RATIONALE: People with liver cirrhosis are at an increased risk of bacterial infections, with a high rate of complications and mortality. Adding albumin to antibiotics may reduce their occurrence.

OBJECTIVES: To assess the benefits and harms of intravenous administration of human albumin in people with liver cirrhosis and bacterial infections versus no intervention, placebo, or other intravenous fluids.

SEARCH METHODS: We identified randomised clinical trials (RCTs) through electronic searches in the Cochrane Hepato-Biliary Group Controlled Trials Register, CENTRAL, MEDLINE, Embase, LILACS, Science Citation Index Expanded, and Conference Proceedings Citation Index-Science. We searched online clinical trial registries for unpublished or ongoing trials and checked reference lists for additional trials (latest search date: 20 August 2025).

ELIGIBILITY CRITERIA: RCTs comparing albumin versus no intervention, placebo, or other intravenous fluids, in people with infection in cirrhosis.

OUTCOMES: Critical outcomes: all-cause mortality, serious adverse events, and health-related quality of life. Among the important outcomes were kidney impairment, acute liver decompensation, acute-on-chronic liver failure, and septic shock.

RISK OF BIAS: We used the Cochrane risk of bias (RoB) 2 tool.

SYNTHESIS METHODS: We used standard Cochrane methods. We meta-analysed outcome data at the longest follow-up. We presented dichotomous outcomes as risk ratios (RR) or odds ratios (OR), with 95% confidence intervals (CI), and a random-effects model. We assessed the certainty of evidence with GRADE. We used 0.25 minimal important difference and optimal information size to assess imprecision.

INCLUDED STUDIES: We included 11 parallel-group RCTs (published between 1999 and 2022), with 1273 adults with cirrhosis and bacterial infections. One trial was a multicentre trial conducted in eight European countries, and ten were conducted in India, Spain, China, France, Egypt, and Taiwan. The trials' settings were clinics or university hospitals. Trial participants had spontaneous bacterial peritonitis (n = 388; five trials), infections other than spontaneous bacterial peritonitis (n = 477; four trials), and infections in different sites and sepsis-induced hypotension (n = 408; two trials). Intravenous albumin was compared with no intervention (eight trials) and other intravenous fluids such as plasmalyte, normal saline, or hydroxyethyl starch (three trials). All participants received antibiotics. The follow-up was four to 90 days. All trials provided data for intention-to-treat analysis.

SYNTHESIS OF RESULTS: Below, we present only the outcome results displayed in our Summary of Findings tables. Risk of bias For overall risk of bias, most of the outcomes had some concerns or a high risk of bias. This was due to missing information regarding the randomisation process, deviations from the intended interventions, missing outcome data, measurement of the outcome, or bias in the selection of the reported result. GRADE We downgraded the evidence to low or very low due to risk of bias and imprecision, meaning that the evidence was either uncertain or very uncertain. This was mainly due to a high or unclear risk of bias or due to imprecision (wide confidence intervals crossing thresholds of interest, or optimal information size not reached). Albumin versus no intervention There may be little to no difference between albumin and no intervention regarding all-cause mortality (RR 0.80, 95% CI 0.53 to 1.20; I² = 45%; 8 studies, 845 participants; very low-certainty evidence), but the evidence is very uncertain. The evidence is also very uncertain about the effect of albumin on serious adverse events (RR 3.15, 95% CI 0.39 to 25.15; I² = 21%; 3 studies, 411 participants; very low-certainty evidence) and about whether it may reduce the risk of kidney impairment (RR 0.51, 95% CI 0.28 to 0.93; I² = 45%; 7 studies, 768 participants; very low-certainty evidence). The evidence is either uncertain or very uncertain about the effect of albumin on acute liver decompensation (RR 1.87, 95% CI 0.36 to 9.81; I² not applicable; 1 study, 118 participants; very low-certainty evidence), on acute-on-chronic liver failure (RR 0.86, 95% CI 0.43 to 1.73; I² not applicable; 1 study, 118 participants; low-certainty evidence), and on septic shock (RR 1.07, 95% CI 0.41 to 2.76; I² not applicable; 1 study, 118 participants; low-certainty evidence). Albumin versus other intravenous fluids The evidence is very uncertain about the effect of albumin on all-cause mortality (RR 1.05, 95% CI 0.85 to 1.29; I² = 0%; 3 studies, 428 participants; very low-certainty evidence). Albumin may increase serious adverse events (Peto OR 9.25, 95% CI 2.66 to 32.18; I² not applicable; 2 studies, 408 participants; very low-certainty evidence), but the evidence is very uncertain. There may be little to no difference between albumin versus other intravenous fluids for kidney impairment (RR 0.77, 95% CI 0.53 to 1.10; I² = 0%; 2 studies, 120 participants; very low-certainty evidence), but the evidence is very uncertain. Albumin may reduce the risk of septic shock (RR 0.91, 95% CI 0.85 to 0.97; I² = 0%; 2 studies, 408 participants; low-certainty evidence), but the evidence is uncertain. No trial reported data on quality of life, acute liver decompensation, or on acute-on-chronic liver failure. Seven trials were without for-profit support, one trial declared partial private funding, and three trials did not declare sources of support. We identified three ongoing trials.

AUTHORS' CONCLUSIONS: Due to the very low certainty and low certainty of evidence, we are either uncertain or very uncertain about the effect of albumin, compared with no intervention, on all-cause mortality, serious adverse events, kidney impairment, acute liver decompensation, acute-on-chronic liver failure, and septic shock. Due to very low certainty of evidence, we are very uncertain about the effect of albumin, compared with other intravenous fluids, on all-cause mortality or kidney impairment. The evidence is also very uncertain whether albumin increases serious adverse events. Due to low-certainty evidence, we are uncertain whether albumin reduces septic shock. No trial in this comparison reported data on acute liver decompensation and acute-on-chronic liver failure. Data on health-related quality of life were lacking. Though clinically relevant, we could not assess the effect of albumin in regard to the site and stage of infection due to insufficient data. We lack trials in children.

FUNDING: None.

REGISTRATION: Protocol available via 10.1002/14651858.CD014636.