Table 3.
Multivariable analyses evaluating role of Arg/DMA ratio in predicting sepsis and outcomes
| Association between Arg/DMA ratio and severe sepsis | |||
| Dependent variable | Independent variables a | OR (95% CI) b | p value |
| Severe sepsis | Arg/DMA ratio | 5.91 (2.45 – 14.26) | <0.001 |
| Charlson score | 3.46 (1.64 – 7.30) | 0.001 | |
| Creatinine concentration | 1.72 (0.87 – 3.38) | 0.118 | |
| Association between Arg/DMA ratio and hospital mortality | |||
| Dependent variable | Independent variables c | OR (95% CI) b | p value |
|
| |||
| Hospital mortality | Arg/DMA ratio | 1.63 (1.00 – 2.65) | 0.048 |
| Age | 1.57 (1.01 – 2.44) | 0.046 | |
| Creatinine concentration | 1.58 (1.00 – 2.48) | 0.049 | |
| Association between Arg/DMA ratio and risk of death over 6 months | |||
| Dependent variable | Independent variables d | HR (95% CI) | p value |
|
| |||
| 6 – month mortality risk | Arg/DMA ratio | 1.41 (1.01 – 1.97) b | 0.043 |
| History of hypertension | 3.91 (1.51 – 10.14) | 0.005 | |
| Creatinine concentration | 1.22 (0.90 – 1.67) b | 0.198 | |
Definition of abbreviations: Arg/DMA ratio = Arginine/(asymmetric + symmetric dimethylarginine); OR = odds ratio; CI = confidence interval; HR = hazard ratio.
The initial model also included presence or absence or pre-existing hypertension, removed without significant deterioration in model fit as described in Methods.
These odds ratios and hazard ratios refer to a one quartile change in the value of the covariate.
The initial model also included Charlson comorbidity score and presence or absence or pre-existing hypertension. Both were removed without significant deterioration in model fit as described in Methods.
The initial model also included age and Charlson comorbidity score. Both were removed without significant deterioration in model fit as described in Methods.