Designing a clinical tool that tells you what to do, not just what to know.

Zenara's Assess module surfaces an average of 8–10 clinically relevant findings per patient, compared to a standard benchmark of 2.8. That's not a rounding error - it's a fundamentally different picture of a patient. But a clinician doesn't have an hour to review it. They have minutes, often right before walking into the room. The design problem wasn't how to display all of that information - it was how to build a hierarchy that tells a clinician, without them having to hunt, what requires a decision right now versus what is important context to have absorbed.

Not all clinical information has the same job. Some findings require an immediate decision - a flag that changes how the clinician opens the conversation, a risk factor that affects the treatment plan. Mixing those with context-only findings in a flat list is one of the most common failure modes in clinical tools - it creates false equivalence and makes clinicians do the triage themselves, every time. The dashboard I designed makes the distinction structural: actionable items have different visual weight, position, and interaction patterns than informational ones, a principle that carried through every module.

Assess and Monitor have fundamentally different jobs - Assess is dense and scannable for the minutes before a visit, Monitor tracks outcomes over time and makes PHQ-9 and GAD-7 trends readable rather than buried in a spreadsheet. The design system had to serve both: shared tokens and components that adapt across different information densities and interaction patterns, one coherent product.