Expert opinion on bleeding risk from invasive procedures in cirrhosis

Abstract

Background & aims: Despite several recent international guidelines, no consensus exists on the bleeding risk nor haemostatic parameter thresholds that define the safety of invasive procedures in patients with cirrhosis. The aim of this study was to establish a position paper on the bleeding risk associated with invasive procedures in patients with cirrhosis among the experts involved in various guidelines.

Methods: All experts involved in recent guidelines on the management of invasive procedures in patients with cirrhosis were invited to classify 80 procedures as "high risk" or "low risk" with respect to bleeding. Procedures were considered high risk when the estimated risk of major bleeding was 1.5% or more, or when even minor bleeding might lead to significant morbidity or death. The experts were also asked to choose safety thresholds for laboratory test values at which elective invasive procedures could be safely performed. The predetermined threshold considered as "consensus" was ≥75% agreement.

Results: Fifty-two experts participated in the study. Out of 80 procedures, a consensus opinion was reached for 52 procedures (65%): 17 procedures were classified as "high risk", primarily interventional endoscopic procedures, percutaneous organ biopsies, or procedures involving the central nervous system; and 35 as "low risk", primarily "diagnostic" procedures. The lowest platelet counts at which performance of a low-risk procedure or a high-risk procedure/surgery were deemed acceptable were 30 × 10⁹/L and 50 × 10⁹/L, respectively. Experts did not believe that international normalised ratio should be considered before performing low-risk procedures; 71% also indicated that it should not be considered before performing high-risk procedures.

Conclusions: This experience-based classification may be helpful to refine future study designs and to guide clinical decision making regarding invasive procedures in patients with cirrhosis.

Impact and implications: Several risk classifications and management guidelines for invasive procedures in patients with cirrhosis have been proposed, but with conflicting recommendations. By providing a position paper, based on the opinion of a broad panel of experts, on the bleeding risk associated with 52 invasive procedures in patients with cirrhosis, this survey will help to provide a framework for future study design. The consensus on platelet count, international normalised ratio, fibrinogen and activated partial thromboplastin time identified in this survey will inform physicians regarding the laboratory test values considered acceptable by the experts prior to the performance of an elective invasive procedure in patients with cirrhosis.

Keywords: INR; aPTT; anticoagulant; biopsy; coagulation; fibrinogen; haemorrhage; haemostasis; platelet; procedural related bleeding.

Graphical abstract

Fig. 1

Summary of the position paper process.

Fig. 2

Classification of the bleeding risk associated with invasive procedures. Green and purple colours indicate when a consensus (≥75% agreement) was reached for low-risk or high-risk procedures, respectively. White colour indicates that a consensus was not reached.

Fig. 3

Results of the survey for laboratory tests thresholds considered safe prior to low-risk and high-risk procedure or high-risk surgery. Percentage of agreement was calculated without taking into account the experts who answered “I don’t know”. To define a threshold, the responses "I do not recommend this parameter to judge bleeding risk in this setting in patients with cirrhosis" were considered as any of the blood test values. aPTT, activated partial thromboplastin time; INR, international normalised ratio.

Fig. 4

Proposed algorithm based on the results of the present survey to stratify and manage bleeding risk following invasive procedures and surgeries in patients with cirrhosis. ∗If the patient has reasonably recent laboratory test results, re-measuring might not be needed. ∗∗Assessing INR to predict procedure-related bleeding may rely more on customary practices than on recent data regarding INR in cirrhosis. Moreover, recent guidelines recommended not to correct a prolonged INR (BSG 2020, ACG 2020, AASLD 2021, AGA 2021, ISTH 2022, EASL 2022). ∗∗∗ The decision to proceed with surgery must carefully balance the patient's individual risks, such as the severity of liver disease and any comorbidities, against the potential harm of foregoing the procedure. INR, international normalised ratio; TPO, thrombopoietin.