[Case study] Kohde for Lyra Southern Africa

The system behind an executive health assessment day

Executive wellbeing system
Live since mid-2024
2024 to present

Lyra Southern Africa, formerly ICAS, supports employee wellbeing across hundreds of organisations. Their executive offering puts a senior leader through a full day of clinical assessment and hands them a report at the end of it. The day itself ran on people, paper and coordination. My job was to design the system that holds it together.

Client: Lyra Southern Africa, formerly ICAS
My role: Sole designer, analysis through QA
Timeline: January to March 2024, support ongoing
Status: Live since mid-2024

Context

An executive assessment is not a doctor's appointment. A senior leader at a large employer, a bank for example, is invited as part of their package. Before they arrive they complete questionnaires covering physical wellbeing, lifestyle, nutrition, physical activity, psycho-emotional wellbeing, sleep and fatigue. That gives the medical team a picture of the person before the person walks in.

On the day they move through standard pathology, clinical assessments covering cardiac and respiratory measures, a functional assessment and a GP consultation, with optional services alongside: a nutritionist, a biokineticist, a life coach, a massage therapist. At the end they receive a report on their physical and mental health with recommendations, sometimes including a specialist referral. Several professionals touch the same executive on the same day, none of them able to wait on another to finish first.

The problem

Underneath the clinical detail, the assessment day is really a coordination problem. Every person in the building needs the same two answers at any moment: where is this executive up to, and what is still outstanding. Without a shared view, that answer lives in whoever happens to be holding the file.

The real design question was how to keep a day with many moving parts legible to everyone running it, without slowing any of them down.

Two constraints shaped everything: the clinical content is not negotiable, and the day is interruptible, because an executive gets called away or a machine is in use and a module is picked up later.

My role and process

I was the only designer on this. I ran it from first conversation through to design QA during the build, which meant the analysis and the interface were never handed between people.

On site with the client

I spent time at Lyra's offices rather than working from a brief. I watched how the assessment process was meant to run and asked the people who run it what actually happens, which is where the interruptions, the skipped modules and the waiting surfaced.

Requirements and functional spec

I ran the requirements elicitation and wrote the functional specification, so the clinical content, the roles and the rules were settled in writing before anything was drawn.

Journeys, then low and high fidelity

I mapped the journey from the executive's invitation through to their report, then worked through low fidelity into high fidelity, reviewing with both the clinical staff who would use it and the business who commissioned it.

Handover and QA

I ran the handover into development and stayed on through the build for design QA, so what shipped matched what was specified.

Low fidelity

The wireframes settled structure before anything else. Three questions got answered here: how a nurse finds the right appointment, how the day's modules are grouped, and how a clinical form behaves when it is long.

Low fidelity wireframe of the module grid — **Modules as a grid.** Each assessment is a card, so a module can be entered in any order rather than as a fixed sequence.

Low fidelity wireframe of a clinical form — **Form structure.** Grouped sections with their own headings, mixing free entry and single-choice fields, established before the clinical content went in.

The module dashboard

This is the screen the day runs on. Every assessment for one executive, grouped into Health and Lifestyle, Core assessments and Optional assessments, each carrying its own state.

High fidelity module dashboard — **Appointment modules, high fidelity.** Status sits on every card as Complete, In Progress or Not started. Optional services carry their own opt-in control, so a nutritionist or biokineticist is added to the day without leaving the screen. Summary of findings and Report sit in the header, since they are the outputs of everything below them.

Decisions worth naming

Status on every card, not a progress bar. A single percentage would say the day is 60% done without saying which room to walk to next. Per-module state answers the question staff actually ask.
Three groups, not one list. Questionnaires the executive completes, core clinical assessments, and optional services are different in kind, funded differently and owned by different people. Flattening them would have hidden that.
In Progress is a first-class state. The day is interruptible by nature, so a module that is open and unfinished had to be as visible as one that is done.
Opt-in on the card itself. Optional services are sold and chosen, so the decision point lives where the service is described rather than in a separate booking flow.

Capturing clinical data

The clinical assessment form is where the design stays out of the way. The content comes from clinical practice, so the work was in everything around it.

Decisions in the clinical form

A flag against each measure. Every reading can be flagged, so a nurse marks something for attention as they capture it rather than remembering it for the report later. The flag is the bridge between capture and the findings the executive eventually reads.
Save progress sits next to Done. Half a form is a normal state here, not an error. Making the partial save equal in weight to completion matches how the day actually runs.
Sections carry their own numbering. Clinical staff refer to measures by name and order, so the form follows their sequence instead of being reorganised for visual balance.
Comment fields under classified results. Where a result is a category, normal, restrictive, obstructive or mixed impairment, a comment sits directly beneath it, because the category alone loses the nuance a clinician needs to record.

High fidelity clinical assessment form — **Clinical assessments, incomplete state.** Blood pressure sitting and lying, body measurements, body fat and BMI, ECG, then spirometry with FVC, FEV1, the ratio, peak flow, CXR and predicted VO2 max.

Outcome

LiveIn production since mid-2024, running continuously
100+Executive assessments a month, confirmed with the client
ThousandsAssessments completed on the platform since launch

The platform has been in continuous use since it went live, pausing only over the holidays. I still support Lyra on it today, which means the decisions above have been tested by two years of real assessment days rather than by a review meeting.

Reflection

If I picked this up again, I would go after the steps it takes to get an executive from arrival to their wellness report. There is still a lot of back and forth between modules. I would streamline that into a role-based stepper, so a clinical professional is guided through their part of the assessment to completion instead of choosing where to go next each time.