Role of Thromboelastography in Septic Shock

Introduction

Coagulation dysfunction is frequent in patients with severe sepsis and it is associated with an increase risk of mortality. During sepsis platelets number usually decreases and their function is reduced and this mechanism is sustained by an inflammatory induced coagulopathy. Several studies have showed a correlation between altered function of platelets and sepsis severity.

Conventional coagulation tests such as activated partial thromboplastin time (APTT), prothrombin time (PT) and international normalized ratio (INR) were developed mainly to monitor anticoagulation treatment and may not fully reflect the complex in vivo coagulation disturbances that usually characterizes sepsis. More recently some studies evaluated the possibility to use viscoelastic whole blood tests of the haemostasis, such as thromboelastography (TEG) and thromboelastometry (ROTEM), which analyze all blood components and their interactions during clot formation and dissolution and might be useful for assessing bleeding risk in septic patients.

A more recent study identified a possible hypocoagulability pattern among septic patients at risk of bleeding using TEG variable and the authors found a strong correlation with increased risk of death. Maximun amplitude (MA) is one of the variables obtained from TEG analysis and it expresses the strength of the clot and the efficacy of platelet function. A low level of MA describes a lower strength of the clot determined by a lower number or a reduced function of platelet. In one recent study it has been described that a lower MA describing a hypocoagulability pattern might be associated with an increased risk of death.

The aim of the present study is to evaluate whether a lower level of MA and a pattern of hypocoagulability might be associated with an increased risk of bleeding and need of transfusion in patients with sepsis.

Methods A prospective multicenter observational study will be conducted in five intensive care units of five Italian university hospitals. We asked for Ethic Committee approval.

Inclusion criteria: : all adult patients with diagnosis of sepsis admitted for more than 48 hours will be enrolled.

Exclusion criteria: patients younger than 18 years old of age; chronic use of oral anticoagulant and anti platelet treatment; hematologic malignancy; congenital bleeding disorders; oral contraceptives; lack of consent.

Primary end point To evaluate whether a lower level of MA might be associated with an increased risk of bleeding.

Secondary end points

1. To evaluate whether a different level of MA correlates with the biomarker of the severity of sepsis such as presepsin (PSEP).

2. To evaluate whether a different level of MA correlates with the biomarker of the severity of infection such as procalcitonin (PCT).

3. To evaluate whether a pattern of hypocoagulability might be associated with a risk of mortality.

Study Design Prospective, multicenter, observational

All demographic and clinical data will be collected for each patient, as following:

Age, gender, weight, simplify acute physiology score (SAPS) II, sequential organ failure assessment (SOFA) score (at admission and every 72 hours), clinical pulmonary infection score (CPIS), if the source of infection comes from lung, type of infection and microbiological isolation, haemodynamic parameters (heart rate, mean arterial pressure, pulmonary capillary wedge pressure, cardiac output, mixed venous oxygen saturation) and renal function (creatinine, urine output and fluid balance).

All enrolled patients will undergo a blood sample at admission (T0), after 72 hours (T1) and after 7 days (T2) and all the following parameters will be measured:

Platelet count, APTT, PT, INR, fibrinogen, procalcitonin and presepsin.

Additionally, all viscoelastic parameters will be performed at bedside, using a TEG 5000 Hemostasis Analyzer System (Haemoscope Corporation, Chicago, Illinois, USA), at T0, T1, T2:

1. reaction time (R), normal range (3-8 sec)

2. clot formation speed (K), normal range (1-3 sec)

3. angle (alpha), which reflects the clot growth kinetics, normal range (55-75°)

4. maximun amplitude (MA), normal range (51-69 mm)

5. fibrinogen maximun amplitude (FFMA), normal range (14-24 mm)

Bleeding definition We defined bleeding event as any intracranial bleeding identified at CT scan or any other bleeding at gastrointestinal tract (hematemesis or melena or frank blood in the stools), at tracheal aspirates or in the urine or bleeding during surgery and from wounds or any bleeding event requiring concomitant transfusion of three units of red blood cells.

Outcome measurements

1. Intensive Care Unit Length of Stay (ICU-LOS);

2. Mortality at 28 days and at 90 days.

Statistical analysis A power analysis was calculated considering primary endpoint: correlation between level of MA and risk of bleeding.

We defined bleeding as intracranial, gastrointestinal, tracheal or surgical bleeding or any bleeding event requiring concomitant transfusion of three units of red blood cells. From previous study we observed an incidence of 26%.

Power 86.4% B (size of the regression coefficient) 0.1 Standard deviation of MA 6 P (event rate) 26% Alpha 0.05

The expected number of patients enrolled in 24 months is about 100 and with a Cox proportional hazard model we can estimate the Hazard Ratios (HR) and 95% CI with the parameters of the following table.

A Cox regression of the log hazard ratio on MA as a covariate with a standard deviation of 6, based on a sample of 100 observations achieves 86.4% power at a 0.05 significance level to detect a regression coefficient equal to 0.1 (log of hazard ratio). The sample size was adjusted for a bleeding rate of 26%.

The characteristics of the patients and clinical measures will be describe using the most appropriate statistics, as mean (SD), median (IQR) or percentages according to variable type and distribution.

The cumulative incidence of bleeding during period of observation (7 days) will be estimated with the Gooley method, to take into account the competitive risk of death, and compared with the Gray test. The Fine and Grey model will be used to estimate adjusted HR and their 95% CI.

Evaluation of associations between MA and severity of disease or infection will be assessed using Spearman's correlation test.