Table 5.

A priori vs. practical approach: Multivariable estimates of risk describing the association between shift length and occupational injury or illness depend on the definition of exposure.

	A Priori Approach N=961,827		Practical Approach N=926,763
	Relative Risk (95% CI)	p-value	Relative Risk (95% CI)	p-value
≤8 hours	Referent	---	Referent	---
>8 & ≤12 hours	1.43 (1.04–1.97)	0.03	1.16 (0.88–1.53)	0.30
>12 & ≤16 hours	1.82 (1.17–2.82)	0.008	1.43 (1.05–1.95)	0.025
>16 and ≤24 hours	2.29 (1.52–3.46)	<0.001	1.93 (1.33–2.81)	0.001
>24 hours	1.68 (0.23–12.42)	0.61	2.88 (1.74–4.77)	<0.001
Continuous (hours)	1.04 (1.02–1.07)	<0.001	1.03 (1.02–1.04)	<0.001
≥10 hours	1.37 (1.05–1.77)	0.018	1.24 (0.99–1.56)	0.06
≥12 hours	1.49 (1.18–1.88)	0.001	1.38 (1.12–1.70)	0.002
8 hour cat	Referent	---	Referent	---
12 hour cat	1.33 (1.05–1.70)	0.018	1.22 (0.96–1.54)	0.10
16 hour cat	2.08 (1.12–3.86)	0.021	1.42 (0.99–2.05)	0.06
24 hour cat	2.05 (1.44–2.90)	<0.001	2.13 (1.54–2.96)	<0.001

Note: Each estimate in the table represents a separate multivariable model adjusted for the previously described covariates.