Table 4.
Comparing multivariate logistic regression analysis for predicting 72-h mortality.
| Exp Var | β | SE (β) | z | P value | VIF | Odds ratio | 95%CI | n (with all Exp Vars) | AIC | |
|---|---|---|---|---|---|---|---|---|---|---|
| Reference model | ||||||||||
| 0 | − 4.586 | 0.7820 | 428 | 505.5 | ||||||
| 1 | Age | 0.022 | 0.0075 | 2.906 | 0.0037 | 1.020 | 1.022 | 1.007–1.037 | ||
| 2 | AST | 0.114 | 0.0229 | 4.965 | 0.0000 | 1.020 | 1.120 | 1.071–1.172 | ||
| Prediction model | ||||||||||
| 0 | − 19.98 | 3.6070 | 303 | 269.7 | ||||||
| 1 | ALP | 0.290 | 0.0601 | 4.824 | 0.0000 | 1.078 | 1.337 | 1.188–1.504 | ||
| 2 | CK | 0.348 | 0.0796 | 4.365 | 0.0000 | 1.112 | 1.416 | 1.211–1.655 | ||
| 3 | Na | 1.766 | 0.5617 | 3.145 | 0.0017 | 1.119 | 5.850 | 1.946–17.59 | ||
| 4 | K | 1.568 | 0.4755 | 3.298 | 0.0010 | 1.522 | 4.796 | 1.889–12.18 | ||
| 5 | P | 0.474 | 0.1145 | 4.145 | 0.0000 | 1.399 | 1.607 | 1.284–2.011 | ||
| P = 1/ [1 + exp (− 19.98 + 0.290(ALP0.2) + 0.348log (CK) + 1.766(Na0.2) + 1.568(K0.2) + 0.474(P0.8))] | ||||||||||
n, number; Exp Var, explanatory variable; SE, standard error; z, z value; VIF, variance inflation factor; CI, confidence interval; AIC, Akaike's information criterion; exp, exponential function; log, logarithm, Sn, sensitivity; Sp, specificity; AST, aspartate transaminase; ALP, alkaline phosphatase; Na, sodium; K, potassium; and P, phosphorus. P-values were calculated using the chi-square test.