patient <> provider messaging

After extensive user research and analysis of patient feedback, the product team decided to focus our efforts for two quarters on improving the messaging platform, i.e., the tool Hello Alpha's patients and healthcare professionals use to communicate and coordinate care.

We conducted interviews with patients to gather problems about the gaps in our platform.

• We exported all 1 and 2 star Trustpilot reviews written by our users and did an analysis of the most frequently used words and phrases.  

• We fed the top words from that list into Trustpilot’s AI-driven “Insights” feature to identify themes among the lowest rated reviews. 

From our user interviews with patients and analyzing their Trustpilot reviews we learned: 

• The time to get a response to a message was too long, making it difficult to have fluid conversations. Especially considering the asynchronous model of communication.

• Oftentimes a patient's message gets sent to a recipient they didn't actually intend.

• Patients disliked having multiple healthcare professionals in a single channel.

• The flow of selecting a subject for a message had a cumbersome user experience.

I collaborated with two product designers on our team to create and wireframe potential solutions. This process helped to uncover gaps in my initial approach and made the final result more concise and simple. We also looked at existing solutions by competitors to give us a benchmark for our ideas.

During the early to middle stages of a project, to optimize our agile process, I worked closely with the software and QA engineers on our workflow diagrams. This helped us to identify gaps early and have both macro and micro frames of reference if the scope of the project were to change. Through this approach we were able to build products that: more closely matched specifications, had a reduced number of QA cycles before launch and had fewer back and forth conversations once the sprint began.

Oftentimes a patient's message gets sent to a recipient they didn't actually intend.

In order to help ensure that patient messages get to the intended team we split the landing page into two sections: first, the list of channels and second, the pathways to each team. We also condensed the channels into a list of the most active/recent messages and hid the remaining channels behind a click. While we wanted to give patients clear avenues to the right team, another priority was to ensure patients don't miss unread messages.  

The chosen avenue then leads to either the filtered list of clinical messaging channels or directly to the support team, making it easier for patients to understand who would be receiving their message.  

The time to get a response to a message was too long, making it difficult to have fluid conversations. Especially considering the asynchronous model of communication.

We added a dynamic badge and subtitle which show the average response time for a team, driven by a content management system. This helped to set expectations ahead of a patient sending a message and also gave the care team more flexibility since they could adjust the response time in high volume scenarios.

Patients disliked having multiple healthcare professionals in a single channel.

Until we could allocate the resources to solve this problem fully, we opted to provide more information to the patient about each person messaging them and why they are present in the channel (not shown here: tapping the info icon leads to a bio of the healthcare professional).

The flow of selecting a subject for a message had a cumbersome user experience.

We also designed a way for patients to be shown FAQ articles which are populated based on their circumstances. This allowed us to answer routine questions through automation, ahead of a patient sending a message, saving time on both ends.

Continued: The flow of selecting a subject for a message had a cumbersome user experience.

In the previous iteration, selecting a subject involved scrolling through a very long list of subjects sorted by popularity, which kept growing. We opted to redesign this interaction by rearranging the list into a "subject tree". This gave us the flexibility to add as many subjects as needed per category, while still keeping the flow manageable and easier to navigate.

We also added a way to search subjects by keywords and hidden meta tags, giving even more power to patients and the healthcare team when categorizing messages. For e.g. if a subject the patient has in mind doesn't specifically match our visible content.