Highmark Capstone Sprint III: The Attack of the MHCI’ers
The A Team is back at it again with another post! The last two weeks for us have been centered around a big research push: deepening our understanding of our target users and broadening our understanding of their care environment. Many meetings, mock hospital sessions, and affinity diagramming pizza parties later, we’re excited to keep moving forward and expanding our research scope to bring clarity to the healthcare landscape.
01. The Pretotype II: The Bill Strikes Back
In our previous post, we were wrapping up our pretotype testing with other students in our cohort. As a quick refresher, we sent our “patients” through a mock hospital visit, where they were examined for a potential broken arm and informed they needed a surgery to set the bone (our “shock factor”). Half of our participants were given a wristband that tracked the care they received and their coverage at each step and displayed it in real time. A few days later, we delivered a bill to our participants and observed their reactions and thoughts as they read through it.
From our hospital experience, we found that our pretotype actually accomplished the opposite of what we intended. While it did increase transparency in terms of care received, it did so at an already very stressful moment. In addition to interfacing with doctors and making decisions about their arm, patients are now presented with how much they might have to pay and start worrying more about that rather than the care they need. This also prompted patients to ask more questions about what the procedure will cost them and what other options were available. While these are good things for the patient to consider, in a stressful or time-sensitive situation they might make decisions based on incomplete information and possibly refuse necessary care.
The reaction to the bill also defied many of our expectations, as patients were generally more trusting of high-priced procedures than lower-priced miscellaneous items (such as surgical gloves). They noted that if the doctor thought it was necessary then it was probably worth the cost, even if they weren’t thrilled with the idea of paying for it. Many patients were also conflicted between what needed to be done and the price for that care. They wondered if there were other options they could have pursued, and felt trapped now that they had to pay the bill and had no other choice.
02. The Affinity Diagram Awakens
Beyond our user testing, we also wrapped up our first round of primary interviews with patients and insurance members. Our task now turned to synthesizing our notes in order to create one “voice” across our interviewees: what did they want from their experience, what frustrated them, and what would they consider a good experience. To find these answers, we turned to affinity diagramming.
We started the process by debriefing each interview in pairs and pulling out “pseudo-insights”: major points the interviewees brought up, their thoughts and experiences, and their general knowledge about their insurance plans and the healthcare system. All together this came to about 300 notes from 9 interviews.
Next, we sat down at 10am on a Sunday, ordered two pizzas, and got to work.
The majority of our time over the next four days was spent grouping our white notes (individual notes from each interview). We first grouped them by general topics such as literacy, positive and negative experiences, and access to care for example. From there we worked on regrouping these as blue notes: a higher-level insight that is supported by some of the white notes. For example, one of our blue notes was “I expect insurance to take care of me.” The process is then repeated by grouping the blue notes under pink notes, which point to a specific issue or concept, and finally grouping the pink notes under green notes, which are general areas of concern. Here’s our final product!
Creating our affinity diagram really validated the problem space we were working in, and helped us uncover some insights about our target user group:
1. Insurance as a *kinda* utility
“I don’t interact with my insurance app except during enrollment.” -P10
“I wish they had told me [that my emergency appointment wasn’t covered] when I called them but they only told me after the appointment was done” -P10
In many of our previous meetings, our client Zach described an ideal insurance experience as similar to a utility: it’s there when you need it but otherwise you don’t want/have to worry about it. We found this idea through our work, but more than with utilities we found that users want to understand what is happening (what they’re being charged for, what care they’re getting, etc.) during the process. This makes it a little more engaging than a utility, and an interesting concept to design around.
2. The strength of personal networks and connections
“I relied on the Carnegie Mellon Moms Facebook group to learn about which doctors to go to. Moms are the most intelligent network.” -P03
Many users expressed more positive opinions when they developed a personal connection with their doctor over time, and when searching for insurance and care options people generally relied on people they were close to for advice. These personal connections can have a huge influence on one’s experience, and begs the question of how we can foster this between members and their insurance.
3. Needing to feel prepared
“I’ve had enough experience taking care of other people to see what the cost of that [insurance] could be, so there was no way I was going through a period of time without health insurance because you’ll be bankrupted so fast.” -P03
The amount of research and financial planning many people do, as well as in some cases taking out multiple insurance policies depending on personal factors (upcoming travel for example), demonstrate a general need to feel prepared. Especially when dealing with health (where a sudden accident can leave you in tens of thousands of dollars of debt) people do not want to take any risks, and when necessary will put in the effort to avoid that risk.
03. The Return of the Primary Interviews
With the members interviewed and the affinity diagram diagrammed, we now get to go back out and talk to the other side of the care equation: doctors, nurses, administrators, and companies. Now that we have an idea of how patients and members view their care and insurance, we want to understand how their needs are being met and what obstacles are present on the service-provider side that prevent doctors from meeting these needs. We also want to see how companies go about choosing insurance plans to offer their employees and how they help connect employees with resources to understand and utilize their insurance. Our set of questions can be found here.
Recruiting users became more of an obstacle at this stage than previously. Whereas for members we were able to leverage our personal connections to recruit a diverse interview pool, in this case we had to make a lot of cold calls to set up interviews, especially when working with companies and HR departments. This slowed down the process significantly from our first round of interviews, but with enough persistence we kept making headway.
These interviews are still in progress, so be sure to check back for an update in our next post!
04. The Rise of the Deep Dive
With all this progress made, we took some time to present our research findings to our cohort, gather some feedback, and start to develop our roadmap moving forward into March. Our peers gave us some great leads on articles to look into about unwanted care at teaching hospitals, some questions to consider in terms of issues with patient trust and procedure transparency, and additional scenarios to consider as we explore our domain further (such as long-term or terminal illnesses).
In terms of additional interviewees, we found some great candidates in an unlikely place: the HCII Connect networking event! Not only were we able to improve our career prospects, but we found people who were working in the same domain and in similar industries. So we grabbed their contact information and hope to reach out to them soon!
Beyond interviews though, we decided it’s time to dig deeper into our problem domain, and this is going to be a big focus of our next phase of research. This is going to include a few different methods:
- First, we want to look at the current healthcare landscape and do a SWOT analysis (strengths, weaknesses, opportunities, threats) on 10 different major healthcare companies to see how these companies tackle the same challenges Highmark faces and has worked and not worked for them.
- Next, we want to experience what the process of shopping for insurance is like firsthand. In the next few weeks we’re going to take a trip out to one of Highmark’s storefronts and shop for insurance as if we were the customers. We also will start running some tests with ourselves and other users in looking up insurance online in a more controlled setting.
- Finally, we want to identify analogous domains that we can probe deeper into but that don’t have the same level of complexity and privacy laws as healthcare. This can help speed up our research and avoid certain restrictions in terms of interviewing people about healthcare.
And these are just a few of our ideas. We have a running list of research methods that we’ll continue to pull from as the project rolls along:
05. A New Sprint
So now what’s on the horizon? In addition to the plethora of research we described, in the next few weeks we’re also going to ideate our next prototype, this time a conceptual one! We also are preparing for Spring Break, which means laying out our goals for mid-March and how we can accomplish them while we are not co-located… and taking a nice breather for a minute!
Hope you enjoyed our Star Wars themed post (+10 health points if you picked up on that in the section headers), and we’ll see you back for the next post!
- The A Team
