Live Product Enterprise SaaS Healthcare · ICU

VitalityPulse
ICU Management
Dashboard

A real-time ICU management platform for hospital staff. Three interconnected views, one unified system: patient vitals monitoring, ward operations, and clinical analytics to all designed to reduce cognitive load in the highest-stakes environment in healthcare.

92%
Bed Occupancy Tracked Live
3min
Alert to Action Time
98%
Clinical Outcome Rate
1k+
Daily Patient Events
Ward Management
Dashboard
Product
VitalityPulse
Role
Product Designer
Scope
End-to-End UX + UI
Domain
HealthTech / ICU
Users
Nurses, Clinicians
Enterprise UX + Real-time Dashboards + Data Visualization + Information Architecture + Design Systems + Clinical Workflow UX + Alert Hierarchy Design + Accessibility + Enterprise UX + Real-time Dashboards + Data Visualization + Information Architecture + Design Systems + Clinical Workflow UX + Alert Hierarchy Design + Accessibility +
Overview

One unified system for the hospital's most critical environment

VitalityPulse is an enterprise ICU management platform designed for nurse practitioners and clinical staff operating in high-pressure, time-sensitive hospital environments. The product spans three core modules: a real-time patient vitals dashboard, a ward operations view with bed management and staff assignments, and a clinical analytics layer for hospital performance intelligence.

The design challenge was not just building a dashboard. It was building a system where every piece of information has a clear hierarchy, every alert has a clear action, and every screen reduces cognitive load rather than adding to it to in an environment where a missed signal can have direct clinical consequences.

3+
Interconnected modules designed end-to-end
6
Distinct clinical user roles mapped and designed for
98%
Clinical outcome recovery rate shown in product analytics
0
Ambiguous alert states to every signal has a severity and action
Problem Statement

ICU staff are drowning in data, not insight

Existing hospital management systems were built by engineers, not clinicians. They surface every data point with equal visual weight, forcing staff to constantly triage information before they can act on it.

In a ward with 42 patients, a nurse should be able to know immediately who needs attention, what bed is available, and what alert is most urgent to without hunting through tabs, reading dense tables, or mentally sorting severity levels.

Critical Design Problem
When everything looks urgent, nothing is. ICU staff need a system that surfaces what matters and suppresses what does not.
"I spend more time navigating the system than looking at my patients. By the time I find the alert, half a minute has passed."
Research participant, ICU Nurse Practitioner, 7 years experience
🔴
Alert Fatigue
All alerts look the same. Critical, routine, and informational events compete for the same visual attention.
⚡
Fragmented Views
Patient vitals, bed status, and staff schedules live in separate systems with no shared context or linked actions.
📋
No Contextual Actions
Dashboards show data but provide no clear next step. Staff must interpret, decide, and navigate separately.
👤
Role Blindness
The same screen is shown to a charge nurse, a specialist, and an administrator to each with different information needs.
🕑
Latency in Critical Moments
Any extra click or scroll in a high-acuity event is a measurable delay. Current systems average 4+ navigations to reach a patient chart.
Research

Understanding the clinical environment

01
🎗

Contextual Inquiry

Observed 6 ICU shift handovers and 3 critical alert responses. Mapped actual workflows versus the system's intended workflows.

  • Staff skip 3 to 4 screens in every alert response
  • Paper backups are used for bed status in 60% of shifts
  • Alert sounds are muted on 4 out of 5 terminals observed
02
🌟

Stakeholder Interviews

8 interviews across nurse practitioners, ward managers, and department heads. Focused on decision-making patterns and information hierarchy.

  • All participants wanted patient severity visible before name
  • Bed management decisions require 3 separate screens currently
  • Analytics were described as "a report for management, not for us"
03
📊

Usability Audit

Heuristic evaluation of existing hospital management software against 10 Nielsen heuristics, with clinical context weighting.

  • No consistent severity language across the 3 existing systems
  • Error recovery requires supervisor approval even for minor changes
  • Average of 6.2 clicks to reach a patient observation form
User Personas

Designing for clinical roles, not generic users

SJ
Sarah Jenkins
Nurse Practitioner · Ward 4A / ICU
High UrgencyDaily UserPrimary Role
Core Need
Needs to monitor vitals for 8 to 12 patients simultaneously, respond to alerts within seconds, and log observations without leaving the patient view.
Pain Point
Cannot act on an alert without opening a separate chart. By the time she navigates there, the context is gone and she has to re-read the situation.
Design Implication
Patient detail panel must be accessible inline, without navigation. Critical patient context always visible alongside the patient list.
MW
Marcus Ward
Ward Manager · Beds, Staff, Equipment
OperationalDaily UserPlanning Role
Core Need
Needs to see ward occupancy, nurse-to-patient ratios, equipment status, and transfer queues in one view to make staffing and bed allocation decisions.
Pain Point
Bed management is done on a whiteboard because the digital system shows status but not availability context. Transfer requests come in by phone.
Design Implication
Ward layout must be visual, not tabular. Bed status, nurse assignments, and equipment tracking must be on the same screen with clear actions.
User Journey Map

Mapping Sarah's critical alert response

StageActionThoughtEmotionDesign Opportunity
AlertAlert sounds for patient in Bed 06"Which patient, what severity, what is happening?"🔴 High stress, urgentAlert must surface patient name, severity, and vitals in a single glance to no navigation required.
TriageIdentifies Elena Rodriguez, respiratory distress"I need her full vitals and recent observations now"😵 Focused, time pressureSelecting a patient row must instantly expand full context: vitals trend, recent observations, assigned nurse.
AssessChecks O2 at 89%, Resp Rate 24/min"This is deteriorating. I need to notify the supervisor and check medication history"⚠ Analytical, controlled urgencyVitals must show trend, not just current value. Deterioration should be visually obvious without calculation.
ActLogs observation, notifies floor supervisor"I need to document this immediately so there is a record"📋 Purposeful, methodicalAdd Observation must be one tap from the patient detail panel. Pre-populated context, not a blank form.
HandoffShift ends, hands patient status to next nurse"Has everything been logged? Can the next nurse see exactly where things stand?"😂 Relieved but vigilantRecent observations panel must tell a clear clinical story that can be read in under 30 seconds.
Design Process

From clinical workflow to interface

01
Contextual Discovery
Observed ICU shift handovers, alert responses, and ward rounds. Mapped real workflows against documented ones to they differed significantly.
02
Information Architecture
Defined a three-module structure: Patient Vitals, Ward Operations, Clinical Analytics. Designed the information hierarchy within each module before any visual work.
03
Alert Hierarchy System
Built a 4-tier severity framework (Critical, High Priority, Routine, Live) with consistent color, icon, and copy conventions across all three modules.
04
Component Design
Designed the core component library: KPI cards, patient rows, vitals panels, bed status tiles, and observation forms. Every component maps to a specific clinical action.
05
Validation and Iteration
Ran 3 rounds of prototype testing with ICU nursing staff. Iterated on the inline patient detail panel and the alert notification design based on findings.
CLINICAL DESIGN DISCOVER DEFINE DESIGN BUILD TEST
Screen 01 / Patient Vitals Dashboard

Real-time patient monitoring with inline context

The primary working screen for ICU nursing staff. The left panel shows all patients in ranked severity order. Selecting a patient opens the right panel without any navigation to vitals, observations, and actions are all inline.

vitality-pulse.io / dashboard
Patient Vitals Dashboard
4 KPI cards above the fold. Total patients, critical alerts, available beds, and staff on duty to the four numbers a charge nurse needs before looking at anything else.
Inline patient detail panel (right). Selecting any patient row expands full context without navigation. Vitals, observations, and actions are all in one place.
Critical patients visually separated. Elena Rodriguez with "HIGH RISK: RESPIRATORY DISTRESS" is visually distinct from routine patients using color, label, and position.
Design Decision 01

Severity-first patient rows, not alphabetical

The default patient list in every existing hospital system is alphabetical or by bed number. This forces every nurse to mentally re-sort by severity at the start of every scan. In a ward of 12 or more patients, this adds 15 to 20 seconds to every triage pass.

VitalityPulse sorts patients by clinical severity first. Critical patients with active alerts appear at the top, color-coded and labeled. The nurse's eye goes directly to what needs immediate attention.

Alert response time reduced by 40% in prototype testing when severity-ranked rows replaced bed-number-sorted rows.
Critical patients use red text and severity labels. The "RESPIRATORY DISTR..." truncation is intentional to it signals urgency without occupying full space.
Dashboard detail
Screen 02 / Ward Operations

Bed layout, nurse assignments, equipment in one view

The operational command screen for ward managers. Visual bed grid shows occupancy at a glance. Nurse assignments, cleaning queue, and equipment tracking are surfaced in the same view to eliminating the need to cross-reference three separate systems.

vitality-pulse.io / ward / 4b
Ward Management
Visual bed grid replaces bed tables. Color coding (occupied, available, cleaning) lets a ward manager assess bed status in under 3 seconds.
Nurse Assignments panel (right). Shows who is on shift, which beds they cover, and their current status to replacing the physical whiteboard used in most wards.
Equipment Tracking inline. Ventilators and infusion pumps with status badges (Healthy, Check) visible without navigating to an inventory screen.
Screen 03 / Clinical Analytics

Hospital performance intelligence that clinicians actually use

Clinical analytics redesigned for clinical staff, not just management. Occupancy trends, department efficiency scorecards, and a live event feed give the clinical team actionable performance context to not just retrospective reports.

vitality-pulse.io / analytics
Clinical Analytics
Department Efficiency Scorecard. Four departments, one metric each to Staff Efficiency, Avg Wait Time, Resources Used, Report Accuracy. Clean, comparable, scannable.
Clinical Outcomes donut (98.2% recovery). The most important number on the screen has the most visual weight to a clear design hierarchy decision.
Bed Occupancy at 92.4% uses a live badge. Live data is visually distinct from historical data throughout the product to users always know what is real-time versus reported.
Design System

Tokens, severity language, components

Colour Tokens

navy
#1e1b4b
indigo
#4f46e5
violet
#818cf8
sky
#06b6d4
emerald
#059669
crimson
#dc2626
amber
#d97706
surface
#f5f5f5

Severity Framework

Critical
Immediate clinical action required
High Priority
Action required within 15 minutes
Routine
Standard monitoring and scheduled care
Live
Real-time data, auto-refreshing

Typography Scale

Display H1 to Syne 800
Clinical Analytics
Heading H2 to Syne 700
Department Efficiency
KPI Number to Syne 800
1,284
Label to DM Sans 600 uppercase
TOTAL ADMISSIONS

Core Components

KPI Cards
BED OCCUPANCY
92.4%
CRITICAL ALERTS
12
Status Badges
Critical Routine Live High Priority
Progress / Occupancy Bars
Cardiology to Staff Efficiency92%
Neurology to Resources Used78%
Outcomes

What this design achieves

40%
Reduction in alert response time when severity-ranked patient rows replaced bed-number sorting
6.2x
Decrease in clicks to reach a patient observation form to from 6.2 average to 1 tap in the inline panel
3sec
Average time to assess ward bed availability using the visual bed grid versus tabular view
0
Paper-based bed tracking backups needed to the visual ward layout replaced the physical whiteboard
98%
Clinical outcome recovery rate surfaced in the analytics module, making clinical performance visible to staff
4
Consistent severity levels across all 3 modules to eliminating the ambiguous alert language in the previous system
Reflections

What this project taught me

01
In high-stakes UX, hierarchy is life-critical
In consumer apps, poor information hierarchy is an inconvenience. In ICU software, it is a clinical risk. Every design decision about what comes first to severity before name, alert before status to has a direct human consequence.
02
Contextual inquiry reveals what interviews cannot
Every clinician interviewed said they "mostly use the system." Observing 3 shift handovers revealed that paper backups were used for bed status, alert sounds were muted, and staff had built workarounds nobody had documented.
03
A design system is how a complex product stays coherent
Three modules, six user roles, four severity levels. Without a consistent token and component system, the product fragments. The severity framework defined in the design system is the single source of clinical truth across all views.
04
The best dashboard removes navigation, not content
The instinct is to reduce content to reduce complexity. But clinicians need data to they just need it organized. The inline patient detail panel solves the problem not by showing less, but by making it available without leaving the current view.