Cost Effectiveness of Early Insertion of Transjugular Intrahepatic Portosystemic Shunts for Recurrent Ascites

Abstract

Treatment options for recurrent ascites resulting from decompensated cirrhosis include serial large-volume paracentesis and albumin infusion (LVP+A) or insertion of a transjugular intrahepatic portosystemic shunt (TIPS). Insertion of TIPSs with covered stents during early stages of ascites (early TIPS, defined as 2 LVPs within the past 3 weeks and <6 LVPs in the prior 3 months) significantly improves chances of survival and reduces complications of cirrhosis compared with LVP+A. However, it is not clear if TIPS insertion is cost effective in these patients. We developed a Markov model using the payer perspective for a hypothetical cohort of patients with cirrhosis with recurrent ascites receiving early TIPSs or LVP+A using data from publications and national databases collected from 2012 to 2018. Projected outcomes included quality-adjusted life-year (QALY), costs (2017 US dollars), and incremental cost-effectiveness ratios (ICERs; $/QALY). Sensitivity analyses (1-way, 2-way, and probabilistic) were conducted. ICERs less than $100,000 per QALY were considered cost effective. In base-case analysis, early insertion of TIPS had a higher cost ($22,770) than LVP+A ($19,180), but also increased QALY (0.73 for early TIPSs and 0.65 for LVP+A), resulting in anICER of $46,310/QALY. Results were sensitive to cost of uncomplicated TIPS insertion andtransplant, need for LVP+A, probability of transplant, and decompensated QALY. In probabilistic sensitivity analysis, TIPS insertion was the optimal strategy in 59.1% of simulations. Based on Markov model analysis, early placement of TIPSs appears to be a cost-effective strategy for management of specific patients with cirrhosis and recurrent ascites. TIPS placement should be considered early and as a first-line treatment option for select patients.