Circadian-aware ED scheduling: the evidence
Emergency medicine is one of the few specialties where the schedule itself is a clinical risk factor. Here is what the literature supports, what it does not, and how to turn it into a number you can act on.
Updated August 2026 · 9 min read
Why the schedule is a clinical variable
Every ED schedule makes the same trade: 24-hour coverage has to come from people whose physiology is built for daytime work. There is no arrangement that removes that cost. There are, however, arrangements that spread it far better than others, and the gap between a well-built month and a rule-compliant but careless one is large.
The practical problem is that most scheduling tools evaluate a month as pass or fail against rules. Two schedules can both pass and still be very different to live through. What is missing is a graded measure.
Four findings that change how a month should be built
Recovery scales with the length of the night block
Re-entrainment after consecutive nights takes on the order of one day per night worked. A single day off after four nights is not recovery; it is a gap. Scheduling a day shift immediately after a night block is the single most common avoidable error we see in imported schedules.
Forward rotation beats backward rotation
The intrinsic circadian period is slightly longer than 24 hours, which makes delaying the clock easier than advancing it. Day to evening to night is tolerated better than the reverse. Backward jumps inside a single week are expensive.
Turnaround time matters more than total hours
Two shifts separated by nine hours do not contain a sleep opportunity once commute and wind-down are subtracted. Total monthly hours can look reasonable while the turnaround structure is punishing. Post-night turnaround is the highest-cost case.
Isolated nights are a pattern problem, not a one-off
A scattered single night prevents any adaptation and then demands re-adaptation. Repeated monthly, this is usually worse than a short, well-recovered block — and it is the pattern that manual scheduling produces most often, because single nights are the easiest holes to fill.
Turning the evidence into a score
The Circadian Alignment Index (CAI) is a 0 to 100 score that applies these findings to an actual month. It penalizes short turnarounds, weights post-night turnarounds more heavily, charges for backward rotation, and treats isolated nights and long night blocks differently. It is computed per physician, then summarized across the group with a median rather than a mean, so one outlier neither hides nor dominates the picture.
90–100
Low fatigue risk
75–89
Moderate fatigue risk
Below 75
High fatigue risk
A score is only useful if you can see what drove it. Every CAI result breaks down into the specific penalties that produced it, so a scheduler can fix the three things that matter instead of rebuilding the month.
What the evidence does not say
It does not support a single universally optimal rotation. It does not let anyone put a credible dollar figure on a fatigue-related error. And it does not mean a higher score is always achievable — coverage requirements, group size, and individual preferences constrain what any schedule can reach. The right target is improvement against your own baseline, not a number borrowed from another group.
Read further
Score your own month
Upload the Excel or PDF schedule you already use and see the CAI, the band distribution, and the top penalties driving it.
Upload a schedule