In adult patients with suspected sepsis, is resuscitation with 20% human albumin solution associated with improved patient-oriented outcomes?
Date First Published:
March 10, 2026
Last Updated:
March 10, 2026
Report by:
Dr Harriet Phelan, Foundation Year 1 Doctor (Shrewsbury and Telford Hospital NHS Trust)
Search checked by:
Dr Cameron Davie, Foundation Year 2 Doctor, Mersey and West Lancashire Teaching Hospitals NHS Trust
Three-Part Question:
In adult patients with suspected sepsis is resuscitation with 20% human albumin solution associated with improved patient-oriented outcomes, such as mortality, length of stay in hospital and Intensive Care Unit (ICU) admission?
Clinical Scenario:
A 65-year-old patient presents to the ED with lethargy and pyrexia. They are haemodynamically unstable and you suspect they have sepsis. While initiating the local sepsis treatment pathways, you wonder if additional resuscitation with human albumin solution (HAS) will improve their outcomes?
Search Strategy:
[Sepsis.mp. OR exp Sepsis/], [Albumin.mp. OR exp Albumins/], AND [(emergency Department.mp. OR emergency room.mp. OR emergency care.mp. OR emergency ward.mp) OR (exp emergency care/ OR exp emergency ward/)].
Search Details:
EMBASE (1974 to February 2026) and Ovid MEDLINE (1936 to February 2026) data- bases were searched using the Ovid interface with the above keyword strategy. No search filters were used. A supplementary search of the Cochrane database was conducted using the same search terms. The citation lists of relevant papers were screened for any studies missed by our search paradigm.
Outcome:
A total of 844 papers were identified in the initial database search, with 81 duplicates. 111 were removed due to being conference abstracts or conference papers. Following abstract and title review, 643 papers were removed after application of the inclusion or exclusion criteria. 9 papers remained, of which 6 were removed after full text review following application of the criteria. 3 papers were retained for full analysis. 3 papers were identified using the Cochrane database, of which 1 paper was retained for full analysis following application of the criteria and removal of duplicates. 2 additional authors were contacted after their papers were identified as applicable to the study, however no response was received and full texts were not available.
In total, four papers were retained for final analysis. These included: one single-site randomized controlled trial (1), one multi-centre randomized feasibility trial (2), one retrospective single-site cohort study (3), and one, double-blind randomized, controlled multi-site trial (4). The results are summarised in Table 1.
In total, four papers were retained for final analysis. These included: one single-site randomized controlled trial (1), one multi-centre randomized feasibility trial (2), one retrospective single-site cohort study (3), and one, double-blind randomized, controlled multi-site trial (4). The results are summarised in Table 1.
Relevant Paper(s):
| Study Title | Patient Group | Study type (level of evidence) | Outcomes | Key results | Study Weaknesses |
|---|---|---|---|---|---|
| Intervention With Concentrated Albumin for Undifferentiated Sepsis in the Emergency Department (ICARUS-ED): A Pilot Randomized Controlled Trial Williams, J.M. et al 2025 Australia & NZ | n = 463 20% albumin administered (400ml / 4hrs) Mean age: 64.1 years (albumin) 66.2 years (no albumin) Male: 62.9% (albumin) 65.4% (no albumin) | Parallel-group, open-label, randomized controlled trial Single site |
Total fluid therapy (including albumin) at 6, 24, 72 hrs, ICU admission from ED, ICU length of stay (LOS), 30-day Mortality | Intervention group received less fluid overall - statistically significant only after 72hr (4275mL [intervention], 5000mL [control]; 95% CI -1284 to -116mL). | The amount of balanced crystalloid given was at clinician discretion, study participants at risk of performance bias and differential treatment. The interventions and analysis were unblinded; results liable to observer and confirmation bias. Single-site data – has been difficult to replicate in equivalent multi-centre trials. |
| No significant reduction in ICU admission (51% [control] vs 50% [intervention], 95% CI -8.1 to 7.1). | |||||
| No significant difference in median LOS (1.9d [intervention], 2.2d [control]; 95% CI -0.9 to 1.4d). | |||||
| No significant difference in 30d mortality (38% [intervention] vs 36% [control], 95% CI -6.5 to 8.1%). | |||||
| Albumin Versus Balanced Crystalloid for the Early Resuscitation of Sepsis: An Open Parallel-Group Randomized Feasibility Trial- The ABC-Sepsis Trial Gray, A.J. et al. 2024 United Kingdom | n = 300 Patients treated with 5% HAS or balanced crystalloid (BC) Mean age (SD) = 69 years old (± 16 years) Male = 151 (50%) | Open, parallel-group randomized feasibility trial Multi-centre |
30-day Mortality, admission to a critical care bed (HDU or ICU), fluid administration in 6h, hospital length of stay | No statistically significant difference in 30-day mortality rate (31% [HAS] vs 22% [BC]. | 22% of participants in HAS arm received crossover fluid for resus - this may have impacted the results. Trial open-label & non-blind; results liable to observer and confirmation bias. Amount of HAS and crystalloid fluid given at clinician discretion, study participants at risk of performance bias and differential treatment. Intervention only mandated for 6hrs - this could be too short for resuscitation. |
| 15% of the cohort who received HAS required HDU/ICU admission, compared to 11% of the cohort who received BC. The HAS group received 1100ml fluid (IQR 600-1600ml) vs 1358ml (IQR 1000-2069ml) in the BC group. | |||||
| Median hospital LOS 6 days for both groups. | |||||
| Effect of albumin on the outcomes in septic patients with hypoalbuminemia in the emergency department: a propensity score-matched retrospective cohort study Hwang, J.E. et al. 2022 Korea | n = 384 Patients treated with albumin or control fluids Mean age = 76yrs Male = 59.2% | Retrospective, propensity score-matched cohort study Single site |
28-day mortality, ICU admission | No statistically significant difference in 28-day mortality (38.3% [Albumin] vs 33.3% [control]; p=0.236). | This is a retrospective study and is at risk of allocation bias. Inclusion criteria changed mid-study due to definitions of sepsis and septic shock changing at the time - potential inconsistencies in the study population. Amount of HAS and crystalloid fluid given at clinician discretion. |
| Statistically significant increase in the rate of admission to ICU in the albumin-treated group (59.9% compared with 41.7% [control]; p<0.001). | |||||
| Fluid resuscitation with 5% albumin versus normal saline in early septic shock: a pilot randomized, controlled trial McIntyre, L.A. et al. 2012 Canada | n = 50 Patients treated with 5% human albumin solution or 0.9% NaCl Mean age = 64.5yr Male = 44% | Concealed, double-blind, randomized, controlled trial Multi-centre |
28-day mortality, median length of hospital stay | Mortality 28% (unspecified). Median LOS 10 days (unspecified) | The study pool is much smaller than the other trials in this review and results may not be applicable to a larger trial. The amount of fluid used was not reported which could have impacted any results from the trial. Feasibility trial – the results were not discussed but more focused on whether clinicians are able to carry out a double-blind randomised trial on larger scale. Statistical analysis also not carried out on the outcomes. |
Author Commentary:
Treatment with HAS did not significantly reduce mortality compared to treatment with crystalloid fluid (1, 2, 4). One trial reported a significant difference in the length of hospital stay, of which the group treated with albumin spent longer in hospital (1). ICU admission rate results were inconclusive, with no significant evidence supporting use of human albumin solution for this outcome. Two trials noted that patients treated with HAS required less fluid administration overall (1, 2) but did not comment on how this related to improved outcomes.
The limited available evidence from this review does not support using HAS for resuscitation in patients with suspected sepsis in the ED. Key patient-oriented outcomes assessed in this review were not significantly improved when compared to treatment with crystalloid, and any small improvement in outcome is not justifiable given the economic impact of implementing this practice. Further study is required on resuscitation with HAS in the ED.
Encouragingly, a large-scale, multi-site, double blind, randomised controlled trial addressing this three-part study has been demonstrated as feasible by one of the trials included in this review (4). Future studies should compare treatment with 20% HAS, 5% HAS and 0.9% NaCl.
The limited available evidence from this review does not support using HAS for resuscitation in patients with suspected sepsis in the ED. Key patient-oriented outcomes assessed in this review were not significantly improved when compared to treatment with crystalloid, and any small improvement in outcome is not justifiable given the economic impact of implementing this practice. Further study is required on resuscitation with HAS in the ED.
Encouragingly, a large-scale, multi-site, double blind, randomised controlled trial addressing this three-part study has been demonstrated as feasible by one of the trials included in this review (4). Future studies should compare treatment with 20% HAS, 5% HAS and 0.9% NaCl.
Bottom Line:
Administration of HAS in the emergency department for patients with suspected sepsis is not recommended. The limited studies available suggest this intervention may be associated with worsened patient outcomes overall.
Level of Evidence:
Level 2: Studies considered were neither 1 or 3
References:
- Williams, J.M. et al. Intervention With Concentrated Albumin for Undifferentiated Sepsis in the Emergency Department (ICARUS-ED): A Pilot Randomized Controlled Trial
- Gray, A.J. et al.. Albumin Versus Balanced Crystalloid for the Early Resuscitation of Sepsis: An Open Parallel-Group Randomized Feasibility Trial- The ABC-Sepsis Trial
- Hwang, J.E. et al.. Effect of albumin on the outcomes in septic patients with hypoalbuminemia in the emergency department: a propensity score-matched retrospective cohort study
- McIntyre, L.A. et al.. Fluid resuscitation with 5% albumin versus normal saline in early septic shock: a pilot randomized, controlled trial