Night Float vs. 24-Hour Call: The Trade-offs

Two models cover the same problem - someone needs to be responsible for the service overnight - and groups tend to pick between them based on tradition more than a clear-eyed comparison. Here's what each one actually costs, and where each one tends to strain.

24-hour call: continuity at the price of fatigue

Under a 24-hour call model, one provider carries the pager (and often stays in-house) for a full day-plus-night stretch, then hands off. The appeal is continuity: the person who admitted a patient at 8am is often still the one managing them at 2am, which can mean fewer handoffs and less information lost in translation.

The cost is fatigue, and it's a well-documented one. ACGME's own duty-hour framework treats 24-hour call as requiring a mandatory 14-hour recovery window afterward (see the ACGME's 24-hour continuous work cap, §6.22., and the associated 14-hours-free requirement, §6.21.a.) - a tacit acknowledgment that a 24-hour shift is not something a person simply walks away from and starts fresh the next day.

Night float: fresher clinicians, more handoffs

Night float assigns one or more providers to work nights only, for a stretch of days or weeks, while day teams handle days. Each individual shift is shorter and the provider working it is (in theory) rested and alert for that specific window, rather than 20 hours into a marathon.

The cost shows up at the handoff. Every admission and every overnight event now has to be communicated to someone who wasn't there for the earlier context, twice a day instead of once. Programs that run night float well invest real structure in the handoff itself - a standard template, protected time, closed-loop confirmation - because the model's biggest failure mode is information loss at 7am and 7pm, not fatigue.

What the numbers say about which one is more common

There's no single universal answer across specialties; the honest picture is mixed and depends heavily on service volume and acuity. What's consistent is the direction of travel: night float has become the more common structure for high-volume inpatient services, in large part because it's simply easier to keep hour totals compliant with a series of 10-to-12-hour night shifts than with recurring 24-hour blocks. Notably, the ACGME's Common Program Requirements don't set a universal cap on consecutive weeks of night float at the common-requirement level (§6.26.) - that detail is left to individual specialty Review Committees, so check your specialty's own requirements rather than assuming a single number applies everywhere.

Which one fits your group

• Lower admission volume, high value on continuity → 24-hour call tends to work better; fewer handoffs matter more than the fatigue cost when nights are quiet.
• Higher admission volume, unpredictable overnight acuity → night float tends to work better; a fresher clinician handling a busy night beats a fatigued one handling the tail end of a 20-hour shift.
• Small groups (under six or seven providers) → 24-hour call is often the only structure with enough bodies to cover it; night float needs a critical mass of people willing to flip their schedule for a stretch.
• Residency programs → the specialty's own Review Committee requirements typically settle this before local preference does.

Emergency medicine and ICU/critical-care services in particular tend to run night float almost by default, given round-the-clock high-acuity volume - see how each specialty structures its coverage in our by-specialty scheduling guides.

Common questions

Can a group run both models at once?

Yes, and many do - night float for weeknights when volume is predictable, with weekend call reverting to a 24-hour block model when staffing is thinner. RotaBay's night-float template pairs a dedicated weeknight shift with weekend call for exactly this reason.

Which model is safer for patients?

The honest answer is that both carry real risk, in different places - fatigue-related errors under 24-hour call, handoff-related information loss under night float. The safest version of either model is the one with explicit structure: enforced recovery time for call, and a disciplined handoff process for night float.