Table 2.

Cross-sectional Associations of Return to Work After ARDS With Pre-ARDS Workload Phenotypes, Functional Impairments, Pain, and Fatigue^a

Variable	Individual Models								Combined Model
	Return to Work		Return to Work		Return to Work		Return to Work		Return to Work
	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P
Pre-ARDS workloadb
Only high psychosocial workload	Reference								Reference
High physical and low psychosocial workload	0.76 (0.25, 2.35)	.64	…	…	…	…	…	…	0.40 (0.13, 1.20)	.10
High physical and high psychosocial workload	0.92 (0.35, 2.40)	.86	…	…	…	…	…	…	1.17 (0.47, 2.89)	.74
Functional impairmentsc
No impairment			Reference						Reference
Only psychosocial impairments	…	…	0.18 (0.06, 0.50)	.001	…	…	…	…	0.40 (0.14, 1.16)	.09
Physical and low psychosocial impairment	…	…	0.08 (0.03, 0.22)	<.001	…	…	…	…	0.28 (0.10, 0.78)	.015
Physical and high psychosocial impairment	…	…	0.01 (0.003, 0.05)	<.001	…	…	…	…	0.07 (0.02, 0.27)	<.001
Pain	…	…	…	…	0.06 (0.03, 0.14)	<.001	…	…	0.21 (0.09, 0.52)	.001
Fatigue	…	…	…	…	…	…	0.07 (0.03, 0.16)	<.001	0.39 (0.16, 0.96)	.041

The intraclass correlation (ICC) for the combined model was 62%.

^aEach column represents one model. Generalized linear mixed model with a random intercept evaluating the association of working status with the variable named in that row, adjusted for age, sex, race, pre-ARDS median household income, APACHE III (Acute Physiology and Chronic Health Evaluation III), hospital length of stay, and follow-up time.

^bPsychosocial workload combines cognitive, emotional, and interpersonal workload categories, with a job exhibiting low (<2 categories), or high (≥2 categories) numbers of these categories.

^cPsychosocial impairment combines cognitive, emotional, and interpersonal impairment categories, with survivors having impairments in low (<2 categories) or high (≥2 categories) number of these categories.