Search all of the Society for Participatory Medicine website:Search
The Journal for Participatory Medicine's website has moved. Please check out the new website for the latest articles.

Abstract

SummaryDigital health tools are increasingly used to engage patients and improve outcomes. Their effectiveness, however, relies on consistent patient use. Developers have struggled to maintain high levels of durable, ongoing patient engagement with digital health tools. This may be related to an element that is often overlooked in the design of these tools: empathy. Empathy is known to be a fundamental principle of medicine, and high empathy doctor-patient relationships have strong positive effects on patient satisfaction and clinical outcomes. By breaking down the definition of empathy and looking to the field of computer science for background, this paper describes a framework that can be used to inform the deliberate design of empathy in digital health tools.
Keywords: Digital health tools, empathy, doctor-patient relationships, patient satisfaction, positive clinical outcomes.
Citation: A framework for designing digital health tools with empathy. J Participat Med. 2017 Feb 17; 9:e4.
Published: February 17, 2017.
Competing Interests: The authors have declared that no competing interests exist.

Introduction


Consumer-facing digital health tools such as mobile health care apps, telemedicine tools, remote monitoring, and virtual care are enthusiastically promoted as a way to improve patient engagement and clinical outcomes. Despite many promising studies linking these tools to better outcomes, [1][2][3][4] the effects are inconsistent. One significant barrier to the effectiveness of digital health tools is continued and consistent use by patients. [5] In fact, a dose-response relationship has been found between the effectiveness of digital health tools and a user’s level of engagement. [6][7] Improving engagement with these tools, therefore, is key to fully achieving the potential of these innovations. Researchers have attempted to determine the most effective ways to improve user engagement. Theoretical models, such as the Fogg Behavioral Model, invoke constructs such as motivation, ability, and triggers to inform the design of persuasive technologies. [8] A recent review of the effectiveness of technology-based triggers in enhancing engagement with digital health tools found borderline positive effects. [5] The most effective way to improve sustained patient use and engagement with digital health tools remains unanswered.

As we continue to think about this challenge, we should consider what we already know about traditional engagement in healthcare. Long before the advent of digital health tools or personal computers, physicians have assumed the responsibility of personally engaging patients to improve health outcomes. Empathy has consistently been a fundamental element of the successful therapeutic relationship between doctors and patients. High degrees of empathy have been shown to improve clinical outcomes, [9] compliance, [10] emotional health, [11] and patient satisfaction. [12] Despite these proven benefits, empathy is often overlooked in the design of digital health tools and may be necessary to improve engagement. Evidence for this comes from what we know about human-machine interaction design. The “uncanny valley,” for example, teaches us that as tools become more human-like in their interactions with us, their capacity to engage us and evoke empathic responses increases to a degree. [13] As the human-likeness of these tools reaches a certain threshold, however, this capacity to engage and evoke response from us precipitously drops to a nadir. This is the uncanny valley. It is only beyond this valley, as the “last mile” of human-likeness of the tool is crossed, that human engagement and empathic response reach their maxima. By breaking down the definition of empathy and looking to the field of computer science for background, this paper describes a framework that can be used to inform the design of empathy in digital health tools.

Defining Empathy

In a widely accepted definition, physician empathy is described by Mercer and Reynolds as the ability of a physician to “(a) understand the patient’s situation, perspective and feelings (and their attached meanings), (b) communicate that understanding and check its accuracy and (c) act on that understanding with the patient in a helpful (therapeutic) way.” [14] In the evolving field of health information technology, “digital empathy” has been defined as “the traditional empathic characteristics such as concern and caring for others expressed through computer-mediated communications.” [15] Using the three core elements of Mercer and Reynolds’s definition (understand, check, and act), designers can strategically incorporate a real sense of empathy into the features of digital health tools. This subset of digital empathy can be referred to as “automated empathy.”

Lessons from Computer Science

Attempting to “automate” empathy may seem contradictory, but in fact, it is not a new concept. In the field of computer science, this paradigm goes back to 1960’s when the first computer program capable of engaging in empathic dialogue was created. [16] Based on simple keyword matching and natural language processing, a computer program, ELIZA, responded to a user empathically and was rated by psychiatric patients as being a “good listener.” More recently, studies have focused on the effect of incorporating empathy into computer programs on improving user interaction. In one study by Brave et al., a computerized blackjack game was manipulated to include empathic emotion via simple agent expressions such as “You won! That’s wonderful!” or “I’m sorry that you lost.” [17] Enabling empathic emotion statements led to more positive user ratings of the computerized agent, including greater likeability and trustworthiness, as well as greater perceived caring and feeling of support compared to the same computerized agent without empathic communication. In another study, Klein et al. studied the effect of empathy on playing time for frustrated users. [18] Players were issued a survey about levels of frustration. The control groups either just answered the questions or were also allowed free text boxes (to “vent”). In the experimental group, players answered the same questions about frustration, but the computer also responded with empathic text dialogue (eg, “Sorry to hear things didn’t go so well.”). The results showed that displaying empathy via simple text dialogues led to a significantly longer playing time in subjects who had been previously frustrated by the same game compared with similar text-based interfaces that only allowed users to vent their feelings, or to those which ignored their feelings altogether.

Framework for Incorporating Empathy into Digital Health Tools

The successes of automating empathy in computer science are encouraging. A systematic way to apply these principles to digital health tools can be conceptualized via a simple framework. The basis of this framework involves the three core elements of the Mercer and Reynolds definition of physician empathy: understand, check, and act.

Step 1: Understand

The first part of displaying empathy is to understand the perspective of the user, or in this case, the patient. User-centered approaches such as design thinking principles are intended to do just that. Design thinking is “a step-by-step approach to problem-solving that involves observing and interviewing people as they go through an experience, and then using that information to prototype and test ways of improving the product or process.” [19] In recent years, this methodology has been increasingly applied to health care innovations. For example, digital health developers of an HIV adherence app held focus groups to determine patients’ opinions about ideal adherence features as well as their responses to app characteristics and understanding of the HIV disease state based on the app features. They used this information to inform their design and content. [20] This type of structured, patient-centered approach to initial and iterative feature design will allow for the discovery of patients’ anticipated thoughts and feelings throughout their intended experiences.

Step 2: Check

Once a well-developed understanding of the patients’ feelings and perspectives has been established, the second element of empathy is to check the accuracy of this understanding. The computer game that was tested by Klein et al. described above checked understanding by simply surveying the users about their levels of frustration. [18]In this way, the computer program could check the accuracy of its assumption that the user is likely to be frustrated at certain time points in the game. By completing a thorough Step 1 (understand), designers of digital health tools can anticipate patients’ feelings at certain points in their courses of care. Then, by incorporating “checks” into the system at pre-defined touch points, an accurate understanding of patients’ feelings can be determined. This will allow for the final element of an empathic interaction: act.

Step 3: Act

Once a patient’s feelings are understood and the accuracy of this understanding is checked, then the system can be designed to act on that understanding in a helpful and therapeutic way. Natural language processing and branching logic are two examples of techniques that can be used to respond to users in a reactive way. ELIZA, the computer program that was deemed a “good listener,” used natural language processing and simple text matching to respond to users empathically.16 The computer game studied by Klein et al. used branching logic based on responses to the survey about frustration to respond with pre-defined, empathic text messages. [18] Similar processes should be built into digital health tools to respond to patients empathically. A successful example of this process in healthcare is the use of text-based smoking cessation programs. The National Cancer Institute has developed a library of short text messages that are programmed to be delivered based on patient inputs. For example, if a patient were to indicate a medium craving level, a pre-scripted text response would be sent such as, “We know how you are feeling. Think about what you are gaining and why you want to quit smoking. Stay focused. It will get easier.” [21] Similar programs, such as Text2Quit, have been shown to improve smoking abstinence in randomized trials. [22] Digital health tool writers and content development teams should take care to include this type of empathic, caring language in pre-scripted content. Following Steps 1 and 2 of the framework, this content can then be delivered based on the system’s ability to understand and check a user’s feelings, completing a thorough yet automated act of empathic communication.

Conclusion

While the doctor-patient relationship will always remain at the core of medicine, digital health tools can extend the physician’s reach outside of the traditional clinical encounter and augment that relationship. To be effective, these tools must be designed to encourage durable and continued user engagement. Empathy is often overlooked in the design of these tools but may be key to promoting engagement. We know that high empathy doctor-patient relationships have many strong positive effects on patients, and computer science has suggested that automated empathy can have similar benefits. Frameworks such as the one described in this paper can be used to inform the design of automated empathic communications in digital health tools.

References

  1. Morrison D et al. Digital asthma self-management interventions: a systematic review. J Med Internet Res. 2014 Feb 18;16(2):e51
  2. Murray E, Burns J, See TS, Lai R, Nazareth I. Interactive health communication applications for people with chronic disease. Cochrane Database Syst Rev. 2005 Oct 19;(4):CD004274.
  3. Pal K, Eastwood SV, Michie S, et al. Computer-based diabetes self-management interventions for adults with type 2 diabetes mellitus. Cochrane Database Syst Rev. 2013 Mar 28;(3):CD008776.
  4. Foster C, Richards J, Thorogood M, Hillsdon M. Remote and web 2.0 interventions for promoting physical activity. Cochrane Database Syst Rev. 2013 Sep 30;9:CD010395.
  5. Alkhaldi G, Hamilton FL, Lau R, Webster R, Michie S, Murray E.. The effectiveness of prompts to promote engagement with digital interventions: a systematic review. J Med Internet Res. 2016 Jan 8;18(1):e6.
  6. Funk KL, Stevens VJ, Appel LJ, et al. Associations of internet website use with weight change in a long-term weight loss maintenance program. J Med Internet Res. 2010 Jul 27;12(3):e29.
  7. Strecher VJ, McClure J, Alexander G, . The role of engagement in a tailored web-based smoking cessation program: randomized controlled trial. J Med Internet Res. 2008 Nov 4;10(5):e36.
  8. Fogg BJ. A behavior model for persuasive design.
    Persuasive ’09 Proceedings of the 4th International Conference on Persuasive Technology
    Article No. 40
  9. Di Blasi Z, Harkness E, Ernst E, Georgiou A, Kleijnen J. Influence of context effects on health outcomes: a systematic review. Lancet. 2001 Mar 10;357(9258):757-62.
  10. Kim SS, Kaplowitz S, Johnston MV. The effects of physician empathy on patient satisfaction and compliance. Eval Health Prof. 2004 Sep;27(3):237-51.
  11. Burns DD, Nolen-Hoeksema S. Therapeutic empathy and recovery from depression in cognitive-behavioral therapy: a structural equation model. J Consult Clin Psychol. 1992 Jun;60(3):441-9..
  12. Menendez ME, Chen NC, Mudgal CS, Jupiter JB, Ring D. Physician empathy as a driver of hand surgery patient satisfaction. J Hand Surg Am. 2015 Sep;40(9):1860-5.e2.
  13. Mori M. The uncanny valley. IEEE Spectrum. 2012; 12 Jun. Available at: http://spectrum.ieee.org/automaton/robotics/humanoids/the-uncanny-valley. Accessed: December 7, 2016.
  14. Mercer SW, Reynolds WJ. Empathy and quality of care. Br J Gen Pract. 2002 Oct; 52(Suppl): S9–12.
  15. Terry C, Cain J. The emerging issue of digital empathy. Am J Pharm Educ. 2016 May 25; 80(4): 58.
  16. Weizenbaum J. ELIZA—a computer program for the study of natural language communication between man and machine. Commun ACM. 1966; 9:36–45.
  17. Brave S, Nass C, Hutchinson K. Computers that care: investigating the effects of orientation of emotion exhibited by an embodied computer agent. Int. J Hum-Comput Stud. 2005; 62:161–178.
  18. Klein J, Moon Y, Picard RW. This computer responds to user frustration: theory, design, and results. Interact Comput. 2002; 14:119–140.
  19. Wykes S. Design Thinking as a Way to Improve Patient Experience. Stanford Medicine News Center. Available at: http://med.stanford.edu/news/all-news/2016/06/design-thinking-as-a-way-to-improve-patient-experience.html. Accessed: December 2, 2016.
  20. Rosen RK, Ranney ML, Boyer EW. Formative research for mhealth HIV adherence: the iHAART app. Proc Annu Hawaii Int Conf Syst Sci. 2015 Jan;2015:2778-2785.
  21. Resources for Health Professionals. Smokefree.gov Available at: /understanding-smoking/help-others-quit/resources-health-professionals. Accessed: December 7, 2016.
  22. Abroms LC, Boal AL, Simmens SJ, Mendel JA, Windsor RA. A randomized trial of Text2Quit. Am J Prev Med. 2014 Sep;47(3):242-50.

Copyright: © 2017 Nina Gonzalez. Published here under license by The Journal of Participatory Medicine. Copyright for this article is retained by the authors, with first publication rights granted to the Journal of Participatory Medicine. All journal content, except where otherwise noted, is licensed under a Creative Commons Attribution 3.0 License. By virtue of their appearance in this open-access journal, articles are free to use, with proper attribution, in educational and other non-commercial settings.

 

Donate