Designing a care plan system for 200+ mental health providers
Less than half of patients had documented care goals — not because providers didn't care, but because the tools made good documentation harder than bad. I spent six months fixing that, coordinating across clinical, operations, engineering, and product to get there.
My role
Interaction design · User research · Prototyping · Information architecture · UX strategy
Collaborators
Engineering, Product, Clinical leadership, Operations
Tools
Figma, FigJam, CodePen
Patient goal documentation rate, 3 months post-launch
Mental health providers on the launched system
of Excellence award for prioritizing user needs
01 — The problem
Good documentation was harder than bad documentation
Before the care plan existed, Aptihealth providers tracked patient goals however they could — mostly in external spreadsheets, sometimes in a free-text field, often not at all. The platform gave them no structure, no guidance, and no way to see what other providers on the same patient's team were working on.
The result: less than 45% of patients had documented goals in the platform. That wasn't a provider problem. It was a tool problem.
Inconsistent care standardsProviders documented goals in their own formats with no clinical guidance — making it impossible to measure or compare outcomes across the network.
No visibility across the care teamProviders couldn't see who else was working with a patient or what goals they were pursuing. Care was fragmented by design.
No data for outcomes reportingWithout structured goal-tracking, it was nearly impossible to demonstrate the platform's effectiveness to insurance companies.
Friction in daily workflowsProviders manually reviewed past session notes to assess patient progress — adding unnecessary overhead to every visit.
02 — Provider research
Four provider types. Four different definitions of useful.
Aptihealth has four distinct provider types, each playing a different role in patient care. A behavioral health specialist documents goals differently than a prescriber. A case manager needs different information visible than a peer support specialist. The care plan had to work for all of them — which meant it had to be designed for each of them.
Behavioral Health Specialist
Primary goal-setters. Needed structured frameworks for documenting patient objectives and tracking progress over time.
Prescriber
Focused on medical context. Needed visibility into care team goals without the burden of maintaining them.
Case Manager
Coordination-focused. Needed a full picture of who was involved in a patient's care and what each person was working on.
Peer Support Specialist
Support-focused. Needed simple, readable goal summaries without clinical complexity in the way.
I interviewed representatives from each group, then synthesized the findings into an affinity map shared with engineering and product. The goal was to get everyone aligned on provider needs before a single design decision was made.
Affinity mapping session — key themes and pain points organized across provider types
03 — The hard part
What clinical leadership wanted vs. what providers actually needed
The most challenging part of this project wasn't the design — it was navigating the gap between what clinical leadership believed the care plan should contain and what providers told us they actually cared about in practice.
Clinical best practice and provider workflow aren't always the same thing. My job was to find a design that honored both — not to pick a side.
Clinical leadership had strong opinions about what needed to be captured for compliance and outcomes reporting. Providers had equally strong opinions about what would actually fit into their workflow without becoming another documentation burden. These weren't always compatible.
The solution was incremental. Rather than designing the entire care plan upfront and hoping it worked, we broke it into phases — designing, testing, and refining each section before moving to the next. This gave us a feedback loop that let provider reality inform clinical requirements, not override them.
04 — System design
Every section had its own permissions. Every role saw something different.
Designing the information architecture required mapping out who could see what, who could edit what, and what happened when a provider tried to do something outside their permissions — across four provider types and multiple care plan sections.
The core structure we landed on included: patient demographics, care team visibility, care goals, patient-articulated goals, patient strengths, medical cautions, and a summary of the patient's initial clinical assessment. Each section had distinct editing permissions depending on provider role.
Example workflow for the goals section, showing state transitions and permission boundaries
The goals section was the most complex — structured enough to guide providers toward clinical best practices, flexible enough to allow for personalized care. Dropdown-based goal selection with a free-text fallback ended up being the right balance: it gave providers a starting point without feeling prescriptive.
05 — Interaction design
Three flows that carried the most design weight
The care plan had many interactions, but these three required the most deliberate design work — each one encoding a clinical decision into an interface pattern.
Adding a goal: dropdown-first selection guides providers toward standardized language while the free-text fallback preserves clinical flexibility. The structured format was a direct response to the inconsistency problem — if every provider used their own words, outcomes data was useless.
Editing vs. achieving vs. removing: three distinct actions that look similar but mean very different things clinically. A completed goal is a data point. A removed goal is a correction. Conflating them would have corrupted outcomes reporting — so the UI had to make the distinction explicit without adding cognitive load mid-session.
Provider-specific details: treatment modality, appointment frequency, session duration. This section directly solved the care team visibility problem — by surfacing each provider's approach in a structured format, the care plan became the single source of truth for how a patient was being treated across the full team.
06 — Prototyping
I built it in the browser, not just in Figma
Once the MVP design was finalized, I created interactive prototypes in CodePen to demonstrate responsive behaviors and edge cases that static Figma files couldn't capture. This let engineering see exact interaction behavior before implementation — and it cut back-and-forth significantly.
Handing off a working prototype instead of static specs changed the quality of the engineering conversation entirely. It's just how I prefer to work — closer to the medium than most designers get.
For example, the Patient Strengths section required a responsive two-column checkbox layout that collapsed into a scrollable list on mobile. Rather than annotating this in Figma, I built it directly in CodePen — complete with the full design token system, hover states, and the "Other" free-text edge case. Engineering had nothing to interpret.
View the prototype →
The pilot ran with 20 providers before the full rollout. I led affinity mapping sessions during the pilot to synthesize feedback, then worked with the team to prioritize changes before launching to the full network.
The launched care plan — structured goal documentation with care team visibility
07 — Outcome
Providers liked it. The numbers showed it.
The care plan launched to 200+ providers and within three months, patient goal documentation went from 43% to 77%. Improved care team visibility was measured through a reduction in cross-provider emails. And for the first time, Aptihealth had structured outcome data it could use in insurance conversations.
The Employee of Excellence award was a nice acknowledgment — but the provider feedback was the part that stuck.
Thank you for the new care plan and the tools to support clinical excellence.
Aptihealth provider
I love the new care plan! Thank you for your time and for the upgrade.
Aptihealth provider
The care plan is very user friendly and the drop downs in the goals section are helpful and clear.
Aptihealth provider
I've found that the care plan is very easy to use — thank you!
Aptihealth provider
If I were to do it again, I'd push harder earlier to get provider representatives into the room with clinical leadership — not just in separate research sessions. The tension between clinical requirements and provider needs was real and took time to navigate. Having both sides in the same conversation sooner would have accelerated alignment.