Prevention of paracentesis‐induced circulatory dysfunction—A systematic review and network meta‐analysis

Background Paracentesis‐induced circulatory dysfunction (PICD) is seen in up to 70% of patients undergoing large volume paracentesis (LVP). We performed comparative effectiveness of available pharmacological interventions for the prevention of PICD in cirrhotic patients undergoing LVP. Methods A comprehensive search of electronic databases from inception to March 2019 was carried out. Adult patients aged > 18 years diagnosed as cirrhosis of liver undergoing LVP comparing active pharmacological interventions with each other or placebo for prevention of PICD (defined by a rise in plasma renin activity by ≥ 50% of baseline or by ≥ 4 ng/ml/h on the fifth or sixth day after LVP) were included. Direct and Bayesian network meta‐analysis (NMA) was performed to assess all the available pharmacological interventions. We followed GRADE (Grading of Recommendations Assessment, Development and Evaluation system) criteria for appraising confidence in estimates. Results Nine RCTs including 620 cirrhotic patients undergoing LVP were included for final meta‐analysis after screening 448 studies. Seven interventions were compared with a common comparator (albumin 8g/L) for the prevention of PICD (n = 605 patients). In a direct meta‐analysis, albumin (8g/L) was superior only to other volume expanders but not to vasoconstrictors. However, in NMA, low‐quality evidence (Grade‐⊕⊕) supported midodrine given for 3 days as superior to albumin (8g/L) in the prevention of PICD (surface under the cumulative ranking curve (SUCRA): midodrine 3 days‐95.05%; albumin 8g/L −73.21%). The cost of albumin therapy was 306.9$, and midodrine was 29.4$. Conclusion Midodrine for 3 days is economical and superior to albumin in the prevention of PICD in cirrhosis patients undergoing LVP.