Eko Digital Experience
A 2012 blind study from Stanford revealed 80% of cardiologists can’t get reliable information from an ordinary stethoscope. Yet the iconic tool remains the first line of defense against heart failure, the leading cause of death worldwide. To address this problem, three Berkeley students invented a device that translates analog stethoscope sounds into digital Bluetooth format, and algorithms for diagnostic support.
I joined Eko in 2013 to create the digital ecosystem in which this audio data can be saved, access, analyzed and shared. As VP and lead designer, this meant building Eko’s user research, visual and experience design, information architecture, branding, and product roadmapping from scratch within a fast-paced health tech startup.
Role: Lead Product Designer (1 designer, 2 developers)
Mobile Engineering: Tyler Crouch
Status: FDA cleared and released (2015)
The Eko mobile app is the centerpiece of Eko’s digital experience - capturing biometric data over Bluetooth and packaging it within a light-grade EMR.
Optimizing for screen sizes as small as iPhone 4s, connectivity speeds common in developing countries, iOS and Android compatibility, and language support for English, German, French, Spanish, and Portuguese, it was essential to design as little as possible, and rely on design systems to maintain consistency and responsiveness.
The Eko web dashboard was designed as a desktop resource for accessing and sharing heart sounds and ECG among teams. This product is particularly useful for medical institutions sharing patient records, and academic institutions collaborating with residents.
Due to HIPPA regulations on EMR data, privacy permissions and security posed the biggest challenge to user experience. By separating UI components related to specific patient information as a standalone modular component (Name, Hospital ID, etc.) I crafted a UI that supports sharable, semi-sharable, and non-sharable content.
A magical experience - no login required
We launched Eko in the App Store in 2015 with the enthusiasm of our team and our clinical ambassadors, yet within a year our KPIs pointed to a big problem - a sharp drop off in utilization during login.
iMedicalApp would later clarify this problem in its 2016 review:
“In a sense [the Eko mobile app] is much easier compared to competitors, whose products usually require downloading pre-recorded sounds to a computer. However I would not say that it is as seamless or effortless as described. At the initial release, it was actually prohibitively cumbersome.”
iMedicalApp review, 2016. Ouch 🤕
Our customers reinforced the value of our mobile app, and shared with us their chief design concern - that manually typing an email and password causes friction and takes critical time away from patient care. As a result, clinicians are less invested to use the powerful in-app experience, and rely instead on their listening training and memory - the very thing our digital experience is meant to improve.
Put a PIN in it
Working closely with our in-house HIPAA expert, my first approach to this problem was to remove as many clicks as possible - reducing the complexity of a full email and password to a 4-digit PIN in keeping with standards set by iOS.
Early research interviews on PIN and TouchID showed good faith, but it was clear that this solution alone would not meet customer expectations. TouchID, for example, doesn’t work with medical gloves - forcing an extra “Cancel” tap before logging in. And while PIN reduces the number of taps, it is not recognized as a secure login method by many hospitals, and is yet another number that must be remembered every time a clinician wants to access the recording screen.
Flip the script
After further HIPAA review, I proposed an alternative strategy - separating the recording screen from patient health information (PHI). This would enable placing the recording screen ahead of the login experience, so clinicians looking to capture a quick 10-second clip can do so without memorizing passwords and PINs.
And while a login-free user experience sounds simple, collaborations with the team unearthed business and security implications which would need to be addressed within the flow:
Upending the existing login experience
Although our clinicians wanted this change, they also appreciated consistency and reliance on our mobile experience as-is. It was important to leverage our clinician ambassadors not only to test our product, but also to rally support around the new feature.
Putting valuable content before login
The alternative design solution hinged on the assumption that by placing valuable information in front of the login experience, we will garner support for our product and our brand - causing downstream improvements to login and conversion rates. To hedge our bets, I worked with leadership to establish a minimum success KPI which would determine if the feature would stay or be pulled within a few weeks of release.
Random edge cases
From a technical perspective, shifting the login experience caused a number of edge cases. For example
What happens when the app is resumed from the background?
What happens when connectivity is lost between recording and patient screens?
What happens when TouchID, PIN, and Email/Password fail?
Working with our in-house HIPPA expert and our CTO, I addressed these issues in sequence to cover all potential cases.