Impact of intravenous fluid composition on SIRS patient outcomes

Shaw AD, Schermer CR, Lobo DN, Munson SH, Khangulov V, Hayashida DK, et al.

crit care 2015;19:334.

This retrospective, observational study used a propensity scoring system to match patients treated with saline with those given a balanced, calcium-free fluid. Patients presented with systemic inflammatory response syndrome (SIRS), and included a total of 3,116 matched patients (1,558 in each cohort). Those initially given saline had higher rates of in-hospital mortality, prolonged hospital stay, more frequent re-admission, and greater frequency of adverse clinical events.

The supraphysiological concentration of chloride in 0.9% saline has been associated with negative clinical outcomes in critically ill patients – either post-surgically or in intensive care. There is a lack of evidence exploring whether a similar deleterious effect can be extrapolated to patients with less severe illness. The aim of this study was to determine whether there was a demonstrable effect of using saline as an early fluid choice in SIRS.

The authors examined a large electronic US health record database in order to identify patients meeting two or more modified SIRS criteria (see below), who were over 18 years, and had received at least 500ml of fluid within 48 hours of diagnosis.

Modified SIRS criteria

• Heart rate >90 bpm

Plus one or more from:

• Temperature >38^oC or <36^oC
• Respiratory rate ≥20 breaths/min or PaCO₂ ≤32 mmHg
• Leukocyte count ≥12.0 x 10⁹/L or ≤4.0 x 10⁹/L

Two cohorts were created based on the type of crystalloid fluid administered; those given saline initially, and those given a physiologically balanced, calcium-free fluid. Patients who initially received balanced fluid (n = 1,558) were successfully matched to the saline cohort using a propensity scoring system – this aimed to minimise bias based on patient and hospital characteristics. The primary outcome was mortality; secondary outcomes included:

• Rate of in-hospital complications (data were further broken down by affected system)
• Electrolyte/acid base disturbances
• Length of stay
• Readmission

Despite the 1:1 matching, some parameters remained different between the two groups – these were adjusted for post analysis. The total fluid received was similar across cohorts, with a significantly higher total chloride load received by the saline cohort. The cohort initially receiving balanced fluid had several improved outcomes:

• Reduced mortality (1.03% vs. 3.27%; adjusted OR = 0.378, 95% CI 0.211–0.676)
• Shorter hospital stay (4.38 days vs. 4.86 days; p =0.016)
• Reduced readmission rate at 90 days (12.58% vs. 16.56%; adjusted OR = 0.741, 95% CI 0.606–0.907)
• Lower frequency of several electrolyte abnormalities(no overall measure given)
• Lower frequency of hyperchloraemic acidosis (adjusted OR = 0.449, 95% CI 0.256–0.786)

In their discussion, the authors remark that their findings contribute to a growing body of evidence demonstrating the possible harms of saline. They were able to evaluate a broad population, and corroborate existing findings. There were no measures that favoured saline as a fluid replacement, while several administrative and clinical outcomes suggested a benefit from balanced fluids. The higher total chloride burden was suggested as the potential reason behind this effect. Contrary to the findings in critical illness, there was no significant difference regarding renal risk in this study. The authors suggest that the size of the exposed population (0.9% saline being the most widely prescribed fluid internationally) necessitates further exploration.

This study adds to the emerging consensus that saline may not be as innocuous as its widespread use suggests. Being an observational study, it is at risk of bias despite efforts to control for it – there is a need for large-scale randomised trials to explore this issue more thoroughly.

Access Article Abstract