Peer-reviewed evidence on shift work, health, schedule design, and business performance, graded by strength and translated into operational guidance. https://soon.works/research en-us https://soon.works/research/8-vs-10-vs-12-hour-shifts-the-safety-tradeoffhttps://soon.works/research/8-vs-10-vs-12-hour-shifts-the-safety-tradeoffSun, 19 Jul 2026 00:00:00 GMTDesigning safer schedulesAccident and injury risk accumulates within a shift as time on task grows. Compared with an 8-hour baseline, a meta-analysis (Fischer et al., 2017) found injury risk of RR 1.54 (95% CI 1.30-1.83) for 10-hour shifts and RR 2.73 (95% CI 2.02-3.69) for shifts over 12 hours, and an earlier systematic review (Wagstaff & Sigstad Lie, 2011) put risk at around 12 hours at roughly double the risk at 8 hours. That is the tradeoff to weigh, not a reason to ban 12-hour shifts outright. (Evidence: strong; 4 cited sources.)https://soon.works/research/do-mandated-nurse-ratios-save-liveshttps://soon.works/research/do-mandated-nurse-ratios-save-livesSun, 19 Jul 2026 00:00:00 GMTStaffing levels and outcomesCalifornia's ratio law, AB 394, was implemented in January 2004 and was followed by an average reduction of 0.98 patients per licensed nurse, so a legislated floor did move real staffing rather than being absorbed by non-compliance (McHugh et al., 2012). Whether that change altered patient survival is not established: the analysis is a within-California pre/post time series whose authors state plainly that they were unable to determine the causal effect of the change. The two figures most often quoted as proof that ratios save lives, the "almost 30% lower mortality" contrast and the "222 fewer deaths" estimate, are model projections built from correlational coefficients, not outcomes anyone measured after a mandate. (Evidence: moderate; 3 cited sources.)https://soon.works/research/does-better-scheduling-actually-payhttps://soon.works/research/does-better-scheduling-actually-paySun, 19 Jul 2026 00:00:00 GMTScheduling and business performanceThe strongest evidence that scheduling practice moves business results comes from the one randomized field experiment on business outcomes we were able to verify: a bundle of responsible scheduling practices at Gap Inc. raised store labor productivity 5.1% on an intent-to-treat basis, from a 3.3% sales increase and a 1.8% reduction in labor hours (Kesavan, Lambert, Williams, & Pendem, 2022). Everything else in this pillar is survey evidence, which links unpredictable schedules to worse worker outcomes but cannot establish that changing a schedule produces them. That asymmetry, one experiment alongside a large body of correlational work, is the honest shape of the field. (Evidence: moderate; 3 cited sources.)https://soon.works/research/does-nurse-staffing-affect-patient-outcomeshttps://soon.works/research/does-nurse-staffing-affect-patient-outcomesSun, 19 Jul 2026 00:00:00 GMTStaffing levels and outcomesThe strongest thing this evidence base offers is convergence, not proof. Two large cross-sectional studies, one in Pennsylvania and one across 300 hospitals in nine European countries twelve years later, produced nearly the same adjusted coefficient for nurse workload against death within 30 days of admission: a coefficient near 1.07 on the odds scale in both Aiken et al. (2014) and Aiken et al. (2002). Both are associations measured on the odds scale, and no study in this literature has established that changing a hospital's staffing changes its mortality. (Evidence: moderate; 4 cited sources.)https://soon.works/research/fixed-nights-vs-rotating-shifts-which-is-healthierhttps://soon.works/research/fixed-nights-vs-rotating-shifts-which-is-healthierSun, 19 Jul 2026 00:00:00 GMTShift work and healthThere is no single healthy shift schedule. An umbrella review of meta-analyses (Cho & Kang, 2026) found that fixed-night and rotating patterns carry distinct, non-interchangeable penalties: fixed nights are associated with cardiometabolic harm, including ischemic heart disease (pooled RR 1.44, 95% CI 1.10-1.89), while rotating shifts are associated with worse sleep quality, cancer, and pre-eclampsia. The right question is not which schedule is safe, but which risk profile your workforce is best placed to absorb. (Evidence: moderate; 1 cited source.)https://soon.works/research/how-many-night-shifts-in-a-row-is-safehttps://soon.works/research/how-many-night-shifts-in-a-row-is-safeSun, 19 Jul 2026 00:00:00 GMTDesigning safer schedulesInjury risk on night shifts is associated with how many are worked back to back, and it is highest on the fourth consecutive night. In a systematic review and meta-analysis (Fischer et al., 2017, as compiled by Garde et al., 2020), injury risk ratios rose from 1.0 on the first night to 1.05 on the second, 1.16 on the third, and 1.36 (95% CI 1.14-1.62) on the fourth. On that evidence, an expert consensus recommends no more than 3 consecutive night shifts (Garde et al., 2020). (Evidence: strong; 2 cited sources.)https://soon.works/research/how-to-design-a-safer-shift-schedulehttps://soon.works/research/how-to-design-a-safer-shift-scheduleSun, 19 Jul 2026 00:00:00 GMTDesigning safer schedulesThree scheduling levers do most of the work in lowering shift-work risk: the rest gap between shifts, the number of night shifts worked in a row, and the length of each shift. This page maps each lever to the evidence and to a deep guide. The consolidated three-limit standard from the expert consensus (Garde et al., 2020), and the caveats that come with it, live in the night shift rulebook. (Evidence: strong; 4 cited sources.)https://soon.works/research/how-to-judge-a-scheduling-studyhttps://soon.works/research/how-to-judge-a-scheduling-studySun, 19 Jul 2026 00:00:00 GMTScheduling and business performanceThe first thing to check in any scheduling study is not the effect size, it is how people ended up in the groups being compared. The Gap study randomized stores into treatment and control, which is why its 5.1% productivity result supports causal language, while the Shift Project studies surveyed workers at one point in time on a non-probability sample and support association only (Kesavan et al., 2022; Schneider & Harknett, 2019, 2021). Same topic, very different warrant, and almost everything else about reading this evidence well follows from telling those two situations apart. (Evidence: moderate; 3 cited sources.)https://soon.works/research/nurse-staffing-and-patient-mortalityhttps://soon.works/research/nurse-staffing-and-patient-mortalitySun, 19 Jul 2026 00:00:00 GMTStaffing levels and outcomesAcross three landmark cross-sectional studies, hospitals whose nurses carried heavier patient loads also showed higher odds of surgical patients dying within 30 days of admission. The largest of them, Aiken et al. (2014), measured the average number of patients per nurse across 300 European hospitals and found each additional patient in that average associated with a little under 7% higher odds of death, OR 1.068 in the fully adjusted model. None of these designs can establish cause, so what the literature supports is a consistent association rather than a demonstrated causal relationship. (Evidence: moderate; 3 cited sources.)https://soon.works/research/rotating-vs-permanent-night-shifts-and-accidentshttps://soon.works/research/rotating-vs-permanent-night-shifts-and-accidentsSun, 19 Jul 2026 00:00:00 GMTDesigning safer schedulesA systematic review (Wagstaff and Sigstad Lie, 2011) found that shift work including nights carries substantially elevated accident risk, whereas pure permanent night work may bring some protection against this effect, because workers on a stable nighttime schedule can partially resynchronize. The honest complication is that permanent nights which look safer for accidents are associated with worse long-term heart health, with fixed nights linked to ischemic heart disease at a relative risk of 1.44 (Cho & Kang, 2026). Confidence in the accident contrast is moderate, resting on a single systematic review. (Evidence: moderate; 2 cited sources.)https://soon.works/research/shift-work-and-health-what-the-evidence-showshttps://soon.works/research/shift-work-and-health-what-the-evidence-showsSun, 19 Jul 2026 00:00:00 GMTShift work and healthThe strongest evidence linking shift work to health sits in one cluster: cardiometabolic risk, where myocardial infarction reaches the highest grade this field has awarded, a relative risk of 1.23 (Wu et al., 2022). Outside that cluster, for outcomes like cancer and psychosocial stress, the evidence stays weak or inconclusive, which is not the same as showing no risk. (Evidence: strong; 4 cited sources.)https://soon.works/research/shift-work-and-heart-diseasehttps://soon.works/research/shift-work-and-heart-diseaseSun, 19 Jul 2026 00:00:00 GMTShift work and healthShift work is associated with a higher rate of myocardial infarction at a relative risk of 1.23 (95% CI 1.15-1.31), the finding an umbrella review (Wu et al., 2022) graded highly suggestive, the top tier this field reaches. For a scheduler, the more useful number is the dose-response: each additional 5 years of night-shift exposure is associated with a further RR 1.07 for cardiovascular incidence (Xi et al., 2025). All of it is observational, so read this as a strong association, not proven cause. (Evidence: strong; 3 cited sources.)https://soon.works/research/shift-work-and-high-blood-pressurehttps://soon.works/research/shift-work-and-high-blood-pressureSun, 19 Jul 2026 00:00:00 GMTShift work and healthRotating shift schedules that include night periods are the specific pattern peer-reviewed research links to higher blood pressure. A meta-analysis of 27 cohort studies covering 394,793 individuals (Manohar et al., 2017) found rotating shift workers had a pooled odds ratio of hypertension of 1.34 (95% CI 1.08-1.67). This is an observed association, not proof that the schedule itself raises blood pressure. (Evidence: strong; 2 cited sources.)https://soon.works/research/shift-work-diabetes-and-weighthttps://soon.works/research/shift-work-diabetes-and-weightSun, 19 Jul 2026 00:00:00 GMTShift work and healthTwo umbrella reviews of observational studies find night-shift work is associated with roughly 10% higher odds of type 2 diabetes (pooled adjusted odds ratio 1.08 to 1.15), and the risk rises with cumulative exposure, by about RR 1.07 for every 5 years worked (Wu et al., 2022; Boini et al., 2022). Because the association is dose-dependent and builds over a career rather than over a single bad week, the schedule-design lever that matters most is limiting a person's lifetime tenure on nights, not fixing one rough rotation. (Evidence: strong; 2 cited sources.)https://soon.works/research/stable-scheduling-randomized-experiment-gaphttps://soon.works/research/stable-scheduling-randomized-experiment-gapSun, 19 Jul 2026 00:00:00 GMTScheduling and business performanceBetween November 2015 and August 2016, Gap Inc. let researchers randomize a bundle of scheduling practices across 28 of its stores, and the intent-to-treat analysis found that responsible scheduling practices increased store productivity by 5.1%, a result of increasing sales by 3.3% and decreasing labor by 1.8% (Kesavan et al., 2022). Randomization is what earns the word increased here, and no other study in this pillar has it. The result is also narrower than the headline suggests, because the control stores were already receiving two-week advance notice and no on-call shifts, so 5.1% is the marginal return on additional practices layered on that base. (Evidence: strong; 2 cited sources.)https://soon.works/research/the-11-hour-rest-rule-between-shiftshttps://soon.works/research/the-11-hour-rest-rule-between-shiftsSun, 19 Jul 2026 00:00:00 GMTDesigning safer schedulesA quick return, less than 11 hours between the end of one shift and the start of the next, is the single shift-design flaw with the strongest evidence against it. In a two-armed cluster randomized controlled trial across 66 hospital units and 1,314 workers, roughly halving quick returns reduced insomnia, sleepiness, and work-related fatigue (Djupedal et al., 2024). Separately, each additional hour of rest between two shifts was associated with about 5% lower injury risk (IRR 0.95, 95% CI 0.93-0.96) in a 69,200-worker Danish cohort. (Evidence: strong; 4 cited sources.)https://soon.works/research/the-evidence-based-night-shift-rulebookhttps://soon.works/research/the-evidence-based-night-shift-rulebookSun, 19 Jul 2026 00:00:00 GMTDesigning safer schedulesA 2020 expert consensus of 15 shift-work researchers (Garde et al., 2020) distilled night-work safety into three scheduling limits: no more than 3 consecutive night shifts, at least 11 hours between shifts, and no more than 9 hours per shift. The panel was explicit that these numbers rest on limited existing literature with major knowledge gaps, so treat them as a defensible default rather than a proven formula. They are also constraints a scheduling tool can enforce automatically, which is what makes them practical. (Evidence: moderate; 1 cited source.)https://soon.works/research/unpredictable-schedules-and-financial-hardshiphttps://soon.works/research/unpredictable-schedules-and-financial-hardshipSun, 19 Jul 2026 00:00:00 GMTScheduling and business performanceMoving from the most predictable to the most unpredictable schedules, the predicted probability that an hourly service worker reports hunger hardship rises from 23% to 50%, adjusting for wages, income, and hours (Schneider & Harknett, 2021). The same upward gradient appears in three other domains: residential, medical, and utility hardship. This is a cross-sectional survey, so it establishes association rather than cause, but the association is large and it survives controlling for pay. (Evidence: moderate; 1 cited source.)https://soon.works/research/unpredictable-schedules-and-worker-wellbeinghttps://soon.works/research/unpredictable-schedules-and-worker-wellbeingSun, 19 Jul 2026 00:00:00 GMTScheduling and business performanceSchneider and Harknett (2019) surveyed 27,792 hourly workers at 80 of the largest US food-service and retail firms and found that routine work-schedule instability is associated with higher psychological distress, poorer sleep quality, and greater unhappiness. Shift cancellation shows the widest gap of any practice measured: among the 14% of workers who had a shift canceled, 64% reported psychological distress, versus under half of those who did not. This is a cross-sectional survey, so these are associations captured at a single point in time, not evidence that canceling a shift caused the distress. (Evidence: moderate; 1 cited source.)https://soon.works/research/what-shift-work-is-not-proven-to-causehttps://soon.works/research/what-shift-work-is-not-proven-to-causeSun, 19 Jul 2026 00:00:00 GMTShift work and healthWhen umbrella reviews graded the research, only four shift-work health outcomes reached the strongest evidence tier: myocardial infarction, diabetes, overweight or obesity, and hypertension (Wu et al., 2022; Boini et al., 2022). Most other harms that headlines attach to shift work, including cancer, cardiovascular mortality, lipid disorders, smoking, sedentariness, and psychosocial stress, rest on weak or inconclusive evidence. Weak evidence means those harms are not established, which is not the same as proven harmless, so the honest reading is to neither panic nor dismiss them. (Evidence: moderate; 2 cited sources.)https://soon.works/research/what-understaffing-does-to-nurseshttps://soon.works/research/what-understaffing-does-to-nursesSun, 19 Jul 2026 00:00:00 GMTStaffing levels and outcomesIn the landmark hospital nurse survey, each additional patient in a nurse's workload was associated with 23% higher odds of burnout (OR 1.23, 95% CI 1.13-1.34) and 15% higher odds of job dissatisfaction (OR 1.15, 95% CI 1.07-1.25), adjusted for nurse and hospital characteristics (Aiken et al., 2002). The design is cross-sectional, so it establishes that heavier assignments and worse nurse outcomes appear together, not that one produces the other. (Evidence: moderate; 2 cited sources.)https://soon.works/research/who-quits-when-schedules-get-stablehttps://soon.works/research/who-quits-when-schedules-get-stableSun, 19 Jul 2026 00:00:00 GMTScheduling and business performanceIn the Gap stable scheduling experiment, a practitioner report states that average tenure of employees who quit treatment stores fell 10.8 months against a 25.5-month base, while overall retention was unchanged (Williams et al., 2018). Read plainly, stable scheduling changed who left rather than how many left: the people walking out had less experience, and experienced staff stayed. That report was not peer reviewed, so treat the tenure figures as a mechanism worth testing rather than a settled effect. (Evidence: emerging; 2 cited sources.)