Bio-mathematical fatigue models predict sickness absence in hospital nurses: An 18 months retrospective cohort study

This study examined the associations between bio-mathematical fatigue-risk scores and sickness absence (SA) in hospital nurses over 18 months. Work schedules and SA data were extracted from the hospital's attendance system. Fatigue-risk scores were generated for work days using the Fatigue Audit InterDyne (FAID) and Fatigue Risk Index (FRI). Over the study period, 5.4% of the shifts were absence shifts. FAID-fatigue ranged from 7 to 154; scores for a standard 9–5 work schedule can range from 7 to 40. Nurses with high FAID-scores were more likely to be absent from work when compared to standard FAID-scores (41–79, OR = 1.38, 95%CI = 1.21–1.58; 80–99, OR = 1.63, 95%CI = 1.37–1.94 and ≥ 100, OR = 1.73, 95%CI = 1.40–2.13). FRI-fatigue ranged from 0.9 to 76.8. When FRI-scores were >60, nurses were at 1.58 times (95%CI = 1.05–2.37) at increased odds for SA compared to scores in the 0.9–20 category. Nurse leaders can use these decision-support models to adjust high-risk schedules or the number of staff needed to cover anticipated absences from work. •Research has shown that sickness absence is related to fatigue in the workforce.•Bio-mathematical fatigue models may be a solution for fatigue monitoring at work.•To date, the predictive validity of these models is related to incidents/accidents.•Higher fatigue-risk scores increased the odds of sickness absence in hospital nurses.•The findings support the use of bio-mathematical fatigue models on nursing units.