Intravenous fluid therapy in adults in hospital

What the guideline recommends

NICE guideline CG174 recommends to offer intravenous (IV) fluid therapy as part of a protocol using the following algorithms:

Methods

The Australasian Resuscitation in Sepsis Evaluation (ARISE) investigators and the Australian and New Zealand Intensive Care Society (ANZICS) Clinical Trials Group (2014) conducted a randomised controlled trial (RCT) to test the hypothesis that EGDT, compared with usual care, would decrease 90‑day mortality in people with early septic shock at the emergency department (n=1,600 participants). Adults were eligible if infection was suspected or confirmed, if they met 2 or more criteria for a systemic inflammatory response, and if there was evidence of refractory hypotension or hypoperfusion. Exclusion criteria were contraindication to central venous catheter insertion in the superior vena cava or to receiving blood products; haemodynamic instability due to active bleeding; underlying disease process with a life expectancy <90 days; death deemed imminent and inevitable; documented limitation of therapy order restricting implementation of the study protocol or aggressive care deemed unsuitable by the treating clinician; in-patient transfer from another acute health care facility; confirmed or suspected pregnancy; inability to commence EGDT within 1 hour of randomisation or deliver EGDT for 6 hours. The study was conducted in 51 tertiary care and non-tertiary care hospitals in Australia, New Zealand, Finland, Hong Kong, and the Republic of Ireland. At time of selection, none of these hospitals had sepsis-resuscitation protocols or resuscitation guided by measurement of the central venous oxygen saturation (ScvO₂). The intervention was provided by a study team trained in EGDT delivery including the insertion of an arterial catheter and a central venous catheter capable of continuous ScvO₂ measurement within 1 hour after randomisation and a resuscitation algorithm followed until 6 hours after randomisation. The control (usual care) was provided by the treating clinical team. ScvO₂ measurement was not permitted in the control group during the 6 hours after randomisation. The primary outcome was death from any cause within 90 days after randomisation. All analyses were conducted according to the intention-to-treat principle.

Results

The intention-to-treat population included 793 participants in the EGDT group and 798 participants in the usual care group.

The primary outcome of death from any cause within 90 days after randomisation was:

Strengths and limitations

Impact on guideline

This RCT provided evidence that EDGT does not seem to be better than usual care for resuscitation in people with early septic shock. EGDT was reviewed during guideline development but no recommendations were developed regarding EGDT. This new evidence was considered to be unlikely to impact on the guideline because EGDT does not seem to have any benefit on the management of intravenous fluid therapy in hospitalised adult patients. Guideline committee members commented during this surveillance review that EGDT was discussed by the committee at the time of the initial development of the guideline. The conclusion at that time was that EGDT did not seem to have any benefit on the management of IV fluid therapy. They also said that this RCT appeared to support that discussion and offered no evidence to refute that assertion. It was highlighted that this RCT did not consider some of the measures that were considered as potential factors or complications with this approach and further evidence to explore EGDT's effect in relation to further complications would be beneficial before revising this element of the guideline.

What the guideline recommends

NICE guideline CG174 recommends using crystalloids that contain sodium in the range 130 mmol/litre to154 mmol/litre, with a bolus of 500 ml over less than 15 minutes if patients need IV fluid resuscitation. NICE guideline CG174 also recommends that tetrastarch for fluid resuscitation should not be used and to consider human albumin solution 4% to 5% for fluid resuscitation only in patients with severe sepsis.

Methods

Rochwerg et al. (2014) conducted a systematic review and network meta-analysis of 14 RCTs (n=18,916 participants) to assess the effect of fluid resuscitation on mortality in people with severe sepsis and septic shock. Parallel-group and factorial RCTs were eligible for inclusion. Quasi-randomised and crossover trials were excluded. The authors also noted that all studies published by Dr Joachim Boldt were excluded because of suspected lack of integrity. Further exclusions were studies in which most participants had the systemic inflammatory response syndrome secondary to other causes, without a clear sepsis subgroup and those focusing on people after elective surgery. There were no restrictions on language or date of publication. The inclusion criteria for participants in the RCTs were critically ill people with severe sepsis or septic shock (aged 16 years or older). Studies were included if there was available data on people with sepsis when mixed critically ill populations were studied. Any fluid or fluid strategy for resuscitation compared with other fluid or fluid strategy were the interventions. The primary outcome was 90‑day mortality. If this outcome was not available, the longest mortality was taken as the outcome either 30‑day, intensive care unit, or hospital mortality. Fluids were classified for analysis as:

There were 2 network meta-analyses (NMA) and a direct fixed-effects meta-analysis:

Results

Fourteen RCTs were included. In 4 of these, people with sepsis were studied as part of a subgroup analysis. The main results are shown below.

Strengths and limitations

Impact on guideline

This evidence supports recommendations 1.3.1 to 1.3.3 for the use of crystalloids and human albumin for intravenous fluid resuscitation and that tetrastarch should not be used for fluid resuscitation. Topic experts highlighted that the NMA included a discussion about the use of balanced crystalloids compared to unbalanced crystalloids but without a definitive conclusion to suggest a change to the original recommendations in NICE guideline CG174.

What the guideline recommends

During guideline development, the scope of the guideline was limited by excluding patients with more specialised fluid prescription needs. One of the exclusions was intensive monitoring in patients under intensive care.

Methods

Young et al. (2015) conducted a 28‑week double-crossover RCT (n=2,278 participants) to compare the effectiveness of 2 types of fluid therapies: a buffered crystalloid and a saline crystalloid at the intensive care unit (ICU). Eligible participants were all ICU patients receiving crystalloid fluid therapy as clinically indicated. Patients were excluded if they were on renal replacement therapy (RRT), or expected to require RRT within 6 hours, or if they were admitted to the ICU only for organ donation, or for palliative care, or those who were previously enrolled in the study. The participant ICUs were adult or mixed (adult and paediatric) general medical and surgical ICUs (3 ICUs). A fourth ICU had a predominance of cardiothoracic and vascular surgical patients. The interventions were 0.9% saline crystalloid (saline) and Plasma-Lyte 148 (PL‑148) which was the buffered crystalloid. Interventions were used in alternating treatment blocks of 7 weeks. The double-crossover occurred in each ICU using each study fluid twice over the 28 weeks of the study. The primary outcome was the proportion of patients with acute kidney injury (AKI).

Results

The primary outcome of AKI was reported within 90 days after enrolment:

Strengths and limitations

Impact on guideline

This RCT was considered to be relevant because acute kidney injury (primary outcome) was one of the outcomes considered in NICE guideline CG174. However, the scope of NICE guideline CG174 does not cover intravenous fluid therapy for patients in intensive care units. This new evidence was considered to be unlikely to impact on the guideline because none of the fluids demonstrated superiority. Topic experts highlighted that this RCT would have been considered as indirect evidence during guideline development and also referred to an editorial by Kellum and Shaw (2015) highlighting several shortcomings of this RCT.