Low-Dose vs. Standard Care Iv Human Albumin During Large-Volume Paracentesis in Patients With Liver Cirrhosis: A Systematic Review

Resumo original em inglês

O que aconteceu

Cirrhosis is a global health concern, with ascites being the most prevalent decompensating event. Intravenous (IV) human albumin (HA) is administered during large-volume paracentesis (LVP) to prevent complications. The standard (6-8 g HA/L ascites) is empirically chosen, potentially leading to overtreatment and added costs. We reviewed evidence comparing the efficacy and safety of reduced versus standard-care HA dosing during LVP. We searched PubMed, EMBASE, Cochrane Library, CINAHL, and Scopus (up to January 31, 2025). We included RCTs and cohort studies in adults with cirrhosis undergoing LVP, comparing low-dose (≤ 6 g/L or ≤ 20 g total) to standard-dose (> 6 g/L or > 20 g total) HA. Outcomes included paracentesis-induced circulatory dysfunction (PICD), renal impairment, hyponatremia, other decompensating events, and survival. Quality assessment used Cochrane risk-of-bias tools and Newcastle-Ottawa Scale. Five studies (three full-text, two abstracts) met inclusion criteria. Low-dose regimens (2-6.5 g/L) showed no significant difference compared to standard care (6-8.3 g/L) in PICD (reported in two studies) and renal dysfunction (reported in five studies). Other secondary outcomes, including hyponatremia, GI bleeding, HE, infections, and reaccumulating ascites, also showed no statistical differences. While these findings indicate potential for dose reduction, definitive confirmation via meta-analysis was not feasible due to methodological heterogeneity. Available evidence suggests that reduced-dose IV HA (≤ 6 g/L) is as effective and safe as standard-dose HA (6-8 g/L) during LVP. However, the quality of the current evidence is limited by small sample sizes and suboptimal study designs, and well-designed, adequately powered future studies are required to confirm these findings.

Resumo publicado sob licença CC BY. A tradução para o português ainda não chegou a este estudo.