Reply Article Swipe

Potential conflict of interest: Nothing to report. We read with interest the correspondence from Amarapurkar and Lo. We thank Dr. Amarapurkar et al. for their comments. We certainly agree that, based on the evidence, there is no need to correct coagulopathy in patients with liver cirrhosis undergoing invasive procedures. Dr. Amarapurkar's correspondence states that correction of coagulopathy is not considered standard of care, when previous guidelines (referenced in our article1) and current guidelines2 recommend such correction. This is the reason why we sought to compare whether thromboelastography (TEG)‐guided correction was associated with a lower transfusion of blood products than correction based on international normalized ratio (INR) and platelet count (PLT). We were surprised to find that even in our cohort of patients with severe coagulopathy (INR, >1.8; PLT, <50,000/mm3) and with a median Model for End‐Stage Liver Disease score of 21, bleeding occurred in only 1 of 60 (2%) patients. It was until we completed the study that we realized that this level of coagulopathy does not define bleeding risk in patients with cirrhosis and should not be corrected.1 In the study mentioned in the correspondence3 where coagulopathy was more conservatively defined (INR, >1.5; PLT, <50,000/mm3), 3 of 128 (2%) that had had a high‐risk procedure developed bleeding, but this bleeding risk was not significantly higher than patients without coagulopathy. Interestingly, compared to patients with coagulopathy who did not bleed, those who bled also had acute kidney injury and/or sepsis as comorbidities and this may represent the difference. Whether correction of coagulopathy in these settings leads to lower bleeding complications remains to be determined.3 These results, together with ours, suggest that, in the absence of comorbidities, the risk of bleeding associated with invasive procedures is quite low and that prophylactic correction of coagulopathy is not warranted. It may be more cost‐effective to correct coagulopathy during or after the procedure if and when bleeding occurs. In our opinion, there is no need to perform a study without prophylactic correction of routine coagulation tests as was suggested by Intagliata et al.,4 because the TEG‐guided transfusion policy in our study required transfusion in only 5 of 30 patients. In our view, there is instead a strong need of prospective, well‐designed studies analyzing predictors of bleeding beyond coagulation tests and the effect of correction of blood products in high‐risk patients. In conclusion, we agree with Dr. Amarapukar on avoiding the widespread use of prophylactic blood products transfusion based solely on INR and PLT count.