Table 2.

Endothelial Dysfunction and Brain Dysfunction: Delirium/Coma-Free Days

N = 134	25th Percentile	75th Percentile	Adjusted Difference 25th vs. 75th (95% CI)*	P-value
Independent Variable
• Reactive Hyperemia Index	1.32	1.81	−0.8 (−1.54, −0.12)	0.02
Covariates
• Age (years)	47	67	1.2 (0.03, 2.37)	0.05
• APACHE APS	17	27	2.2 (1.11, 3.35)	<0.001†
• ICU Type	Medical	Surgical	0.5 (−1.34, 2.28)	0.61
• IQCODE	3.00	3.19	0.1 (−0.42, 0.53)	0.83
• Prehospital Statin Use	No	Yes	−0.4 (−2.29, 1.46)	0.66
• Severe Sepsis	No	Yes	1.2 (−0.87, 3.18)	0.27

^*In this analysis, a positive value is indicative of better outcomes

^†Nonlinear relationship

Abbreviations: APACHE, Acute Physiology and Chronic Health Evaluation;²⁷ APS, Acute Physiology Score; ICU, intensive care unit; IQCODE, Informant Questionnaire on Cognitive Decline in the Elderly.²⁸

Endothelial function, the independent variable, was assessed using the Endo_PAT (Itamar Medical Ltd, Franklin, MA) reactive hyperemia index (RHI), a validated assessment of systemic endothelial function where <1.67 is considered endothelial dysfunction.^18;19 Linear regression was used to study the association of endothelial function (N = 134) with acute brain dysfunction after adjusting for age, acute physiology score of the Acute Physiology and Chronic Health Evaluation II,²⁷ severe sepsis, Informant Questionnaire on Cognitive Decline in the Elderly²⁸ score, medical versus surgical ICU admission, and prehospital statin use. The adjusted difference represents the difference in delirium/coma-free days between the 25^th vs. the 75^th percentile for continuous variables or no vs. yes for dichotomous variables. Interpretative example: Other covariates being equal, patients with a RHI at the 25^th percentile (1.32, poor endothelial function) would have, on average, 0.8 fewer delirium/coma-free days than patients with a RHI at the 75^th percentile (1.81, good endothelial function).