Chloride Content of Fluids Used for Large-Volume Resuscitation Is Associated with Reduced Survival
Sen A, Keener CM, Sileanu FE, et al. Crit Care Med. 2017;45:e146-e153.
Study Question: How do chloride load and fluid volume impact hyperchloremic acidosis (HCA), acute kidney injury (AKI), and mortality in critically ill patients?
Study Description: This single-center, retrospective study utilized the High-Density Intensive Care (HiDenIC-8) database (eight ICUs at University of Pittsburgh) for patients admitted from 2000 to 2008. Patients who received 60 mL/kg or greater in a 24 hour period, called large-volume resuscitation, were eligible for inclusion. Unadjusted and adjusted analyses were performed including regression analysis, receiver operating characteristic curve, and hazard modeling.
Results: Of the 4,710 patients receiving large-volume fluid resuscitation, unadjusted rates of HCA, AKI, and mortality were statistically higher in patients with greater chloride load based on quartile. Ordinal logistic regression of unadjusted data found HCA was associated with AKI (OR 1.24, p = 0.025); however, the model was unable to be adjusted for baseline covariates of age and acute physiology score-III (APS-III) differences between quartiles. Odds of mortality at 30 days was 70% higher in quartile four versus quartile one, which persisted after controlling for APS-III and fluid volume. One-year mortality worsened as chloride load increased in continuous model evaluation (Hosmer-Lemeshow p = 0.76; first quartile versus fourth quartile OR 1.49, 95% CI 1.16-1.92). Cox proportional hazard ratio plotting adjusted for fluid given, age and APS-III demonstrates a hazard ratio of 1.055 (p = 0.0015) for increased mortality with each 100 mEq increase in chloride load.
Conclusion(s): Increased chloride load is associated with higher rates of HCA, AKI, and mortality. Highest quartile (849 – 5,432 mEq) was 1.49 times likely to die at one year with a 5.5% increase in mortality over one year.
Perspective: Multiple studies have demonstrated increased AKI with supraphysiologic chloride containing fluids. This is the first study to examine the contribution of HCA on clinical outcomes. Impact of baseline covariates, specifically age and APS-III, were unable to be accounted for in the regression analysis. However, these results add to a growing body of literature which associates chloride load with mortality. Prospective, randomized, controlled studies should be performed to confirm the impact of chloride load on clinical outcomes.