Reflexion provides virtual at-home physical therapy. An engaging avatar, 3D motion-capture technology, and telehealth visits make recovery convenient for patients, insightful for clinicians, and cost-effective for payers. The system (and avatar) is named Vera.

Perform user research to aid in the creation of a mobile app version of Vera. At the time, all physical therapy had to be done in front of the system installed by the patients' TV as it operated similarly to an Xbox Kinect. Reflexion wanted to use the sensors built into most smartphones to create a smaller and more portable companion to Vera. This would allow patients the freedom to practice their physical therapy in any location and would simplify the on boarding process as there would be no equipment to install into their home.

The actual customers of Reflexion are the hospitals and the physical therapists who buy and proscribe the software. There were many, many interviews conducted with both of these groups. The two primary concerns of this group were being able to increase the number of patients each physical therapist could see at a time (which also meant decreasing the amount of time spent on each individual patient) and being able to give better care. These were both successfully addressed by Vera. However, since physical therapists viewed this data on a web portal and this project was concerned with the creation of an app for patient use, most of that research is not relevant now. Suffice to say that the user research was done and successfully baked into the main product of Reflexion.

We all know that user research is important, but working with Reflexion Health was one case that really reinforced that point for me personally. As much as we try not to, it can often unavoidable to go into user research with your own preconceived ideas of "This is what I think the user will be looking for." Reflexion was a special case for me because not only did my perception change dramatically while conducting user research but I learned things about our users I never would have even thought to ask questions about ahead of time. This is why UX designers can't make up user personas out of nothing or on a hunch because sometimes you just don't know what you don't know.

Reflexion Health makes software that can track a person's movements so that patients can receive real-time feedback as they do physical therapy in their home. Before I joined the company, there was an initial idea that their primary user base would be made up of sports injuries. Early user personas are for people such as a 21 year old woman who got an ACL injury playing soccer.

However, the company quickly learned that most people who need physical therapy and were interested in the software were above the age of 55. This group would clearly be less used to technology than an 18-30 year old but great pains had been taken to make the software intuitive so there wasn't a perceived problem. However, when I joined the company as usability testing was beginning, we were forced to realize that what was intuitive to us would not be to someone so unused to dealing with technology. In many of the lower-income communities that Reflexion started working with, users had never used a keyboard before and so making secure, HIPAA-compliant passwords meant teaching patients the difference between typing upper and lower case letters. For a group of people working at a technology startup, this was just not something that anyone had remembered having to learn. We didn't know what we didn't know.

We onducted focus groups as well as direct, non-direct, and ethnographic interviews with patients and physical therapists as well as non-direct interviews with hospital administrators. The top takeaways from this were:

1. The software was helping as intended. Where before it would take about 6-8 weeks to heal, with Vera this was reduced to only 3-4 weeks, allowing patients to go back to daily life faster. Additionally, with typical physical therapy, patients would get seen about every 3-4 days (~2-3 times a week), but with Vera therapists could check up on a patient every day. This means that patients could report any pain and have it corrected much sooner and that they can get their exercise routine changed up much faster to accelerate recovery.
2. The software was intuitive once patients were past the process of creating a username and password to log into the software. The issue was getting through to that point.
3. The gamification of PT was working quite well. Patients were very excited to improve their "stats". Where therapists were quite happy with a patients above 70% adherence (since it is higher than the typical average of ~50-60%), patients tended to view this as a "C" grade and worked to improve their adherence to above 90% to get an "A". They liked to see their "score card" and to be able to quantify how they were doing.
4. The patients really enjoyed the techie-ness of the system. It was commonly described as "geeky," "futuristic," or having "all the bells and whistles."
5. Installing the physical equipment was still the biggest issue. It only took ~5-10 minutes to install, but a technician was still sometimes required to ensure it was quick and completely painless. The kit was around 15-20 pounds which was a worry for injured elderly patients, especially if they needed to install the system upstairs.

Those months of research were than condensed down into two user personas. Before user research, I would not have thought to separate the patients into these two groups, but they really did use the product differently and had differing concerns that needed addressing.

One of the personas (John) had never used a computer before, while the other (Helen) was used to the technology side but had concerns about other logistics.

John embodied the patients with technical concerns. He has never used a computer before, partly due to habit and partly due to price. He's retired from the blue collar work he used to do, and as such never needed a computer at his job and never had a need to buy one for himself at home. He has a mobile phone but it is not a smartphone. When his physical therapist recommended a new technology to help his knee heal faster, he was a bit unsure. But he tried it out at the clinic and really liked it, saying "It's got all the bells and whistles, doesn't it!" He's excited to try it, but asks for some extra help to set it up. He is very nice and willing to learn during the training but doesn't like feeling that he's not good at something and will react badly if the person teaching him the system is demeaning or assumes he should already know something.

Helen embodies the patients with non-technical concerns. She is very active and technologically savvy. She always has her cell phone close in case her grandchildren need her, and she loves writing emails to her friends. She fell and broke her hip half a year ago and is frustrated by her slow recovery. She's worried about how Vera will impact her family's routines and doesn't want to be a bother, so she needs to know the logistics before she agrees to try the new software. Will she need help carrying the physical system? Is it easy to disconnect so her husband can watch tv? Will the system be large or take over the living room? She likes the idea of not needing help to drive to her physical therapy appointments and of getting better faster, but she doesn't want to bother her family.

The idea of an accompanying app was enthusiastically embraced but the Helens but didn't offer as much value to the Johns of the group. Since the Johns of user testing probably did not have a smartphone, the app would not add additional benefit to them. The Helens, however, were quite excited about the idea of an app since it would allow them to practice their physical therapy in any location, would ensure there was no bothersome equipment they would need to carry or set up, and would mean that they did not need to "commandeer" a communal space every time they did their physical therapy. Knowing this, we were able to tailor the designs of the app to be more suited to what would be its primary users.