Table 2.

Cox Proportional Hazards Regression Models^a

Model	Race Parameter Estimate (SD)	Hazard Ratio (95% CI)	P Valueb	Reduction in Race Parameter Estimate Compared With Model 1, %
Model 1 (no covariates)	0.81 (0.19)	2.26 (1.56–3.26)	<.01	NA
Model 2	0.39 (0.17)	1.47 (1.05–2.05)	.02	0.53
Model 2c,d	0.42 (0.17)	1.52 (1.08–2.13)	.01	0.49
Model 3	0.16 (0.16)	1.18 (0.87–1.60)	.30	0.80
Model 3d,e	0.27 (0.17)	1.31 (0.94–1.82)	.11	0.67

Abbreviation: NA, not applicable.

^aThe models illustrate the reduction in the race parameter estimate after balancing African American and white children with respect to biological, environmental, disease management, access, and socioeconomic hardship variables using inverse probability of treatment weighting.

^bP value obtained from the likelihood ratio test comparing African American and white children using adjusted Cox proportional hazards regression modeling.²⁸

^cModel 2 includes measures of financial and social hardship, caregiver educational attainment, and caregiver marital status.

^dTrimmed inverse probability of treatment weighting for a more stabilized result as a sensitivity analysis.

^eModel 3 includes outdoor allergen sensitization, salivary cotinine, traffic-related air pollution, running out of or missing dose of medication, and vehicle ownership alongside measures of financial and social hardship, caregiver educational attainment, and caregiver marital status.