Table 3.
Variables Significantly Associated With Treatment Failure in Logistic Regression Analysesa
Variable | OR | 95% CI | P |
---|---|---|---|
Univariate logistic regression analysis | |||
Transferred from another hospital | 7.1 | 1.2–43.1 | .033c |
Hepatomegaly | 13.2 | 2.0–85.8 | .007c |
Severe hepatitisb | 24 | 3.2–177.4 | .002c |
Bilirubin total (mg/dL) | 106.6 | 2.2–5191.9 | .019c |
Chloramphenicol use during admission | 20 | 2.1–192.7 | .010c |
Multivariate logistic regression analysisa | |||
Transferred from another hospital | 2.7 | 0.2–29.1 | .420 |
Severe hepatitis | 13.1 | 1.4–119.9 | .023c |
Chloramphenicol use during admission | 7.6 | 0.6–102.7 | .128 |
Abbreviations: CI, confidence interval; OR, odds ratio.
aVariables included in the multivariate logistic regression analysis limited to “transferred from another hospital,” “chloramphenicol use during admission,” and “severe hepatitis” as the representative hepatic variable to minimize the risk of multicollinearity. Odds ratio for the multivariate log regression analysis is adjusted OR.
bDefined as an aspartate aminotransferase or alanine aminotransferase level of ≥5 times the upper limits of normal.
cStatistically significant result.