congruent (Con) (happy face with positive words or sad face with negative words), negative words with happy face incongruent (NWHF) and positive words with sad face incongruent (PWSF) trial types (see Figure 1) (Hofelich & Preston, 2012; Preston & Stansfield, 2008; Rameson et al., 2011). Participants were instructed to press one key for happy face and another key for sad face as soon and accurate as possible. After the response, there would be a feedback signal (500 ms) with a green square around the face picture for correct responses and a red square for incorrect responses. The Emotional Stroop effect was defined as longer reaction times for the incongruent trials than the congruent trials.

In the current experiment, participants performed emotional Stroop task under 5 different empathy contexts including neutral viewing condition, medical cognitive interference condition, medical empathy induction condition, non-medical cognitive interference condition and non-medical empathy induction condition. There were four blocks for each empathy context with each block contained six pictures belonging to a single emotional context. To prevent the emotional Stroop task of neutral viewing condition from interference by other EI and CI conditions, the neutral viewing condition was presented first and then the other 4 conditions would show in random order. For

In the current experiment, participants performed emotional Stroop task under 5 different empathy contexts including neutral viewing condition, medical cognitive interference condition, medical empathy induction condition, non-medical cognitive interference condition and non-medical empathy induction condition. There were four blocks for each empathy context with each block contained six pictures belonging to a single emotional context. To prevent the emotional Stroop task of neutral viewing condition from interference by other EI and CI conditions, the neutral viewing condition was presented first and then the other 4 conditions would show in random order. For