Buffered solutions versus 0.9% saline for resuscitation in critically ill adults and children

Fluid therapy is one of the main interventions provided for critically ill patients, although there is no consensus regarding the type of solution that should be used. There are two main types: colloid and crystalloid. The most commonly administered crystalloid solution is 0.9% saline. Buffered solutions may offer some theoretical advantages (e.g. less metabolic acidosis, less electrolyte disturbance), but the clinical relevance of these remains unknown. This is an update of a review published in 2019.

To assess the effects of buffered solutions versus 0.9% saline for resuscitation or maintenance in critically ill adults and children.

We searched CENTRAL, MEDLINE, Embase, CINAHL, and four trial registers in July 2023. We checked references, conducted backward and forward citation searches for relevant articles, and contacted study authors to identify additional studies. Although we updated our search in June 2025, the results have not yet been fully incorporated into the review.

We included randomised controlled trials (RCTs) with parallel or cross-over design that examined buffered solutions versus 0.9% saline in a critical care setting (resuscitation or maintenance). We included studies with participants who required intravenous fluid therapy due to critical illness (including trauma and burns) or undergoing emergency surgery during critical illness. We included studies of adults or children (or both). We excluded studies of people undergoing elective surgery and studies with multiple interventions in the same arm.

Our critical outcomes were overall (in-hospital) mortality and acute renal injury. Our important outcomes were organ system dysfunction, need for renal replacement therapy, days without organ support, electrolyte disturbances, blood loss or transfusion, coagulation disorders, total resuscitation fluid volume, quality of life, and cost. To populate a table summarising the findings of our review, we selected key outcomes for decision-makers, which were our two critical outcomes and two of our important outcomes (organ system dysfunction and electrolyte disturbances).

Two review authors independently assessed the risk of bias of each included study using the Cochrane risk of bias tool RoB 1. We considered pharmaceutical industry funding as a potential source of bias.

Where possible, we synthesised results for each outcome using random-effects meta-analysis. We reported outcomes using the odds ratio (OR) and 95% confidence intervals (CIs). We used the GRADE approach to assess the certainty of evidence.

We included 34 studies, with a total of 37,859 participants. Two RCTs with 26,854 participants contributed more than 70% of the total sample. Adults were the participants in 22 trials, and children in 12. All studies enroled critically ill participants: people with diabetic ketoacidosis (six studies), acute pancreatitis (five studies), severe dehydration (five studies), sepsis or septic shock (four studies), severe trauma (three studies), dengue shock syndrome (two studies), and mixed conditions (nine studies). The studies took place in 16 countries. All studies were published in English. We judged 16 studies to have an overall low risk of bias (i.e. low risk of bias for allocation concealment, blinding of participants and blinding of assessors, incomplete outcome data, and selective reporting). In the remaining trials, we judged that some form of bias had been introduced or could not be ruled out.

We found that buffered solutions result in little to no difference in overall (in-hospital) mortality (OR 0.95, 95% CI 0.90 to 1.01; I² = 0%; 23 studies, 36,452 participants; high-certainty evidence), when compared to 0.9% saline. Based on a mortality rate of 147 people per 1000, buffered solutions could reduce the number of deaths by 13 per 1000 or could increase deaths by 1 per 1000. We found that buffered solutions likely result in little to no difference in acute renal injury (OR 0.87, 95% CI 0.75 to 1.02; I² = 51%; 17 studies, 30,832 participants; moderate-certainty evidence). We downgraded the certainty of the evidence because of the risk of bias. Based on an acute renal injury rate of 140 per 1000, buffered solutions could reduce acute renal injury by 31 per 1000 or could increase acute renal injury by 2 per 1000. We are very uncertain of the effects of buffered solutions versus 0.9% saline on organ system dysfunction (OR 0.83, 95% CI 0.41 to 1.70; I² = 0%; 5 studies, 266 participants; very low certainty evidence), and on sodium (MD -0.26, 95% CI -2.29 to 1.77; I² = 79%; 7 studies, 1246 participants; very low certainty evidence) and potassium (MD 0.11, 95% CI -0.04 to 0.25; I² = 41%; 5 studies, 1086 participants; very low certainty evidence). Compared to 0.9% saline, buffered solutions may reduce chloride (MD -2.39, 95% CI -3.77 to -1.00; I² = 90%; 11 studies, 1981 participants), and may increase pH (MD 0.06, 95% CI 0.02 to 0.10; I² = 88%; 6 studies, 1224 participants) and bicarbonate (MD 2.16, 95% CI 1.06 to 3.25; I² = 87%; 9 studies, 1368 participants) (all low-certainty evidence). We downgraded the certainty of the evidence because of the risk of bias and imprecision.

Buffered solutions do not reduce overall (in-hospital) mortality compared to 0.9% saline solution in critically ill patients, and probably do not reduce acute renal injury. Evidence for organ system dysfunction and electrolyte disturbances is of low or very low certainty. We have high-certainty evidence about the outcome of mortality, but further trials are needed to clarify the impact of buffered solutions on acute renal injury and other outcomes. Future studies should involve underrepresented populations (paediatric, neurocritical, female) and adopt standardised, patient-centred outcome measures to broaden the evidence base. Once the 38 relevant ongoing studies are published and the nine studies that await classification are evaluated, the inclusion of new studies in this review may alter its conclusions regarding acute renal injury, organ dysfunction, and electrolyte disturbances.

The original review and this update received no funding.

This 2026 review is an update of the 2019 review. Both versions were conducted according to the published protocol. Protocol (2016) available at https://doi.org/10.1002/14651858.CD012247 The protocol was registered with PROSPERO (CRD42016045988). Original review (2019) available at https://doi.org/10.1002/14651858.CD012247.pub2.