Table 2.
Independent predictors of mortality using logistic regression
Variable | Value (mean ± SD) | Odds ratio | 95% CI | p |
Lactate (mEq/l) | 5.7 ± 5.8 | 1.145 | 1.107–1.184 | <0.0001 |
Age (years) | 57.6 ± 16.4 | 1.023 | 1.014–1.034 | 0.0011 |
SIG (mEq/l) | 5.8 ± 5.2 | 1.065 | 1.032–1.099 | 0.0012 |
Phosphate (mg/dl) | 4.0 ± 2.2 | 1.214 | 1.118–1.319 | <0.0001 |
With the use of stepwise logistic regression, these variables were identified as independent predictors of mortality (C statistic = 0.76). The order of the variables listed reflects their discriminating value in the model. The odds ratio for phosphate seems larger than the other variables because of the narrower distribution of phosphate than that of lactate or SIG. CI, confidence interval; SIG, strong ion gap or unmeasured ions.