Table 4.

A priori approach: Multivariable model results for predicting injury as a function of shift length and other characteristics

	Relative Risk (95% CI)	p-value
Shift Length (category)
≤8 hours	0.70 (0.51–0.96)	0.029
>8 & ≤12 hours	Referent	---
>12 & ≤16 hours	1.27 (0.91–1.77)	0.158
>16 and ≤24 hours	1.60 (1.22–2.10)	0.001
>24 hours	1.18 (0.16–8.49)	0.87
Overnight shift	0.78 (0.65–0.93)	0.005
Consecutive shift	0.84 (0.62–1.15)	0.29
Crew Familiarity
0–1 shifts	Referent	---
2–7 shifts	1.05 (0.88–1.24)	0.59
8–19 shifts	1.08 (0.90–1.30)	0.39
≥20 shifts	1.14 (0.91–1.41)	0.25
Hours of recovery	1.00 (1.00–1.00)	0.85
Employment Status
Full-time	Referent	---
Part-time	0.91 (0.78–1.07)	0.26
Midpoint Agency Size (20 person units)	0.99 (0.86–1.13)	0.88

^*Model includes 961,827 observations. The hours of recovery was not available for the first shift in the dataset.