Pain is a basic and necessary experience that alerts us to physical harm or infection. However, pain is notoriously difficult to describe, and it is harder for one person to understand what another person’s pain is like. The International Association for the Study of Pain defined chronic pain as pain that persists for more than 3 months [1], whereas acute pain refers to pain that persists for a shorter time and is expected to subside with healing (eg, pain from surgery or childbirth) [2]. Although most people have experienced acute pain, chronic pain persists well beyond the expected time of healing and can last a lifetime [3]. Although chronic pain is now considered to be a condition [1], its cause and cure have yet to be discovered, and it has no clear biomarkers. In addition, chronic pain often has associated sequelae, from anxiety, depression, insomnia, and cognitive impairment to decreasing mobility [3].

Conservative estimates suggest that 1 in 5 people of all ages live with chronic pain [1,3]. Despite this high prevalence, people who live with chronic pain also face social stigma and high rates of social isolation [4]. More than loneliness, social isolation is correlated with a decreasing quality of life and earlier morbidity [3,5]. As treating this long-term pain means managing it, it is important that caregivers and health professionals find ways to understand the lived experience of chronic pain sufferers—in other words, what it is like to actually live with the debilitating effects of long-term pain and how it impacts a patient’s biopsychosocial realities, their ability to function, and their quality of life [3].

Definitions of empathy are wide ranging: the most common is “our ability to perceive, understand, and respond to the experiences and behavior of others” [6]. According to Davis [7], empathy has 3 primary dimensions: (1) physical sensations, (2) emotional sensations, and (3) cognitive awareness (perspective taking). In many studies from the medical and cognitive science domains, the emotional and perspective taking dimensions of empathy were the predominant dimensions that were evaluated, along with their effects in both clinical and nonclinical environments [8-13].

However, despite renewed attention on empathy, some studies found that teaching empathy has declined in undergraduate education both in medical schools [8,14,15] and colleges in general [16]. Researchers believe that we can mitigate this situation by teaching empathy in medical schools by embedding it in students’ actual experiences with patients [17], which can lead to better diagnoses [17-19]. Consequently, empathic communication may help patients manage their health by fostering trust between them and their health professions [9]. Hence, empathy training may similarly improve the attitudes of nonpatients toward patients with chronic pain. (Here, we define nonpatients in terms of 2 populations: (1) doctors, nurses, and health professionals who work with patients with chronic pain and (2) caregivers and family members of chronic pain sufferers.)

Previous studies demonstrated that games and video games have a high potential for fostering empathy toward a certain population, even if the players are concerned about winning the game [20-24]. The sense of embodiment in virtual reality (VR) games can appear significantly more effective compared with watching a 2D video [12,25,26]. Evidence also showed that VR games and environments can facilitate a significantly higher level of empathy than videos or traditional media forms [27]. The immersive and convincing nature of VR has profound effects and may confer meaningful benefits for an individual’s cognition or behavior [28]. For instance, Bailenson et al [27] found that subjects showed a higher occurrence of thinking from the place of their partners in a VR perspective taking experience. They also observed an improvement in participants’ empathy levels toward their partners when they considered the patients’ perspective [27]. In other words, two experiential aspects of VR—an immersive sense and an embodied sense—appear to play important roles in empathy [27,29,30].

To develop or strengthen a more empathic connection between nonpatients and patients with chronic pain, we designed an interactive VR game entitled AS IF [31]. In this VR game, participants inhabit an avatar—a 3D character who has chronic pain—from a first-person perspective. Participants were then asked to perform a series of activities and tasks as patients with chronic pain. In other words, AS IF was developed to enable participants to not only inhabit the virtual body of a patient with chronic pain but also to metaphorically walk in the shoes of this patient by experiencing physical limitations during tasks and to hear that patient’s self-talk during each task. We designed AS IF with a view to eventually deploying it for clinicians and caregivers. Overall, the goals of this study were to (1) increase empathy toward the growing number of people who live with chronic pain, (2) evaluate the prototype, and (3) offer design suggestions that may guide future research.

Initially, a non-VR desktop version of AS IF was developed and tested by the authors using Microsoft Kinect [31]. Later, we conducted a study to evaluate the effectiveness of this interactive game system [32]. In that version [31,32], the game tasks involved completing connect-the-dots puzzles from a third-person perspective. The player inhabited and could move the limbs and head of a virtual avatar—a grandmother living with chronic pain. Players listened to the grandmother’s self-talk while solving puzzles; each solved puzzle resulted in the completion of a kitchen task. From analyzing the feedback collected in playtesting, we found that participants were more willing to help people with chronic pain and exhibited positive attitudes about playing from the first-person perspective. However, participants wanted to interact with virtual objects more directly, rather than solving the more abstract connect-the-dots puzzles, and some suggested we design a comparable game in VR for that reason.

Therefore, we redesigned the game and switched from Microsoft Kinect’s platform to an immersive VR platform, the HTC VIVE [33]. In addition, we made significant changes to the game using the Unity3D [34] game engine. Here, we outline the revision of AS IF.

In this VR version, from the first-person perspective, the player inhabits an avatar of a grandmother who lives with chronic pain. Participants attended to tasks that required physical movements and, as the grandmother, experienced limitations of those movements as she would, a point that was made in her self-talk or audible narration. The embodiment aspects of the VR game feature multimodal feedback; design decisions such as this were based on the theory of embodied simulation [35].

Altogether, 18 participants (4 females), aged 19-39 years (mean 24.8, SD 3.8 years), were recruited through convenience sampling method for this study. We placed advertisements in the university campus media and sent emails to the faculty and student groups. The inclusion criterion was anyone older than 19 years, and the exclusion criteria were anyone who had a pain condition or was a pain patient or did not understand English. No participants had a reported history of a chronic pain diagnosis; this was an essential requirement because personal experience of chronic pain could bias expectations of AS IF. Nevertheless, 3 participants reported that they had contact with patients with chronic pain.

The goal of this research study was to determine if a serious VR game such as AS IF may influence participants’ empathy for patients with chronic pain (raising awareness and fostering positive attitudes toward patients) and to explore what factors may be important to elicit empathy. Furthermore, based on participants’ feedback on this version of our design, we summarize the fundamental design principles and game components to help guide future works.

Participants experienced and could interact with objects in the virtual environment via a wired, stereoscopic HTC VIVE head-mounted display (HMD) and its handheld controllers. We developed the game using the Unity3D game engine, which was responsible for rendering and running the game during the study. These real-time rendered scenes of AS IF were sent to the HMD through SteamVR (HTC VIVE) suites; the software was responsible for data between Unity3D and the devices via an API called OpenVR.

A mixed-method design approach was adopted in this pre-test, post-test study. Participants’ empathy levels towards chronic pain patients before and after playing AS-IF, and quantitative questionnaires and qualitative interviews were used in parallel to derive the findings [37]. The study lasted for 35 to 40 min in total. On arrival, participants were briefly introduced to the purpose of the study and the entire procedure and were then asked to read and sign the informed consent form. Next, participants were asked to fill out the preintervention questionnaire, which included the Empathy Scale (revised from the Compassion Scale [38]), the Willingness to Help Scale, the VR-adapted Other in the Self Scale, and the Emotional Wheel evaluation (Multimedia Appendix 1). During the intervention, participants were first shown how to play AS IF, and then, they played it for approximately 10-15 min.

When they finished the game, participants were asked to fill out the posttest questionnaires to assess their level of empathy toward patients with chronic pain. The posttest questionnaires were a repeat of the Empathy Scale, the Willingness to Help Scale, and the Emotional Wheel evaluation. The Other in the Self Scale was added to assess the relationship between self and the first-person perspective in the VR experience (Multimedia Appendix 1). In addition, participants were given a Sense of Embodiment questionnaire to evaluate their perceived level of immersion in the VR game. Finally, through a 15-min semistructured interview, participants discussed their experience, provided feedback, and offered researchers their ideas about the game. The interview topics were primarily about the game’s interactions, depictions of pain, and physical impact on participants’ empathetic attitudes in AS IF. Meanwhile, the session was audio recorded to ensure that the captured data were accurate.

We used multiple instruments to measure various aspects of empathy. For instance, the Empathy Scale (adapted from Pommier Compassion Scale [38,39]) measures multiple-dimensioned implicit cognitive changes, whereas the Willingness to Help Scale detects explicit cognitive changes. Wheels of Emotion reflects emotional changes, and the VR-adapted Other in the Self Scale assesses the perspective taking aspect of empathy. As chronic pain is a complex process that involves physical, emotional, and social aspects, we adopted instruments that account for cognitive and emotional perspectives regarding pain (refer to Multimedia Appendix 1 for more details on each scale).

To determine which instruments were validated and the most appropriate, we compared existing instruments, such as the Basic Empathy Scale [40], Toronto Empathy Questionnaires [41], and Baron-Cohen and Wheelwright’s Empathy Quotient for people with autism [42]. After thoroughly examining questions from each scale and evaluating how the scales may fit and be adapted for our purposes and population (patients with chronic pain), we selected the Compassion Scale [38,39] for its appropriate number and types of questions.

In our previous study of the initial version of AS IF, the Empathy Scale, the five-point Numerical Rating Scale (NRS), and the Willingness to Help Scale (10-point NRS) were used to understand if and how that game may have affected participants’ cognitive empathy levels. A total of 2 pretest questionnaires were also used in the previous version: (1) the Empathy Scale was used to assess implicit empathy (unconscious emotion) and (2) the Willingness to Help Scale was used to assess explicit empathy (conscious emotion).

For the new VR version of AS IF, the Wheel of Emotion [43] and the VR-adapted Other in the Self Scale [44] (Figure 4) were adopted to measure any emotional changes associated with empathy and the degree of perspective taking according to the suggestions of Carey et al [45]. The Wheels of Emotion provides a comprehensive list of emotions and standardizes the data collected (according to Plutchik’s Psycho-Evolutionary Theory of Emotion), whereas the Other in the Self Scale helps players articulate how they related to the grandmother’s avatar. Moreover, to further investigate embodiment in the VR version, we adapted a sense of ownership (SoO) and a sense of agency (SoA), rated on an 11-point NRS (from −5 to 5). The adapted questions are listed in Table 1.

All statistical analyses of the quantitative data were performed using SPSS version 22 (IBM) software [46]. For qualitative data, in-person interviews were first transcribed as electronic textual data. These data were then coded into categories by 2 authors based on preexisting knowledge or hypotheses. After comparing the results, the researchers highlighted significant patterns and summarized and grouped them into themes.

Here, we discuss what the participants thought of AS IF and evaluate the strengths and weaknesses of the VR game. We also discuss which of the game’s main features may be useful for future research.

We explored a significant redesign and study results of a VR game, the AS IF. It is aimed at motivating people who do not live with chronic pain (nonpatients) to better understand the lived experience of chronic pain by increasing empathy. In general, the findings demonstrate that participants had greater degrees of empathy toward patients after playing the VR game. Furthermore, from the semistructured interviews, we were able to gather essential feedback about the strengths and limitations of the current VR design, such as the effectiveness of pain representations. Finally, we extracted critical design issues, implications, and protocol suggestions and offered them to potentially benefit similar research in the future.

Overall, after playing the VR game AS IF, participants scored significantly higher on the Willingness to Help Scale and the kindness subscale—an adaptation of the Empathy questionnaire. These two scores revealed that not only could one VR experience of AS IF raise people’s awareness of chronic pain but it could also increase their implicit and explicit empathy. Data from the kindness subscale showed implicit cognitive changes, whereas data from the Willingness to Help Scale revealed changes in explicit empathic attitudes toward patients with chronic pain. The other three subscales showed nonsignificant differences. Furthermore, the qualitative interview data show that most participants reported that playing this game helped them to understand what a chronic pain patient’s life would be like, and that they had never thought about that before. We assume these findings may result from 2 potential reasons: (1) indifference, disengagement, and separation are difficult to affect or change during a single, short period, as in this study, and (2) the design of the VR game AS IF focused more on the perspective taking and emotional aspects of empathy, but it did not have specific game features that were meant to increase the four subscales.

The VR-adapted Other in the Self Scale suggested that most participants felt that the virtual body overlapped with their real identity. The findings from the VR-adapted Other in the Self Scale also overlapped with the interview results. Some participants said they felt embodied in the grandma avatar who has chronic pain through narrative storytelling, the immersive environment, and the game tasks. Thus, in AS IF game, participants were able to understand the perspective of the grandma who has chronic pain.

The sense of embodiment scores showed that, on average, participants could sense owning and controlling the virtual avatar. However, the Pearson correlation test revealed no statistical significance between the sense of embodiment in VR (comprising SoO and SoA) and the Empathy Scale (posttest) or the sense of embodiment in VR and the Willingness to Help scale. We conjecture that 2 reasons might account for this nonsignificant outcome. First, there could be multiple factors that affected empathy levels besides embodiment, such as the narrative and the game’s specific tasks (and the fun/frustration behind that). Hence, a single factor might not be strong enough to alter overall feelings of empathy. As mentioned in the interviews, participants suggested that tactile feedback might be a better way to indicate pain or the association of visual effects with pain. The second possible explanation could be that the game did not provide a strong enough sense of embodiment to reach statistical significance. In interviews, participants said they wanted the virtual avatar to more closely match their own gender and ethnicity and perhaps even body height and shape. A few participants—a male and a participant whose skin color differs from the avatar’s—reported they felt disembodied with the virtual avatar because of its divergent characteristics.

Admittedly, the overall empathy scores and the 3 subscales (indifference, disengagement, and separation) did not change significantly. A crucial issue is how long it takes to affect and change empathy and what factors are important in facilitating such change. For example, implicit empathy may be difficult to change in a short time, in part because of its mental cost [47]. However, in this study, the VR-adapted Other in the Self Scale findings suggest that participants identified with the VR avatar, insofar as the avatar was felt to overlap with their real self. Therefore, a perspective taking ability may be critical to being able to influence one’s empathetic attitudes toward patients with chronic pain and painful experiences. From the qualitative interviews, 2 approaches to representing pain in the virtual body may also facilitate empathy. Although a few participants found the movement restrictions confusing, most reported it made them realize how pain would impact one’s range of motion and emotion. Most liked the idea of using the red flashes to represent pain spikes and reported that the visual effect felt like a headache or pain (a synesthesia effect of transferring a visual sense to an emotional sense).

Besides facilitating nonpatients’ (game players who are health care givers or family members of patients with chronic pain) empathy toward patients with chronic pain, we believe that a VR game such as AS IF has the potential to benefit patients and other researchers. Chronic pain experiences are notoriously difficult to describe and are often out of the experiential scope for most people. Hence, VR is one method that may provide clinicians and family members or friends who play a caregiving role with a deeper understanding of a patient’s chronic pain condition. VR appears to have the potential to change the player’s mind significantly [48] and to stimulate perspective taking and/or behavioral changes that are associated with empathy toward others like patients [49,50]. For instance, in previous research, Platt et al [51] showed how empathic communication, such as clinicians’ awareness of the patients’ affective states and showing appreciation of the patient’s feelings, may reduce patients’ feelings of isolation [43]. In addition, the findings, experiences, and design suggestions from this paper may directly benefit other researchers in the future developing empathic games for patients with chronic pain specifically or for patients who must manage similarly invisible chronic conditions. Given our aging population, this may be particularly useful as an approach to medicine shifting from treating acute conditions to managing chronic conditions and promoting wellness.

Some of the participants questioned, “Was my pain sensation the chronic patients’ pain too?” Although it is impossible to make this determination in this study, we assume that the sensations were not the same. For one thing, chronic pain is unique and difficult to describe, let alone recreate specific perception. For another, the interview findings suggested that current gameplay increased the participants’ awareness of chronic pain patients’ situation through the narrative, visual, and audio feedback in AS IF.

This study also had a few limitations that may affect the empathy outcomes and bring a risk of bias regarding the conclusions. First, the sample size for this preliminary study was small, and a larger number of clinicians will be tested once the prototype is revised and prepared for clinical deployment. Next, to avoid overwhelming participants with assessment instruments, the level of immersion in VR was not measured, so we do not know if there are potential relationships between immersion and changes in empathy. Moreover, we only conducted a preliminary, in-laboratory study using questionnaires and interviews to evaluate changes in empathy, and no real-life assessment has been implemented. However, evaluating the pragmatic aspects of the VR game is definitely something we are planning. This involves running a practical test immediately after the subsequent study, asking participants to donate a portion of their study compensation to a hospital’s foundation or a nonprofit organization for patients with chronic pain. Finally, we did not have a control group and did not conduct a follow-up study to see if any long-term empathy behavioral changes persisted. Investigating these factors to determine whether they affect changes in empathy are important next steps.

For a long time, researchers have been looking for evidence about how the sense of embodiment in VR may impact one’s cognitive perception [48]. Various potential impact factors of virtual embodiment on empathy (perspective taking) have been investigated, notably, attitudes toward racial bias [52], gender bias [53], and age [54,55]. We explored the relationship between SoO, SoA, and empathy through correlation tests. However, although participants reported a medium-level SoO and SoA over the virtual body, no significant relationship was discovered.

In summary, to put nonpatients in the shoes of patients with chronic pain, players inhabit a virtual body of a patient with chronic pain who attends to everyday tasks in the VR game AS IF. It simulates several experiences common to chronic pain: physical limitations of movement and a patient’s verbally articulated self-talk. The visual-motor synchronicity of a player’s full-body movements mirrored by the avatar appears to elicit identification with the avatar. The results from the mixed methods study revealed that the game was effective in improving implicit and explicit empathy. Furthermore, the findings showed that the game raised the emotional and perspective taking aspects of players’ empathy. However, no associations were found between the sense of embodiment (SoO and SoA) and the empathy scales in this game. On the basis of the analysis of participants’ feedback, we developed and proposed design suggestions for empathic games designed to facilitate understanding of pain patients. In future work, we plan to iterate the design features and study protocols according to participants’ feedback. We also plan to conduct a randomized controlled study with a larger sample size that is more diverse in terms of gender and age. As for the game, we plan to implement tactile feedback in the controllers (or body sensors) that matches the game tasks, and we would also prepare virtual avatars of different genders and ethnicities.