Thromboelastography Phenotypes and Clinical Outcomes in Gastrointestinal Bleeding

Rapid thromboelastography (r-TEG) may be useful for guiding transfusion needs and prognosis of emergency department (ED) patients with gastrointestinal bleeding (GIB). However, guidance is needed on how to classify r-TEG values for their use in clinical decision-making. Using two different methods to classify r-TEG values according to coagulation phenotypes, we assessed clinical outcomes among adult ED patients with GI bleeding who were not taking anti-coagulants.

We collected r-TEG and clinical outcome data on 193 non-anticoagulated adult ED patients who presented with a GI bleed. Using r-TEG values (ACT, K time, α-angle, MA), we classified r-TEG coagulation phenotypes according to two different methods. Method 1 phenotypes were defined as normal, hypo- or hypercoagulable according to our hospital’s r-TEG normal range cut-offs. Method 2 phenotypes were based on our hospital’s trauma transfusion guidelines: normal, hypocoagulable if meeting transfusion thresholds, hypercoagulable if above our upper range cut-offs. Mixed phenotypes had conflicting r-TEG values. Primary clinical outcomes measured were in-hospital and 30-day mortality, 5-day rebleeding; secondary outcomes were ICU admission, 24-hour transfusion, total transfusion, and length of stay (LOS). We assessed clinical outcomes by phenotype categorization method using X2 testing for categorical and Kruskal-Wallis for continuous outcomes.

Method 1 classified 38.3% of patients as normal, 23.3% hypercoagulable, 30.6% hypocoagulable, and 7.8% as mixed. Method 2 classified 51.3% as normal, 28.0% hypercoagulable, 17.6% hypocoagulable, and 3.1% as mixed. As compared to their respective normal phenotypes, Method 1’s abnormal phenotypes were associated with greater occurrence of rebleeding (p<0.004), blood product use (p<0.03), ICU admission (p< 0.02), and longer LOS (p=0.02), while Method 2’s abnormal phenotypes were associated with rebleeding (p<0.01) and ICU admission (p<0.04).

Abnormal r-TEG phenotypes are associated with clinical outcomes that could guide treatment decisions and prognosis in the care of adult ED patients with GIB. However, how r-TEG results are classified can affect these relationships, and in turn clinical decision-making. Further research is needed to establish a phenotype classification system to guide the management of these patients.