On-call and incident response

    For on-call rotations: sleeping with the pager, and the week after

    You were not paged and you still slept badly. The phone was on the nightstand and some part of you stayed on duty all night. The week after the rotation you are flat, short-tempered, and behind on everything.

    The day

    Where the sessions go

    6 sessions · 44 minutes across a working day

    1. Two nights before the rotation

      Pre-Sleep Breathing

      Bank sleep deliberately. Going into a rotation with a deficit is the single biggest controllable factor.

      10 min
    2. Each on-call night

      4-7-8 Breathing

      Long exhales before sleep, phone face down and out of arm's reach. A rehearsed pattern is what you fall back on at 03:00.

      10 min
    3. During the incident

      Physiological Sigh

      Two inhales, one long exhale, before you type the first command. It costs twenty seconds and it prevents the decisions people write postmortems about.

      1 min
    4. After the incident

      4-7-8 Breathing

      Leave the room you worked in, then five minutes of long exhales. Do not re-check unless it pages again.

      5 min
    5. End of the rotation

      Extended Exhale

      A deliberate handover ritual. Off-duty has to be signalled to the body or it does not register.

      8 min
    6. The two days after

      Coherent Breathing

      Protected days planned in advance: no early starts, no consequential decisions. Plan them before the rotation, not during.

      10 min

    Anticipation alone fragments sleep

    Studies of on-call workers find worse sleep quality, more arousals and higher stress-hormone markers on on-call nights than on ordinary nights — including nights when nothing happened. The possibility of being woken keeps sleep lighter. So an uninterrupted on-call night is still a partial deficit, and planning as if it were a normal night is where most people go wrong.

    The deficit outlasts how you feel

    Under sustained sleep restriction, objective performance keeps degrading while subjective sleepiness plateaus — you stop noticing before you stop being impaired. That mismatch is the specific hazard of incident work, because the judgement required is highest exactly when the measurement of your own state is least reliable.

    The spike after a page is manageable; the ninety minutes afterwards is the cost

    Being woken into a production incident produces a full sympathetic response. The incident itself takes twenty minutes; getting back to sleep is what takes ninety. A rehearsed long-exhale pattern and a short physical transition out of the room reliably shorten that, which over a week of nights is the difference between one rough rotation and a month of recovery.

    What this will not do

    This makes a rotation survivable. It does not make an unsustainable rota sustainable. If you are on call one week in two, or paged most nights, or the only person who can resolve a class of incident, the fix is staffing, escalation policy and alert quality — and no breathing protocol should be used as a reason not to fix those.

    The full protocol

    7-Day On-Call Recovery

    7 days, written day by day with what to measure. It is inside the app, and the free Control Pause test is the number it starts from.

    Questions people in this role ask

    Why do I sleep badly even when nothing pages?
    Because anticipation alone lightens sleep — on-call nights show more arousals and higher stress markers regardless of whether a call comes. Worth knowing, so you stop counting a quiet night as a full one.
    How long does recovery from a bad rotation take?
    Longer than it feels. Subjective sleepiness recovers faster than actual performance, so plan two protected days rather than one and avoid consequential decisions in them.
    Should the phone be in another room?
    Only if the alert still wakes you reliably — that is a role-specific judgement. What is always available: face down, out of arm's reach, and nothing else read on it.

    What this page is built on

    • — Ziebertz et al., 2017, and the on-call sleep literature — sleep quality and stress markers on on-call nights without calls
    • — Van Dongen et al., 2003, Sleep — cumulative performance deficit under sleep restriction versus subjective sleepiness
    • — Balban et al., 2023, Cell Reports Medicine — cyclic sighing and acute arousal reduction
    • — Sonnentag & Fritz, 2015, Journal of Organizational Behavior — detachment and recovery

    What this usually shows up as

    Other roles