Images are representative as designs are locked under NDA.
Safer observations and simpler shifts for busy nurses
Ward is a tablet and desktop app for ward nurses that brings observations, NEWS2 scores, alerts, and handover into one workflow. I designed the information architecture, safety flows, and key screens so nurses move from scattered paper and EPR screens to a single place for “who needs me next” and “what changed since last shift.”
My role
Product Manager (for discovery)
Product designer (end to end)
Skills used
Product discovery, service mapping, interaction design, information architecture, design systems, stakeholder facilitation, clinical safety by design, UX writing
Duration
Roughly 6 months from idea to built MVP
Nurses wasted time and took on risk because essential observations and notes lived across paper, ward booklets, and generic EPR screens, with no clear digital view of their workload.
Fragmented patient data and risk identification
Observations and NEWS2 sat on paper charts, ward booklets, and generic EPR screens
Positive ID relied on call slips and bed-board flags, not a guided flow
Nurses had to juggle charts, screens, and wristbands while under pressure
Wrong-patient observations and missed checks were a constant risk
No single view of workload or priorities
Nurses relied on mental lists and scraps of paper to decide what to do next
To find overdue obs, they jumped between charts and patient records
New or bank staff needed extra time to orient every shift
Weak support for handover and ongoing narrative
Handover lived in ward booklets, separate journal screens, and verbal updates
Notes, observations, and NEWS2 were split across different tools
Building a picture of a patient meant stitching together numbers and stories by hand
Cross shift and cross role handovers were fragile and inconsistent
Define a safe structure for ward work
The first step was to turn a one-line brief, “an app for nurses on a ward to enter results and manage workload”, into a structured flows. I mapped existing flows from paper charts, ward booklets, and Compucare screens, then grouped them into four functional areas: Reminders, My patients, Forms, and Handover.
I took early prototypes into a working group with product, engineering, and clinical stakeholders, then refined flows based on feedback and safety points.
Patient confirmation became part of the design rather than an optional habit.
The flow requires a scan of the wristband or a confirmation of demographic details through explicit checkboxes before observations or chart access.
Ward dashboard gave nurses a shared mental model
“I orient on the dashboard, then I go to Reminders or My patients to decide what to do next.”
That replaced ad hoc navigation across multiple screens and paper. Nurses now reach key tasks in one or two taps from a single starting point.
Make workload and priorities visible
To tackle workload, I focused on two views.
Reminders: a list of upcoming and overdue observations, sorted by urgency.
My patients: a ward list with two modes, “My patients” and “All patients”.
This changed the decision process from “hunt across screens and paper” to “scan one list and follow the next action”.
Reminders target the question; what is most urgent right now?
My patients targets; who am I looking after and how are they doing?
Simplify NEWS2, alerts, and narrative through focused views and AI summaries
NEWS2 data and alerts are dense. To reduce confusion, I introduced a dedicated NEWS2 dashboard per patient:
A patient banner with clear identity.
A mini NEWS2 chart with an option for full-screen when needed.
Static actions such as “Add observation”, “Adjust observation interval”, and “Escalate”.
A reverse-chronological table of observations.
Alerts (allergy, clinical, financial, general) appear in one grouped view with severity indicators, accessible from the same context, so nurses do not search across separate areas for safety-critical information.
Alongside this, I designed a concept for an AI-powered handover digest that uses existing platform summarisation features. The digest aggregates:
NEWS2 changes and current escalation state
New or changed alerts
New journal entries (notes and attachments)
Key orders and results
Outcome
Ward progressed from a one-line description and scattered requirements to a structured application that:
Enforces clinically safe and appropriate workflows
Gives nurses a single dashboard for a summary of their workflow
Provides focused views for critical data like NEWS2 and alerts
Integrates narrative and prepares for AI-driven digests
The first customer, Foscote, signed up for Ward with a subscription for around 30 named nurses and dedicated training on Ward and clinical forms.
The AI digest concept sits in the near-term roadmap and builds directly on existing summarisation features in the platform.
What I would do differently next time
Stronger quantitative baselines:
I would run time-on-task and error-rate studies on existing paper and EPR flows at the start, so later comparisons for “time saved” and “errors reduced” rely on measured baselines, not only on task modelling.
Earlier observation of real shifts:
I would schedule structured observation sessions on wards before finalising navigation, to validate the placement of Reminders and My patients against real shift rhythms and interruptions.
Tighter co-design with nurses around handover
I would run focused co-design sessions only on handover, to refine how narrative, tasks, and risk signals appear together, and to align Ward more directly with existing handover practices.