Table 3.
Univariate logistic regression model to test association of AKI and increase in serum chloride (ΔCL = [Cl-]max - CL0) in all patients and patients without hyperchloremia
| Variable | AKI stage 1 to 3 | p | AKI stage 2 and 3 | p |
|---|---|---|---|---|
| Odds ratio (95 % CI) | Odds ratio (95 % CI) | |||
| All patients | ||||
| Δ[Cl-] | 1.25 (1.16–1.36) | <0.001a | 1.14 (1.07–1.20) | <0.001a |
| Δ[Cl-] ≥ 5 mmol/l | 5.70 (3.00–10.82) | <0.001a | 3.40 (1.95–5.94) | <0.001a |
| Patients without hyperchloremia | ||||
| Δ[Cl-] | 1.37 (1.20–1.56) | <0.001a | 1.25 (1.13–1.38) | <0.001a |
| Δ[Cl-] ≥ 5 mmol/l | 8.25 (3.44–19.78) | <0.001a | 4.77 (2.13–10.70) | <0.001a |
AKI acute kidney injury, Δ[Cl - ] increase in serum chloride, [Cl - ] max maximal chloride concentration in the first 48 hours, [Cl - ] 0 initial chloride concentration, CI confidence interval
aIndicates statistical significance, p < 0.05