Shift Management and Circadian Cycles
What to observe when comparing shifts and how to avoid attributing differences to circadian rhythms without supporting evidence.
Updated:
Night work and shift organisation can affect rest and safety. An activity study describes the process; it does not diagnose fatigue, chronotype or illness, or show that a shift difference is caused by the biological clock.
What to compare
Define tasks, demand, staffing, product, equipment and represented hours. If one shift maintains equipment and another produces, direct-activity shares may not be comparable. Differences require process data and a suitable design; they must not be automatically attributed to shift timing or used to diagnose individual fatigue.
Sampling design
Cover included shifts and retain their exposure. Define execution, preparation, coordination, confirmed waiting, breaks and unobservable states. Required breaks and safety tasks are not waste.
For an independent simple-sample proportion, planning starts with n₀=ceiling(z²p₀(1−p₀)/e²). At two-sided 95%, p₀=0.5 and five-percentage-point error, this gives 385. Two hundred and fifty readings per period do not guarantee detection of a 10% fall. A comparison needs a relevant difference, expected proportions, power, significance level and design. Dependence within people, days and rounds may reduce effective information.
Differences and causes
A differing share prompts investigation; it is not a “circadian-adjusted” measure. Record load, breakdowns, experience and supply. To attribute change to a redesign, use comparable periods, a documented intervention and, where feasible, a comparison group. Neither large n nor random timing removes confounding or reactivity.
Activity sampling does not directly estimate critical-error risk. That requires defined events, exposure denominators, suitable data and specific safety criteria. OEE also needs unit counts, ideal cycle time and quality, not operating states alone.
Organisation and health
Review shifts, rest and load with occupational-prevention professionals and worker representatives. Do not restrict twelve-hour shifts to “evening types” or use chronotype questionnaires for automatic assignments. Decisions require risk assessment, individual considerations and confidentiality.
Forward rotation usually means morning–afternoon–night. It should not be justified by claiming that advancing the clock is always easier: that pattern tends to delay schedules. Suitability depends on the system. Do not prescribe 480–500 nm blue light to “reset” the clock; lighting interventions need competent assessment and consideration of subsequent rest.
Legal scope and checks
In Spain, review Workers' Statute article 36 and applicable provisions: night work is between 22:00 and 06:00, and night-worker rules have conditions and exceptions. Directive 2003/88/EC defines night time as at least seven hours including 00:00–05:00, not a mandatory 00:00–07:00 interval.
Sampling does not replace individual working/rest-time records or certify compliance. Separate estimated activity, working-time records and risk assessment.
Improvement plan
Document the baseline, preventive proposal, owners, consultation and follow-up. Check that changes do not compromise safety or rights. Worksamp supports activity studies; attendance recording answers another question. Do not guarantee productivity, health or a fixed financial return.
Sources: Spanish Workers' Statute, Directive 2003/88/EC, NIST.