PurposeTo determine the rate of platelet transfusion in patients with cirrhosis and severe thrombocytopenia (platelet counts <50 × 109/L) undergoing high-risk invasive procedures when prescribed by thromboelastography (TEG) compared with empirical and on-demand transfusion strategies. Materials and MethodsThis was a single-center, single-blinded, randomized controlled trial. Patients with cirrhosis and severe thrombocytopenia undergoing high-risk invasive procedures were randomized into 3 groups: TEG group, transfusions based on TEG parameters; standard of care (SOC) group, 3 units of random donor platelets before procedure; and on-demand group, transfusions based on procedural adverse events/clinician’s discretion. The primary outcome was periprocedural platelet transfusion in each arm. ResultsEighty-seven patients were randomized (29 in each group) with no significant differences in demographics/coagulation profile/procedures. The median platelet count was 33 × 109/L (interquartile range, 26–43 × 109/L). Percutaneous liver biopsy was the most common procedure (46, 52.9%). Significantly lower number of patients in the TEG group received platelets (4 cases, 13.8%; 95% CI, 3.9–31.7) compared with SOC group (100%; 95% CI, 88.1–100; P < .001). Four patients in the on-demand group received platelets (13.8%; 95% CI, 3.9–31.7). Minor (World Health Organization [WHO] Grade 2) procedure-related bleeding occurred in 3 (10%; 95% CI, 2.2–27.4) patients in the TEG-guided transfusion group compared with 1 (3.4%; 95% CI, 0.1–17.8) patient each in the SOC and on-demand groups (P = .43), although the study was not powered for comparison of bleeding rates. No bleeding-related mortality was observed in any of the 3 groups. ConclusionsTEG-prescribed transfusion reduced prophylactic transfusions in patients with cirrhosis and severe thrombocytopenia undergoing high-risk invasive procedures. The study was not powered for comparison of bleeding rates.