Table 3. Multivariable logistic regression modeling, with a dependent variable of mortality by 60 days post-ARDS diagnosis.

Dependent Variable	Independent Variable	Model 1	Model 2
Day 60 mortality	Total FGF23	1.43 (1.07, 1.91), p = 0.014	1.62 (1.07, 2.45), p = 0.023
Day 60 mortality	Intact FGF23	1.44 (1.02, 2.05), p = 0.039	1.30 (0.89, 1.90), p = 0.17

In these models, the independent variable is Day 1 plasma total FGF23 or intact FGF23, both of which were log-transformed to correct for skewness. Model 1 is unadjusted. Model 2 is adjusted for age, sex, the presence or absence of Day 1 AKI, P/F ratio, and PRISM score. Data shown are odds ratios and 95% confidence intervals. As only subjects with complete covariate data are included in the regression analysis, n = 150 for both models.