Table 2:
Random-Effects Logistic Regression Model Showing Pairwise Associations between Statin, CRP, and Being Free of Delirium
Statin: Free of Delirium | Statin: CRP | |||||
---|---|---|---|---|---|---|
(n = 228; Person-Days = 1,246) |
(n = 226; Person-Days = 1,123) |
|||||
OR | 95% CI | P Value | β | 95% CI | P Value | |
Statin | 2.28 | 1.01 to 5.13 | <0.05 | −0.52 | −0.70 to −0.33 | <0.01 |
Age, per yr | 0.99 | 0.96 to 1.03 | 0.71 | −0.00 | −0.01 to 0.00 | 0.35 |
Sex, women vs. men | 0.48 | 0.20 to 1.19 | 0.11 | 0.03 | −0.19 to 0.24 | 0.81 |
mSOFA, per point | 0.59 | 0.47 to 0.73 | <0.01 | 0.13 | 0.09 to 0.18 | <0.01 |
Sepsis, yes vs. no | 1.53 | 0.62 to 3.76 | 0.36 | 0.29 | 0.07 to 0.51 | 0.01 |
Propensity score* | 2.90 | 0.35 to 23.83 | 0.32 | 0.02 | −0.49 to 0.53 | 0.94 |
Ventilated, yes vs. no | 0.78 | 0.72 to 0.84 | <0.01 | −0.01 | −0.03 to 0.00 | 0.14 |
Emergency vs. elective | 12.81 | 3.05 to 53.82 | <0.01 | 0.36 | 0.05 to 0.66 | 0.02 |
Definition of abbreviations: CI = confidence interval; CRP = C-reactive protein; mSOFA = modified Sequential Organ Failure Assessment (minus Glasgow Coma Score); OR = odds ratio.
β is the slope from linear regression.
CRP quantities log(e) transformed to obtain normal distribution. When used as a dependent variable, OR is per increase in standard deviation of log(e) CRP.
All models allow for random-effects for each individual.
Propensity score according to age, sex, primary hypercholesterolemia, ischemic heart disease, diabetes, peripheral and cerebrovascular disease, and admission for aortic aneurysm surgery.