Cannes, France – Every 73 days the amount of patient data doubles. Most of that data (80% of it!) is invisible. That means it’s not easily read by digital systems, including video, images, audio files, clinician notes. All the things that computers are really bad at reading and interpreting.
Using the data we can read and cracking into what we can’t read requires cognitive computing. Gorham Palmer, Executive Creative Director and Creative Practice Lead for IBM Interactive Experience, came to Lions Health to tell us about some of their newest systems and the role of the designer in large system design.
Let’s start with their innovations across the continuum of care. Three took us across an incredibly sophisticated amalgam of individual and population data with a careful pairing of information and decision support:
- IBM Population Manager: Creates retrospective analyses to discover patients across a large population who might have symptoms or signs that indicate a treatment option or question that might have been missed
- IBM Care Manager: Aggregates a complete visual view of the patient, supports clinicians showing patients a treatment path, and suggests next best questions.
- The Distiller: Currently in development, this view of the patient is based on indicators from structured and unstructured data, shown on tiles and cards. A click or tap reveals both the detail and providence of each data point. It also has a tab that suggests most likely diagnoses by degree of confidence. One more click shows the evidence of each hypothetical diagnosis and suggests additional clinical questions. The Distiller can also show population treatment outcomes and potential clinical trials that could be a good fit for both patient and disease.
To design these systems, IBM has had to explore every element of the experience with users. One issue: information density in these elaborate and deep profiles of individual patients. They tried to simplify the system to lighten the visual load, but doctors say they like the density. One said, “I have to consume tons of data very fast. I’m very comfortable flipping through it quickly.”
Stepping back, Palmer said there are four questions to ask when thinking about cognitive computing from a designers perspective:
- Is it broad and shallow? Or deep and narrow? Is it a generalist you can ask for directions or a sports score? Or a subject matter expert that can answer incredibly detailed question on one topic.
- How much should we show we know? (i.e. controlling for the creepiness factor)
- Is there a visual interface? Or is it more like Siri?
- What’s its persona? What’s the personality and voice? How do you write for it?
The role of designer is creating human-centered experiences in things that could otherwise be simply objective. If you trace the trail of intent back, there are always business rules at the other end. Designers challenge those rules against user needs.
“Traditionally designers have been creators of beautiful artifacts,” said Palmer. “But the more important role is that of design thinker applying design to the problems that need to be solved.”
Design thinking isn’t a new idea but IBM has customized it for the big problems they take on:
- Understand people’s needs with robust user research
- Form intent
- Deliver outcomes at speed and scale
See lots more at IBM’s open source site for design thinking.
