OverviewProblem DefinitionDesignOutcomes & Learnings
A mockup of an electronic medical record with a Telehealth video window over it

athenaTelehealth

The Covid-19 pandemic created an urgent need for a telemedicine as a way of connecting physicians and patients in the uncertain first year of the crisis. The athenaTelehealth project began in April of 2020.

My team was tasked with taking our minimally viable beta experience bring it to maturity as a commerically available GA product. My focus was on researching and redesigning the physician’s telehealth experience to allow them to provide the best care possible.

Role: Lead designer on scrum team
Timeline: Sept 2020–August 2021

Problem Definition

Virtual Observation

Early beta feedback had revealed a number of potential product enhancements: screen-sharing, whiteboarding, chat, and more direct client requests. However, it was difficult to understand the themes in these requests and the underlying issues that they were meant to address. In the taking on this project, I knew we needed to take a step back and understand how the existing product was working.

In a typical year at athena, exploratory research involves site visits to medical practices to observe the clinical workflows first-hand. In 2020, both we and our clients had to figure out new methods. I piloted a series of 3 virtual “site visits”  where a physician would share their screen with us as they conducted a Telehealth visit. I was able to perform contextual inquiry as we normally would on site, with the added bonus of having a clear view of how they were interacting with athena’s EHR system during the visit.

Finding 1:
Physicians struggled with the balance between documentation and maintaining patient rapport

Every provider we spoke to emphasized the importance of moving the video window near their laptop webcams so that they could appear to make eye contact with the patient and establish rapport. However, this means that on every screen, the video would be covering up important documentation. Providers were constantly adjusting their video during the visit to access what was behind it, which led to a decrease in efficiency and an increase in documentation time.

A mockup of an electronic medical record with a Telehealth video window over the center, covering the documentation space behind it
Providers did not keep the video in the default bottom right location, but consistently moved it to the center of the screen.
Finding 2:
Tasks that were done by staff previously in the physical office settings now relied on the physician

In a physical office, every physician is supported by a team of staff members who handle the administrative tasks that surround the visit. With Telehealth, physicians now had to do this work themselves, especially if they are physically separate from their team.

These tasks included things like:

—all of which were tasks that physicians were generally unfamiliar with, but were now responsible for in the age of remote work.

These findings helped us contextualize the feature requests that we were hearing and helped us prioritize them into themes. We learned that before we pursue any enhancements, we needed to address the glaring usability issues of the beta application. Following our redesign, we would be able to pursue improvements in the vein of reducing physician burden during the whole visit, not just in the video call.

Re-Designing the Video Call

While video-calling is a somewhat "solved" problem (that has also had lots of room for improvement during the pandemic!), Telehealth has slightly unique needs compared to a typical Zoom meeting.

  1. There is always a featured user on the call—the patient.
  2. Telehealth visits nearly always involve multi-tasking.

Introducing Mini View

The first major change we made was the creation of a “mini view” for the Telehealth window. The thumbnail-style view allowed physicians to place the window in their preferred top center location without blocking crucial documentation areas. Sizing controls allow the user to quickly move to standard or fullscreen mode—users could quickly expand to get a better look at a rash, then minimize to resume their documentation.

2 images of the Telehealth mini view. One shows the controls, which are visible on hoverA mockup of an electronic healthcare record screen with a thumbnail-sized Telehealth window on the top center of the screen

Redesigning Video Layouts

Beta metrics showed that the majority of patients were joining from mobile devices. Our MVP video layout was extremely unfriendly for mobile video screen ratios, which likely contributed to the complaints of the video being "too small." The redesigned layout of the standard view is a portrait view, allowing the best visuals of patients joining from mobile.

A before and after image of the Telehealth video design
Before and after of the Telehealth standard view

In our research, we also learned that there are two camps of users - those who wanted the video smaller for multi-tasking, and those who wanted it bigger. For some physicians, such as those in mental health specialties, documentation was less important than having a focused face-to-face conversation with the patient. To better serve these users, we also designed a true fullscreen layout.

Optimized fullscreen view. Up to 4 users are supported on a call, with one "pinned" user (the patient).

Increasing Physician Awareness

As our research found, providers were also struggling with tasks outside of the video call that were previously handled by staff. To combat this, we explored a series of features to increase physician awareness. By giving physicians more visibility into the status of the patient and allow them to share their own status, we could reduce some of the uncertainty and inefficiency around the Telehealth visit. One such feature was the running late status.

Communication Prior to the Call

"Chat" came up frequently as a feature request, with many users referencing other video-calling software. However, the main user need that was expressed related to chat often had nothing to do with chatting during a Telehealth call. Through resonance interviews, we learned that physicians wanted to to reduce confusion and no-shows from waiting patients by messaging them prior to the call – a marked difference from the in-call chat features that many other platforms provide.

Through our interviews, it also became clear that practices wanted to acknowledge the patient's presence and thank them for waiting, but generally did not want a response; responses were impossible to monitor, especially if they were already busy enough to be running late.

I introduced the concept of a one-way "running late status" exclusively for the purpose of letting patients know that their doctor is, in fact, running late. Users could access this from the main schedule as they are monitoring upcoming patients, without needing to start a Telehealth call with a patient.

We chose a pre-written message for consistency and efficiency, both important when the user is in a rush.

Outcomes & Learnings

User Reception

Following the initial re-design, we successfully launched our commercial GA in March 2021. We evaluated the combined usability improvements via athena's Standard Feature Perception Instrument, a standardized survey tool for measuring user satisfaction. Our improvements were able to bring the average satisfaction to a 4.0, our target score for released features. We also managed to eliminate any "1" or "extremely dissatisfied" responses in our post-GA survey.

Overall Satisfaction
3.79 → 4.0
out of 5
Attractiveness
4.1 → 4.2
Ease of Use
3.9 → 4.1
Meets Requirements
3.8→ 4.0
Usefulness
4.3 → 4.3

Post-GA enhancements included the Running Late Status. This feature was received quite positively by many users, and showed us that we made the right bets even though this feature was quite different from the original chat feature that had been requested:

The generic message overall is great. Occasionally it may be nice to type a direct message to the patient depending on the circumstance that's causing the delay.
We are LOVING this new feature!

Future improvements could include allowing the practice to configure a set of pre-written messages to share with patients that would address different reasons for running late, but we were able to validate that this was not a core feature.

Final Musings

The Telehealth project was a gratifying experience with quick, iterative development, something that can be rare in healthcare. These features were just a few of the many changes that we were able to release during my year on the Telehealth team.

This project was a fantastic opportunity to think outside the box in all regards, from my understanding of our users' workflows and needs, to my own processes in how to conduct exploratory research. More importantly, these deviations from the norm have allowed us to experiment with tools and processes that will continue to serve us in post-pandemic life. Remote shadowing has since been adopted as a part of the athena research toolkit, which will allow us to better engage with clients across the country in a low-cost, minimally disruptive manner. Similarly, Telehealth is surely here to stay as a way for physicians to engage existing patients and reach marginalized communities.

Next Project →
Wanna chat about this?
→ mailforyiran@gmail.com
→ LinkedIn