Overview

Mitigating Patient Safety Risks

Alarm fatigue is an increase in a healthcare provider’s response time to an alarm as a result of experiencing excessive alarms, potentially risking patient safety. These alarms go off when, say, a patient's heart rate is lower than normal.

NurseWav is a one-of-its-kind mobile app built to help nurses tackle alarm fatigue in hospitals by moderating and organizing the influx of excessive alarms. This three-month project was a research-heavy, cross-functional effort with 4 others as part of the Patient Safety Technology Challenge, pitched to Arizona State University’s eSeed Challenge for funding.

The work involved a two-month strategic research deep dive followed by a month of solution ideation. As part of the challenge, we researched existing issues in the medical field under the guidance of our mentor, Ariana Longley, a patient safety expert, and chose to solve alarm fatigue based on the results and our personal experiences.

Challenges

• Excessive number of false alarms.
• Difficult to differentiate between actionable and non-actionable alarms.
• Too many trips to the patient's bedside to manage alarms.
• Existing solutions have complex/cluttered interfaces and face challenges integrating with existing hospital systems.

Goals

• Reduce false alarms by organizing and filtering them out.
• Diminish the need for frequent trips to the patient's bedside whenever an alarm activates.
• Ensure the solution seamlessly integrates with existing hospital systems.
• As always, make the solution simple and user-friendly.

Role

Product Design Lead

My responsibilities included leading the branding strategy for market viability (pitch deck) and driving the high-fidelity prototype design in Figma. It also included defining the user personas and patient safety workflows, and documenting the core solution architecture with the team.

Design Process

From Research Deep Dive
to Prototype

• Conducting research by sending out user surveys and performing user interviews.
• Ideating the solution — figure out a foolproof alarm management flow.
• Designing wireframes.
• Designing a high-fidelity interactive prototype.

Empathize

Understanding Alarm Fatigue

Our research unfolded in two stages — the first phase was researching the broader problem area. We began by addressing questions like which problem category we want to tackle: Diagnostic errors? Infections? Surgery-related issues? And so on. We read several research papers and articles on the internet, discussed our personal experiences, brainstormed with our mentors, and chose to address alarm fatigue as a result.

Show less

We received 24 responses for the survey, with the majority of participants aged over 40, boasting a minimum of 20 years of experience in the medical field, and all participants being familiar with alarm fatigue:

• Majority of participants encounter over 40 alarms in a typical shift.
• Vital sign alarms are triggered the most often.
• Majority of the participants face difficulty in distinguishing between false and true alarms, and are frustrated with the lack of customization options for non-voluntary alarms.

We analyzed all of the interview and survey data in a fishbone diagram and created two user personas and a journey map to understand the highs and lows of our potential users.This research data resulted in us narrowing down our problem area a little more and giving us a more concise problem statement, i.e., we now started looking at solutions targeting nurses and involuntary alarms in less-critical post-surgery units.

Ideate

Defining a Precise,
High-Impact Solution Scope

Why a mobile app?

This form factor enables nurses to manage alarms remotely, and thus, diminishes frequent, non-actionable trips to the patient's bedside whenever an alarm activates. Additionally, it's easier to integrate with existing hospital infrastructure compared to a tangible product since nurses already have hospital-provided mobile devices, making it simply a matter of installing an application on them.

We conducted a rapid task analysis of the features we wanted our app to have — Nielsen’s ten heuristics helped us out a lot here. With this information mapped out, I set out to design the screens.

Design & Prototype

Designing for Criticality & Clarity

I started the branding process by building out a style guide for the designs -

• Color palette: The color palette I chose was WCAG compliant and included a dark matte for the background with bright, contrasting colors for alerts, adhering to Gestalt's law of figure-ground for grabbing attention. I added hints of green to primary components to symbolize health, and eye-catching orange/yellow shades for various levels of critical alerts to ensure immediate cognitive differentiation and quick reaction time.
• Typography: I focused on a clear, easy-to-read font with adequate size, contrast, and spacing to enhance readability even under pressure.
• Screen layout: I opted for a clean, sleek, and minimalistic layout with clear hierarchy and ample white space, minimizing visual clutter and allowing nurses to scan information quickly and efficiently. Gestalt’s law of good figure helped me out here.
• Touch dimensions: I chose appropriate touch dimensions, keeping Fitts' law and the Aesthetic-usability effect in mind, i.e., both usable enough and visually appealing.
• Accessibility: The color shades and typeface I chose are all WCAG AAA-compliant and don't contrast too much with the background to not strain the eyes of users.

The 3 Core Flows

Critical Patient Metrics Simplified

Design Goal: Provide a scannable overview of key patient data for quick assessment.

Key Design Elements:

• Key information in a scannable list format.
• Prioritization of the three most vital signs based on user research (surveys and interviews), which are customizable.
• Filters and sorting options for empowering nurses to personalize their view and access specific data efficiently.
• Prominent alerts with three clear interaction options, based on Hick's law, for immediate action.
• Highlighted vital(s) causing the alert for user clarity.
• Navigation menu at the bottom to meet mobile users' expectations.

User Impact: Nurses can quickly assess multiple patients, focus on critical information, personalize their view for efficient data retrieval, and easily manage alerts with minimal cognitive load.

Vitals View

Alert Interaction Options

Comprehensive Patient Vitals at a Glance

Design Goal: Provide in-depth vital details of a patient, allowing remote monitoring of all the key information.

Key Design Elements:

• Display of all vitals and monitor readings for a comprehensive overview.
• Consistent alert handling with the same interaction options for efficient management.
• Accessible basic patient information like surgery type and date of birth for convenience.

User Impact: Nurses can quickly view all vitals and basic patient details without needing to go anywhere else in a seamless manner.

Detailed Vitals View

With Alert

Patient Details Overlay

Empowering Decision-Making with Actionable Data

Design Goal: Enable informed decision-making with detailed nurse shift and alert statistics.

Key Design Elements:

• Card layout with concise, quick stats for fast information scanning.
• Nurses can analyze detailed vitals, understand alert patterns, and make informed decisions based on comprehensive data visualization.
• Increased font size for key statistics for improved readability.

User Impact: Nurses can analyze their workload, understand alert patterns, and make informed decisions based on comprehensive data visualization.

Reflection & Next Steps

End of a Journey

This was an extremely fun and challenging project, given how big the initial scope was. It solidified my understanding of user-centered design by directly translating research findings into actionable design decisions, thus strengthening my core UX skills.

• The ROI of research: The two-month, two-stage research phase was critical for risk mitigation, allowing us to identify a precise problem (less-critical post-surgery units) that maximized the potential for organizational adoption and minimized implementation resistance.
• Designing for data abstraction:  In a data-dense domain, the primary design challenge was abstracting critical, real-time vital signs into simple, scannable, and actionable mobile alerts that adhere to WCAG standards.
• Strategy for cross-functional alignment: In a tight, three-month team environment, we found that establishing a clear process for managing our timeline and deliverables was essential. We leveraged tools like Airtable to set deadlines, track research interviews and findings, and Google Drive to manage all our assets. This organized approach wasn't just about deadlines; it was about ensuring our cross-functional handoff was smooth and our technical feasibility was secured from the start.

The next steps would be to complete designing the high-fidelity prototype, test it, implement the application, and develop it further with the funding received, with a core objective to reduce the number of clinically insignificant alarms presented to the nurse by 75% and eliminate 50% of unnecessary bedside verification trips in the first six months after launch.