How to Create a Fair Residency Call Schedule (Without Losing Your Mind)
Key Takeaways
Manual scheduling fails to create provably fair outcomes because it can't mathematically balance complex variables like shift timing, difficulty, and physician preferences at scale.
The resulting perception of unfairness erodes team trust, accelerates burnout, and generates a constant stream of change requests and complaints for chief residents.
The solution is a shift from manual balancing to mathematical optimization, which uses data to produce schedules that are demonstrably fair and maximize resident satisfaction.
Thrawn provides this solution as a managed service, delivering finished, optimized schedules that reduce bias and help end fairness debates.
You've gotten the angry emails. You've had the hallway conversations. You've tried explaining why the holiday schedule looks the way it does—but without objective proof, it always feels like a losing battle.
Fairness complaints are the number one source of scheduling drama in residency programs. And the frustrating truth is that no matter how many hours you pour into that spreadsheet, the schedule will never feel defensibly fair to everyone—because without mathematics behind it, it isn't.
This guide covers what physician scheduling fairness and equity actually require: how to define it, why manual balancing fails at scale, and what a mathematical approach to equitable distribution looks like in practice.
Why "Fair" Is Harder Than It Looks
Most chiefs approach fairness intuitively—counting nights, eyeballing weekends, trying to remember who worked New Year's last year. That instinct is right. The execution is the problem.
True equity in scheduling isn't just about equal shift counts. Petal Health's analysis on shift equity defines it as a fair distribution that accounts for the nature of the shift (a routine overnight vs. a high-acuity holiday), its timing (Christmas Eve vs. a random Tuesday night), and physician-specific factors like PGY level or rotation demands.
A schedule where every resident works the same number of call nights isn't necessarily fair if one resident's nights are all Friday-into-Saturday while another's are all midweek. Equity requires tracking multiple dimensions simultaneously—and that's where manual scheduling breaks down.
The consequences of perceived unfairness go beyond hurt feelings:
Eroded team cohesion. Feelings of favoritism poison group dynamics and redirect frustration toward the schedule-maker.
Accelerated burnout. Consistently assigning demanding shifts to the same residents contributes directly to the burnout patterns documented across residency programs.
Endless change requests. When residents don't trust the process, swap requests multiply. As one physician noted, "the swaps, tracking fairness, and just carrying it in your head all month… that's what really adds up."
Why Manual Balancing Always Fails at Scale
Excel is the default tool because it's flexible and free. But flexibility without optimization is just organized guessing.
Here's the structural problem: manually balancing a schedule for 15+ residents across nights, weekends, and holidays involves hundreds of interdependent variables. Change one cell—a vacation request, a sick call, a swap—and the fairness you built over hours can unravel in minutes. Chiefs describe this as rebuilding a house of cards.
Beyond the time cost, there are three failure modes that are essentially unavoidable in manual scheduling:
Unconscious bias. Without an objective system, your working memory favors certain residents when filling gaps—often the most recent name you thought of, or the one least likely to push back. This creates actual bias even when none was intended.
Unprovable fairness. A spreadsheet can count shifts, but it can't prove that the distribution was globally optimal. When a resident complains, all you have is your word. As reported in one emergency medicine residency thread, programs "have lately had problems with schedules containing duty hour violations and inequality between individual residents"—often discovered after the schedule was already published.
No institutional memory. When the chief year ends, the scheduling logic walks out the door. The next chief inherits the same spreadsheet, the same pain, and none of the hard-won context.
A Mathematical Framework for Provably Fair Schedules
The good news: operations research has solved this problem. The challenge is that most chief residents don't have the time—or the background—to implement it themselves. This mathematical approach breaks down into three concrete steps.
Step 1: Define Your Fairness Metrics
Before you can balance anything, you need to agree on what you're measuring. Programs that skip this step end up defending schedules without a shared definition of fair. At minimum, track:
Total weekend shifts per resident
Total holiday shifts worked (and which specific holidays)
Distribution of night float blocks
Variance in assignment to high-burden rotations (ICU, overnight call)
Satisfaction rate of vacation and rotation preferences
The specific thresholds matter less than the consistency. Agreeing upfront that "every resident will work between 8 and 10 weekend shifts this year" gives you a defensible target—and a number you can verify.
Step 2: Move from Rules to Optimization
Most scheduling tools—and all spreadsheets—operate on rule-based logic: they check for hard violations (Dr. Smith is on vacation, so she can't be scheduled that day) and flag conflicts. That's useful, but it doesn't produce an optimal schedule. It just produces a valid one.
True optimization, as Lightning Bolt's analysis of optimized scheduling explains, uses combinatorial methods to assess all possible configurations and find the one that best satisfies all constraints simultaneously—including fairness targets, not just hard rules.
The specific technique used in academic scheduling research is integer programming (IP). A 2024 study on the Resident-to-Rotation Assignment Problem developed an IP model specifically designed to maximize satisfied vacation requests while minimizing disparities in satisfaction among residents—essentially encoding equity as a mathematical objective rather than a manual balancing act.
Step 3: Validate Against Real Outcomes
If you're skeptical, the data is convincing. A study published in PLOS ONE evaluated AIMS, an automated scheduling tool built on these principles, in a live residency program. The results were significant:
96% of interns received their first-choice rotation with the automated schedule, compared to 69.4% with manual scheduling
80.5% of residents received their first-choice assignment, up from 30.5%
Conflicts between night and day shifts were more than halved
Resident-reported satisfaction and perceived fairness scores increased substantially
These aren't marginal gains. They're the difference between a schedule residents trust and one they argue about for months.
From Theory to a Finished Schedule
The practicality gap is real. Integer programming is the right approach—but building your own optimization model isn't a reasonable ask for a third-year resident running a residency program on the side.
This is where a managed service built on these principles becomes relevant. Thrawn, founded in 2024 by a team of mathematicians, computer scientists, and logistics experts from MIT, built its scheduling engine specifically around this problem. Its proprietary Scheduling Programming Language (SPL) is rooted in mathematical programming and operations research—the same foundational methods described above.
Thrawn's Fairness & Equity Engine treats equitable distribution as a core constraint in the optimization, not an afterthought. That means:
Mathematically provable fairness. Night, weekend, and holiday assignments are balanced across all residents by the optimization itself—not estimated by hand.
Bias removed by design. Because assignment decisions are made by the optimization engine against your defined constraints, unconscious preferences don't enter the process.
Fewer complaints. When a resident challenges the schedule, the answer isn't "trust me"—it's a distribution that can be shown to meet the equity targets your program agreed to upfront.
Critically, Thrawn operates as a done-for-you managed service. Programs don't operate scheduling software themselves—they send their constraints (rotation requirements, vacation requests, Accreditation Council for Graduate Medical Education duty hour rules, resident preferences) and receive finished schedules for review.
As Dr. R. Kapoor, a Clinical Fellow in Neurocritical Care, described the process: "We provided the team with the vacation requests of our clinical fellows and scheduling requirements for various rotations, and Thrawn quickly followed up with a couple of clarifying questions. Within such a short time, our yearly block fellowship schedule was complete!"
Dr. Kapoor noted that "scheduling can be one of the most stressful and time-consuming parts of the role, but Thrawn made the entire process seamless."
The scheduling knowledge is also retained by Thrawn across chief resident transitions—so the incoming chief doesn't inherit a cryptic spreadsheet and a steep learning curve every July.
Stop Defending Schedules, Start Approving Them
Defensible fairness in Graduate Medical Education scheduling doesn't come from working harder in a spreadsheet. It comes from using a fundamentally different method—one where equity is a mathematical objective, not a manual approximation.
The shift matters practically, too. Chief residents who stop building schedules and start reviewing them get back significant time, reduce the political friction of fairness complaints, and hand off an institutional system—not a personal spreadsheet—when their year ends.
If your program is still manually tracking holiday equity and weekend rotation fairness in Excel, Thrawn's managed scheduling service is worth a conversation. Programs at multiple top-20 academic health systems have already moved to optimization-based scheduling. Thrawn offers a personalized consultation to see if the approach fits your program's constraints and workflow.
Frequently Asked Questions
What is the main cause of scheduling unfairness?
The primary cause is manual scheduling. Spreadsheets can't mathematically balance the complex variables of shift timing, difficulty, and physician preferences at scale, leading to schedules that are unintentionally biased and difficult to defend as truly equitable for the entire team.
How does mathematical optimization create fairer schedules?
Mathematical optimization evaluates all possible schedule configurations to find the one that best satisfies all constraints simultaneously. This process treats fairness metrics—like balanced weekend and holiday shifts—as core objectives, resulting in a provably equitable distribution for all residents.
What are fairness metrics in physician scheduling?
Fairness metrics are specific, measurable targets your program agrees to track. Common examples include the total number of weekend shifts, specific holidays worked, night float blocks, assignment to high-burden rotations, and how many vacation or rotation preferences were granted per resident.
Why can't our current scheduling software guarantee fairness?
Most scheduling software is rule-based, not optimization-based. It checks for hard violations (e.g., someone is on vacation) but can't find the best possible schedule. It produces a valid schedule, not a provably fair and optimal one that balances all fairness metrics across the team.
How does Thrawn handle last-minute schedule changes?
Thrawn handles changes through rapid re-optimization. When an unplanned absence occurs, the system doesn't just patch the hole; it re-runs the optimization to generate a new, globally fair schedule that accounts for the change while preserving fairness and respecting all other constraints.
Who is responsible for building the schedule with Thrawn?
Thrawn builds the schedule for you. As a managed service, you provide your program's constraints—rules, requests, and fairness goals—and receive a finished, optimized schedule for review and approval. This eliminates the manual work for chief residents.