CHALLENGE

Patients recently discharged from hospitals are more likely to be readmitted in the first 7 days. Confusion around discharge instructions or ignoring symptoms after a procedure can lead to complications that could result in readmission or a costly emergency room visit.

solution

Daily Check-Ins is a feature of Identifi Engage app that leverages low touchpoint monitoring to capture daily, self-reported health status updates directly from the patient. A patient’s Care Advisor receives real-time alerts when a check-in response merits further follow- up.

Healthcare professionals can effectively maximize their time compared to traditional methods of direct outreach.

Role

Lead designer— discovery, user research, design, prototyping, and testing

PhoneMock.png
 

Research

Will people want to answer health questions?

To test the viability of the feature, we set out to run a pilot that enrolled recently discharged patients in a 30-day program. Patients were asked 5-8 Yes/No questions everyday regarding their physical state and if they’ve experienced any changes since leaving the hospital. 

The pilot assessed patient willingness to answer the questions and the care advisor’s effectiveness in responding to alerts and preventing readmission. If successful the feature could expand to more comprehensive health assessments.

Health assessments are typically 80+ questions that care advisors need to conduct with patients at the start of their work together either over the phone or in person. It can be difficult for care advisors and patients to coordinate a time to complete the assessments. If the assessments were for patients to complete on their own time, it could free up time for care advisors to focus on other tasks or help more patients.  

 
 
Carol.png
It’s not always easy to complete an assessment in one phone call. Betty doesn’t always have the energy to stay on the phone with me for that long.
— Carol, RN Care Manager
 
 
Betty.png
With the app I can complete the check-ins on my own time and I don’t have to worry about getting back to Carol if I miss her call.
— Betty, Patient
 
 

Metrics of success

For the pilot team, we would focus on the following metrics to better assess the effectiveness of the Engage apps for both patients and clinicians.

• Inpatient admission rate decreased by x% compared to non users

• Readmission rate decreased by x% compared to non users

• Emergency room visits decreased by x% compared to non users

• Usage decrease cost of patient’s care by x%

• Care Advisor increased their case load by 3x

 

What if we use a chatbot?

During the early ideation phase, we explored if a chatbot could be an effective method to administer questions. It could bring a friendlier tone to typically cold survey questions. The chatbot could be built into the existing care team group conversation as an additional member.

Executive stakeholders were in favor of an opportunity to market a product with AI functionality and dev teams were excited at the possibility to build with new technologies, but is it the best experience for users?

Why it won’t work

Chatbots work best as conversational AI where users can both answer and ask questions. But for our use case the chatbot was asking all the questions and the user was doing all the answering. We were trying to force a survey into a chatbot experience and testing revealed confusion.

The development team was also facing technical challenges in finding ways to implement a chatbot. And since we were trying to get to an MVP for the initial pilot, why would we invest in such sophisticated technology for a 6-question survey?

As these challenges began revealing themselves we pivoted our design to a more focused survey experience that was separate from the app’s chat screen.

ChatBotScreen3.png
 

Mobile survey inspiration

I explored mobile survey apps to gain understanding of how others handle various interactions.

App Examples.png

Open Design Questions

  • Do we show progress? Is progress a percentage, number of questions left?

  • Can the user skip or leave answers blank?

  • Should the user be forced to take an action to see the next question?

  • How do we design the layout in a way that is consistent with the visual language of our platform

  • Consider that our users are older and potentially visually impaired, how can we make this an easy experience for them?

 

Design

UI Design

I explored various button designs and layouts paying attention to targets areas and screen real estate. With each iteration and testing session, I gathered feedback from teammates and users that helped lead me closer to the final design.

 

Through testing and iterations based on user feedback we landed on a final survey design that best suited an experience that would encourage completion. Following Steven Hoober’s research on thumb zone in Designing Mobile Interfaces, I placed secondary actions in harder-to-reach upper edges. The more easily accessible centered text and buttons keeps the user’s attention focused.

 
ThumbZone.png
AnsweredQuest1.png
 
 

In-app survey launch

We also needed to solve for how the user would launch the survey from inside the app. The persistent banner of Option C best grasps the user’s attention without taking away from the chat experience. At any point users could close the check-in to ask a question in the chat and resume their place when they return.

We also designed a solution leveraging the phone’s notifications to initiate the check-in without forcing the user to navigate to it from within the app.

 

On web Care Managers would use a calendar view in the patient’s profile to review their alert history and progress.

Calendar.png