Schedule Solver
A case study about a nurse scheduling solution.
This application helps users manage the creation of shift schedules. The previous experience was fragmented and inconsistent, relying mostly on manual edits after printing the schedule on paper.
As one of the designers for the nurse manager experience, I collaborated closely with PMs and developers to create a scheduling system by gathering user research insights and expanding our existing design system. I joined in spearheading this design-driven initiative to create a more seamless application, setting the foundation framework for future applications in the Commure OS.
Role
Product Designer
Responsibilities
UX/UI, prototyping, research
Timeline
2022
Problem
HCA needed a way to transform its scheduling and staffing process to improve employee satisfaction, provide equitable and balanced schedules, and have the ability to match prospective patient volume to staff.
They already had developed a staffing optimization algorithm, so it was up to us to operationalize it via a new UI.
Solution
The Commure Scheduling App would seamlessly integrate into their existing system as a user-friendly interface for head nurses to manage schedules across different facilities or departments. The various data points, including facility scheduler data, live and forecasted patient volume, nurse schedule preferences, and nurse talent profiles, would feed into their staffing optimization algorithm, HCA Schedule Solver. The Schedule Solver would generate proposed schedules and staffing data, and our solution would enable adjustments to these schedules. Final schedules would then be stored in the Facility Scheduler App for nurses to view.
Research & Discovery
This project was accompanied by a clearly defined set of requirements and expectations. However, we also prioritized collecting firsthand insights. Opportunities were provided for team members to visit various facilities, observe nurses in their work environments, and gain a direct understanding of their needs as end-users of our product. Our objectives included observing their existing scheduling procedures, identifying key pain points, and analyzing their temporary solutions to utilize the staffing optimization algorithm.
Once we had a first iteration of several core workflows, we conducted two design feedback sessions at different HCA facilities. Nurses participated by dropping in throughout the day to provide comments. The participants were very engaged and expressed enthusiasm for the ongoing efforts to enhance their scheduling process. They showed satisfaction with the current trajectory of our solution, with the most common requests being modifications to how staffing issues are communicated, a desire for a more comprehensive overview of all shifts, and an easier method for matching nurses with the requisite skills to open shifts. Their handwritten note suggestions were effectively translated into well-defined objectives for subsequent iterations.
highlight week they are under and show options for best shift
cancel / call off reasons drop down
also search other nurses on roster
important for shifts to be balanced - different skill levels
MVP
#1 Show a more comprehensive display of shifts
Initially, shifts were displayed using the AM and PM labels; however, due to overlaps in nursing shifts and multiple start times throughout the day, a more effective way was needed to show staff gaps and surpluses during specific shift periods. This was accomplished by adding a collapsible section that lists each shift window, showing the targeted staff versus the scheduled staff, with the difference represented in each table cell. This calculation was previously part of the client's informal solution, ensuring that presenting information this way would be familiar to our users.
#2 Update shift error indicators
In our initial iteration, we highlighted the date of each imbalanced shift, prioritizing understaffed shifts when an AM shift had too few employees and a PM shift had too many. After our feedback sessions, we learned from nurse schedulers that understaffing in shifts is common, and highlighting every day in red wouldn't be helpful. What would be most useful for our users is to identify overstaffed shifts so they can reallocate those nurses.
We tried to solve this by changing the alert hierarchy to prioritize overstaffed shifts, but found this still caused a lot of cognitive load and stress for our users. Seeing a newly completed schedule already filled with alerts and icons indicating problems isn't a pleasant experience. Instead, we decided to adjust the summary periods accordion to include the number of nurses needed along with smaller indicators that show if each window is over- or understaffed. This way, schedulers can decide to look for staffing information and quickly determine which shift windows to address first.
Original indicators highlighting dates
Second iteration within the summary period section
Final iteration using both colors and icons as indicators
#3 Match needed skill sets
The last adjustment we made was to the shift management drawer, which went through several rounds of iterations until it met the user's needs and expectations. It was crucial that we received regular feedback from the nurses at our target pilot site after each iteration to ensure we delivered a product that would be a significant improvement to their current workflow.
In establishing our MVP, we needed to make a drawer component that was able to:
Assign additional nurses to an understaffed shift
Cancel surplus nurses from an overstaffed shift
Ensure that nurses scheduled had the skills required for a shift
First Iteration
Second iteration including skills
The first change we made to the drawer component was to add the ability to match nurses with the needed skill set to an open shift.
We achieved this by adding an alert element to the drawer that displays the required roles for that shift and lists available staff by role underneath.
The final version of the drawer included further improvements to its functions. We removed the alert at the top to cut down on visual clutter and instead added drop-downs by role needed, with a tag indicating how many staff members must be selected to balance the shift.
We replaced checkboxes next to available nurses with a searchable dropdown that lists all appropriately skilled employees. Each employee's name is accompanied by a tag indicating the nurse's availability to the scheduler.
Final iteration for MVP
Reflections
This project provided many opportunities for me to learn and grow as a designer. I was able to work closely with various roles such as UX researchers, product managers, and developers. I am grateful that our team had the chance to observe users in their normal work environment and collect direct feedback from them on our ideas. It is rewarding to know that our solution helped an HCA facility adequately staff their nurses, enabling them to deliver better quality care to patients.
This solution went live in one HCA facility before the project was put on hold. I would have loved to continue working on bringing other requested features to life and seeing this app go live in more locations.
Other Projects
