ACM Interactions

Authors: Jina Huh
Posted: Mon, May 20, 2013 - 12:55:32

---

In this blog post, I want to share, from the perspective of a junior researcher, some reflections on how I observed HCI work has been discussed by the medical community, the changes happening in the medical community toward evaluating technology design for health, and the different perspectives on what constitutes novel research between the HCI and medical communities. People who have been pioneering this area (e.g., those who have helped build and organize the Workshop on Interactive Systems in Healthcare (WISH) over the past couple of years, guide the CHI health community, and organize HCI in health conferences) will be able to provide broader insights from their experience working closely with the medical community. WISH is probably the closest place where the themes I am about to explore here have been, and will continue to be, discussed in depth. This post may relate to Richard Anderson’s recent post, “What designers need to know/do to help transform healthcare,” in terms of connecting the two communities together. The difference may be that while Anderson looked at what designers could contribute to healthcare, I am looking at how the two communities differ in culture and the implications for considering close collaboration.

Julie Kientz, Eun Kyoung Choe, and I carpooled to the Fred Hutchinson Cancer Center to attend a talk given by Dr. David C. Mohr, Professor of Preventive Medicine and Director of the Center for Behavioral Intervention Technologies (CBITs) at Northwestern University. The title of the talk was “Going Digital: Building eHealth and mHealth Technologies.” Mohr suggested the seminar to be an interactive session rather than a formal talk. The audience (about 60 people) consisted mainly of people from the medical field—at least we (Julie, Eun Kyoung, Betsy Rolland, and I) were the only ones from the DUB community. Long, enthusiastic discussions with the audience allowed Mohr to cover only his first few slides. We did not get to hear details of the projects going on at CBITs, but we were able to hear general perceptions about designing and evaluating technology for health coming from people in medical fields. (Later I will discuss what my frequent use of the terms medical fields and medical community entails, just in case it bothers you. I also have to mention that what I mean by we will not be consistent, but as a person working to participate in both HCI and medical informatics fields, it is really hard to not use we in multiple ways. Perhaps I will write a separate post about my meaning of we.)

Mohr was quite knowledgeable about the HCI design process and appreciated multidisciplinary teams. He shared insightful challenges in funding and development, contrasting the huge time length that it takes to receive an R01 grant (the biggest funding mechanism at National Institute of Health (NIH)) and finish an RCT with how fast technologies evolve. He shared the changes in how the medical community sees randomized controlled trials (RCTs), or at least the people gathered for an NIH workshop last year to explore alternatives to RCTs. We also got to hear how the audience, mainly coming from medical fields, perceives the role of technology development in general. Hoping that these are not too overwhelming to cover in this one article,  I will now present some burning themes I want to discuss.

Designers make things look pretty

We have heard this many, many times. Ever since I moved from being an art and design student to being an HCI researcher, I constantly have had to remind myself and others that designers are not there to make things pretty but to design the entire interaction. As designers of technology, interaction designers, as opposed to graphic designers, for instance, design interaction, not just the appearance. Many CHI scholars who are designers by training have continued to push this argument for years. 

We saw that sentence again on Mohr’s slides, along with another line that said: “UX/HCI Engineers: Make things usable.” While it is true that UX/HCI people are there to make things usable, “going beyond usability” has been another dominant theme that people have been discussing at CHI for many years. I remember trying to be philosophical about it in my preliminary exam in 2006, proposing to study “Postmodernism in HCI.” 

It was sad to see that statement (designers make things pretty) appear again, especially from somebody who has a good understanding of the HCI field. So how can we solve this? There really is no easy solution. Perhaps, just as designers with a capital “D” continued to fight to become primary stakeholders of project development rather than staying on as consultants or as some third-party group of people providing labor to half-finished projects, we should continue building deeper relationships with the medical community, enough to help both of our communities better understand our roles as stakeholders holding common goals. 

The R01 cycle and the technology adoption cycle: Perspectives on the evaluation outcomes

Mohr showed the tedious timeline of an R01 cycle. In total, it takes about 15 years for one to start synthesizing problems and reporting results from RCTs. He contrasted the timeline with how our phones have evolved in 15 years. Mohr questioned, considering how fast our environment changes, if R01 can really control for any external influence and how we can say that the outcomes from the RCTs are applicable in the real world. This is when he mentioned that the medical community is also seeking alternative ways to evaluate outcomes and that NIH persons actually had a workshop to discuss the matter. The funny part was that the NIH wants to change, but the reviewers with strong methodological views are the ones who are resistant. 

Then Mohr and someone in the audience introduced iterative design cycles as one of the solutions—one of the main methods HCI uses. However, the ultimate goal in the medical community being health outcome changes, not engagement with technology, iterative design approaches may not be appropriate. I remember, with my brave social skills, sharing my NIH career grant idea with one of the leadership faculty at the Duke Center for Health Informatics and the Department of Biomedical Informatics at Columbia University at a hallway at American Medical Informatics Association (AMIA) symposium for brief feedback. The first question I was asked was, “What is the ultimate outcome you are trying to test?” I could not respond to that question. I was used to working for innovative design company and evaluating user experience, not measuring users’ A1C level (a measure for patients with diabetes) in response to using a diabetes mobile app, for instance. This is when I realized that there is a huge gap between the  HCI and medical communities—that we measure different things. 

Evaluation versus innovation

From hearing the audience’s discussion, both Eun Kyoung and I sensed that the granularity of technology design they are describing were at a very high level, such as “Internet use,” or “mobile use.”

I remember going to talks at AMIA and around campus on electronic medical records (EMRs), to which my initial response was, “What about EMRs? Which EMR?” The speakers presented outcomes on how EMRs shifted medical errors, work efficiency, or billing issues, without giving details on the design of the EMR that potentially could have shifted the outcome. After the talk, Eun Kyoung shared how challenging it is to design a really good notification system, and how a small little feature can change behavior outcomes. 

One of the salient differences that is often true between the HCI community and the medical community is that the HCI community is interested in innovation, how novel the technology is, and the medical community is interested in evaluation, regardless of how novel the technology is. Accordingly, strict methods matter (e.g., evaluating their own company employees to test their technology is not valid) for the medical community, while it may not be so important in the CHI community, although some may strongly disagree.

Final note

I am hopeful, with many researchers from the medical community increasingly working with, and hiring, students and researchers at CHI. Several leaders from the CHI community have been working hard to create a close connection between the NIH and the NSF worlds. This is evidenced by a recent grant call on smart and connected health that NSF and NIH collaborated on and both CHI and AMIA’s continued efforts to work together using venues such as WISH. At the same time, as a new faculty member in the field, I feel under pressure to accomplish many of the changes that I believe should occur. 

What is for certain is that designing technology for health is an exciting field to be in. The field continues to evolve and requires multidisciplinary collaboration, creating necessary cultural clashes among disciplines. People in this field are continuing to challenge themselves, such as the NIH workshop people finding alternative solutions to RCTs; the audience at the seminar and Mohr, who initiated conversation in finding best ways to design technology for behavior change; and the CHI community, which is continuously expanding its horizon in designing technology for people.

About my use of the terms medical fields and medical community: I assume that people from the medical community will read this post and think that I may be oversimplifying what I mean by medical community. In my definition, the medical community includes any people coming from academic homes that constitute medicine and public health. Accordingly, in my mind, as an extreme example, biomedical informatics people are also "medical people." Using my definition, I technically should be a medical person too since my current fellowship comes from a biomedical informatics department that is part of a medical school. However, at AMIA, they (or we) call themselves (or ourselves) the informatics people, without the word medical. As much as what I mean by medical may greatly differ in the medical community, what the medical community sees as informatics may greatly differ in the information science community. Who will we consider the closest “informatics person” if people from the American Society of Information Science and Technology (ASIST), iSchool, computer science, and HCI gather together in one room? As much as the role of designers is misunderstood in the medical community, the role of “being medical” may be as equally misunderstood in the HCI community. And my broad use of the term medical fields is evidence of that.

*I also want to thank Julie Kientz, Eun Kyoung Choe, and Wanda Pratt for their feedback and help on this piece.

Jina Huh is NLM Postdoctoral Fellow at the University of Washington Medicine Division of Biomedical and Health Informatics. Starting in fall 2013 she will be an assistant professor at Michigan State University's College of Communication Arts and Sciences.

---

Posted in: on Mon, May 20, 2013 - 12:55:32

Jina Huh