Table 4.

Sepsis as an independent risk factor for 30-day hospital readmission

Model	OR (95% CI)	P Value
Sepsis, population-averaged effect	1.96 (1.81–2.12)	<0.001
Sepsis, subject-specific effect	1.70 (1.55–1.87)	<0.001
Sepsis, subject-specific effect, adjusted for presepsis characteristics and center*	1.51 (1.38–1.66)	<0.001
Sepsis, subject-specific effect, after forcing each presepsis characteristic and center into the model	1.36 (1.24–1.48)	<0.001

^*Mixed-effects multivariable models adjusted for comorbidity of malignancy. The OR (odds ratio) was not significantly altered (<10%) by adjustment for number of hospitalizations in the prior year after accounting for subject-specific random effects, nor did age, sex, race, marital status, insurance status, Charlson Comorbidity Index, or center alter the OR significantly. CI = confidence interval.