Thromboelastography and mean platelet volume as prognostic markers in sepsis:
an observational study
DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE RULES AND REGULATIONS FOR THE MD GENERAL MEDICINE EXAMINATION OF THE TAMILNADU DR. M.G.R.
MEDICAL UNIVERSITYTO BE HELD IN MAY 2019
DECLARATION
This is to declare that the dissertation entitled “Thromboelastography and mean platelet volume as prognostic markers in sepsis: an observational study” is my original work done in partial fulfilment of rules and regulations for the MD General Medicine examination of the Tamil Nadu Dr.M.G.R Medical University, Chennai to be held in May, 2019.
Dr Fibi Ninan K Post Graduate student
Register Number: 201611457 Department of General Medicine Christian Medical College
Vellore October 2018
BONAFIDE CERTIFICATE
This is to certify that the dissertation entitled “Thromboelastography and mean platelet volume as prognostic markers in sepsis: an observational study” is a bona fide original work done by Dr Fibi Ninan K during her academic term April 2016 to March 2019, at Christian Medical College, Vellore in partial fulfilment of rules and regulations for the MD General Medicine examination of the Tamil Nadu Dr.M.G.R Medical University, Chennai to be held in May, 2019. This work was carried out under my guidance in the department.
Dr Ramya I Professor
Department of General Medicine Christian Medical College
Vellore
BONAFIDE CERTIFICATE
This is to certify that the dissertation entitled “Thromboelastography and mean platelet volume as prognostic markers in sepsis: an observational study” is a bona fide original work done by Dr Fibi Ninan K during her academic term April 2016 to March 2019, at Christian Medical College, Vellore in partial fulfilment of rules and regulations for the MD General Medicine examination of the Tamil Nadu Dr.M.G.R Medical University, Chennai to be held in May, 2019.
Principal
Dr Anna B Pulimood The Principal
Christian Medical College Vellore
Head of the department
Dr Thambu David Sudarsanam Head of the Department
Department of General Medicine Christian Medical College Vellore
ACKOWLEDGEMENT
“Thus far has the Lord helped us....”
I express my sincere gratitude to all who contributed to this thesis:
Dr. Ramya I for her expert guidance from the conceptualization of the theme, till the final full stop; there is not one area of this thesis where her hands or mind have not touched.
Dr. John Victor Peter for providing patients from the ICU and for his help in valuable critique and suggestions,
Dr.SukeshChandran for guiding me through the analysis of results of Thromboelastography,
Dr.Thulasi Geevar for her support in counter checking thromboelastography results, Ms Preethi Lilly, for promptly providing thromboelastography and mean platelet volume results,
Mr BijeshYadav for helping me with the statistics, All the consultants for their constant encouragement,
My fellow registrars for being very helpful and considerate all throughout,
My husband Dr. Justin and our parents for their constant support, encouragement and prayers all throughout our lives,
My sister Dr Fini for her constant support in writing up the thesis,
My relatives and my fellowship members of Christian Revival Fellowship for constant prayers and well wishes,
Last but not least, I thank all my patients for their co-operation and trust. Without them my training would be incomplete and this thesis would not have happened.
ANTIPLAGIARISM CERTIFICATE
This is to certify that this dissertation work titled - “Thromboelastography and mean platelet volume as prognostic markers in sepsis” of the candidate Dr. Fibi Ninan K with registration Number 201611457 has submitted her dissertation for verification and I have personally verified the Urkund.com website for the purpose of plagiarism check. I found that the uploaded thesis file contains the introduction to conclusion pages and the analysis shows eight percentage of plagiarism in the dissertation.
Dr Ramya I
Professor
Department of General Medicine Christian Medical College
Vellore
TABLE OF CONTENTS
INTRODUCTION...…...14
AIM AND OBJECTIVES …….………….….………...17
LITERATURE REVIEW……….………….………..………...19
MATERIALS AND METHODS………….………...49
RESULTS……….………...60
DISCUSSION……….………...75
CONCLUSIONS ………...…...………...85
LIMITATIONS OF THE STUDY..…….………...88
REFERENCES………..………... ... 90
ANNEXURE I: IRB APPROVAL FORM……….…...104
ANNEXURE II: INFORMATION SHEETS AND CONSENT FORMS...108
ANNEXURE III: CLINICAL RESEARCH FORM …………..…...…...125
ANNEXURE IV: EXCEL SHEET...127
ANNEXURE V: STROBE CHECK LIST...132
Abbreviations
ALI: Acute Lung injury
APTT: Activated partial thromboplastin time CFT: Clot formation time
DAMA: Discharge against medical advice DAMPS: Damage associated molecular patterns DIC: Disseminated intravascular coagulation
ISTH: International Socitey on Thrombosis and Haemolysis JMHLW: Japanese Ministry Health, Labour and Welfare JAAM : Japanese Association of Acute Medicine
FFP: Fresh frozen plasma MA: Maximum amplitude MCF: Maximum clot formation MOF: Multi organ failure MPV: Mean platelet volume PT: Prothrombin time
PAMPS: Pathogen associated molecular pattern ROTEM: Rotational thromboelastometry
SAPS II: Simplified acute physiology score II SOFA: Sequential organ failure assessment TEG: Thromboelastography
TAFI: Thrombin activable fibrinolysis inhibitor
LIST OF TABLES AND FIGURES
List of figures
Fig 1: Demostrates the various signalling pathways involved in sepsis …………...20
Fig2: Demonstrates endothelial dysfunction in sepsis: ………. ..22
Fig 3: Demonstrates the conventional method of coagulation………...24
Fig 4: Demonstrates the initiation phase ………. ....25
Fig 5: Illustrates pathophysiology of DIC ………..28
Fig 6: Demonstrates the principle of TEG……….. ...,...32
Fig 7: Demonstrates normal TEG tracing ……….33
Fig 8: Demonstrates TEG tracing in various clinical conditions………...35 Fig 9: Demonstrates the role of platelets in activation of immune system in sepsis…. 43
Fig 10: Shows a resting platelet and an activated platelet………...44
Fig 11: Shows picture of TEG machine………....52
Fig 12: Demonstrates tracing of TEG on the system ………...53
Fig 13: Demonstrates analysis of TEG being done………...53
Fig 14: Detailed Diagrammatic Algorithm of the Study: ……….56
Fig 15: Pie diagram demonstrating gender distribution of patients ………. 59
Fig 16: Demonstrates the aetiology of sepsis………. 60
Fig 17: Demonstrates the primary outcome ………62
Fig 18: Summarises coagulation abnormalities in patients with severe sepsis based on various parameters in TEG ………...65
List of tables Table1: Demonstrates the ISTH score for DIC ……….. 29
Table: 2 Illustrates definition of hyper and hypocoagulability based on TEG parameters ………..34
Table 3: Illustrates various TEG parameters in different clinical settings...…...…36
Table: 4 Illustrates transfusion of blood products based on TEG...…………..38
Table 5: Various modifications in TEG ………..….42
Table 6: Parameters of platelets were studied in the past as markers of disease severity……… 44
Table 7: Normal values of TEG parameters……… 51
Table 8: Calculation of sample size...54
Table 9: Age distribution of patients with severe sepsis ……….58
Table 10: Demonstrates the co-morbidities of the patients ……….59
Table 11: Shows the spectrum of patients with bacteraemia ………..61
Table 12: Demonstrates various coagulation parameters ………...61
Tables 13: Compares the mortality in bacteraemic and non bacteraemic patients ………...62
Tables 14: Comparison of mortality in patients with hypercoagulable, normocoagulable and hypocoagulable state based on R time ………... 63
Tables 15: Comparison of mortality in patients with hypercoagulable, normocoagulable and hypocoagulable state based on K time ……… ...63
Tables 16: Comparison of mortality in patients with hypercoagulable, normocoagulable and hypocoagulable state based on alpha angle………...…. 64
Tables 17: Comparison of mortality in patients with hypercoagulable, normocoagulable and hypocoagulable state based on maximum amplitude………...64
Table:19: Demonstrates correlation between TEG parameters and mortality .………...66
Table 20: Demonstrates association of TEG parameters and requirement of transfusion ………... 67
Table 21: Demonstrates association of TEG parameters and requirement of transfusion of
fresh frozen plasma ………...…… 67
Table 22: Demonstrates association of TEG parameters and requirement of transfusion of platelets ………...68
Table 23: Demonstrates association of TEG parameters and requirement of transfusion of cryoprecipitate ………...68
Table 24: Shows correlation between PT and TEG parameters………...69
Table 25: Shows correlation between APTT and TEG parameters ……….. 69
Table 26: Shows correlation between Platelets and TEG parameters ………... 69
Table 27: Demonstrates correlation between various TEG parameters and ICU stay and hospital stay ………...70
Table 28: Demonstrates correlation between various TEG parameters and SOFA score ………... 71
Introduction
Sepsis remains to be a life threatening condition despite having insight into its aetiology, pathophysiology and advances in early goal directed therapy(1). As per 2016 Society of Critical Care Medicine and European Society of Intensive Care Medicine task force, sepsis is defined as ‘life threatening organ dysfunction caused by dysregulated host response to infection’(2). The annual incidence of sepsis is 20 million cases worldwide of which 5 million patients do not survive in the low and middle income countries(3).
The knowledge of pathophysiology of sepsis is incomplete till date. The key event in sepsis is endothelial dysfunction which causes microvascular dysfunction. This is associated with hypoxia, oxidative stress and release of cascade of inflammatory mediators and in turn multiorgan dysfunction syndrome(1,4). As part of this inflammatory response various coagulation abnormalities occur that ranges from minor drop in platelet counts to severe forms of coagulopathy like disseminated intravascular coagulation(5).
Coagulation system is activated by various proinflammatory mediators that are released as a host response to infection. The initial hypercoagulable state causes deposition of fibrin in the micro vasculature leading to multi organ dysfunction syndrome(5). Over a period of days platelets and various coagulation factors are consumed due to progressive generation of thrombin resulting in overt DIC(6). ISTH defined pre DIC as compensated coagulopathy, a stage before overt DIC sets in(7). The diagnosis of sepsis induced coagulopathy in its pre DIC stages may aid in early goal directed therapy.
Thromboelastograph is a single test which assesses various stages of coagulation cascade ranging from fibrin formation to platelet aggregation and fibrinolysis (8). Mean platelet volume (MPV) is a measure of average size of the platelets (9).We decided to determine
whether abnormal thromboelastography and mean platelet volume is associated with adverse outcome in patients with severe sepsis.
Kansuke et al in their study demonstrated that a single measurement of plasminogen activator inhibitor-1, thrombin-antithrombin complex, and protein C activity helps in the diagnosis of coagulopathy early in sepsis(6).The hypothesis is that severe sepsis is characterized by a pro-coagulant state in the initial stages, and hypocoagulation and disseminated intravascular coagulation (DIC) in the later stages. Thus hypercoagulation detected by thromboelastography (TEG) in early sepsis in patients without clinical bleeding could predict subsequent DIC and multi-organ dysfunction. This observation was noted in a preliminary study on a small sample population by Turani et al(10).Our study was determined to determine association of abnormal TEG (either hypercoagulable or hypocoagulable state) and mean platelet volume with adverse outcomes in patients with severe sepsis.
Aim and Objectives
Aim
To determine whether abnormal thromboelastography and mean platelet volume is associated with adverse outcome in patients with severe sepsis.
Objectives
1. To study the prevalence of abnormal thromboelastography and mean platelet volume in patients with severe sepsis.
2. To evaluate if thromboelastography and mean platelet volume are associated with number and severity of organ dysfunction as assessed by the sequential organ failure assessment (SOFA) score.
3. To study the association between an abnormal thromboelastography and high mean platelet volume, at the time of diagnosis of sepsis, on mortality and duration of ICU stay and hospital stay in patients admitted with severe sepsis
Literature Review
Incidence and definition of sepsis
Sepsis is the leading cause of death in non coronary critical care unit(11) and has been termed as the ‘hidden public disaster’(12). There is a steady trend towards increase in the incidence of sepsis which reflects the aging population with multiple comorbid conditions(13). The global burden of severe sepsis ranges from 13-300/100,000 and the case fatality rate is as high as 50% (14). The incidence of sepsis in India as reported by Todi et al is 16.5 % of all admissions and has a hospital mortality of 65.2% (11).
Sepsis has been referred to as ‘garbage code’ being a pathway leading to death from an infection. However most of these patients die from sepsis than from the infection as such(15). As per the third international consensus, sepsis is defined as ‘life-threatening organ dysfunction caused by a dysregulated host response to infection’ and organ dysfunction is defined as ‘acute change in total SOFA score ≥2 points consequent to the infection(13). Septic shock is defined as the condition in which a patient requires vasopressors to maintain a mean arterial pressure > 65 mm of Hg and serum lactates >
2mmol /L in the absence of hypovolemia. Patients with organ dysfunction is known to have an in hospital mortality >10%. Septic shock is associated with in hospital mortality
>40%(13).
Pathophysiology of sepsis
The normal physiological response of the body helps in eradication of pathogens.
Sepsis is characterized by inappropriate regulation of these mechanism(16). Hence the
degree of damage depends both the pathogenicity of the invading organisms and the exuberant host response which causes exaggerated tissue damage(12).
The pathophysiology of sepsis can be explained under the following headings a)Inflammation
Bacteria, virus and fungi release various substances like endotoxins and betaglucans which are called pathogen associated molecular patterns (PAMPS). Similarly there are various damage associated molecular patterns (DAMPS) which are endogenous molecules released from damaged host cells like ATP, mitochondrial DNA etc. Various pattern recognition receptors are located in the cell receptors or in the cytosol , toll like receptors /NOD like receptors, to name a few of them. Innate immune system is activated by PAMS and DAMPS through pattern recognition receptors and thereby initiating transcription of cytokines like TNF alpha, IL-6, IL-1 etc. These inflammatory mediators help in programmed cell death and when it exceeds a certain threshold it can lead to activation of complement cascade , widespread thrombosis and release of reactive oxygen species(17).
In sepsis, various signalling pathways are activated leading to expression of several genes involved in inflammation. Various pro inflammatory cytokines are released in recognition of various bacteria, fungi and viruses which in turn lead to phosphorylation of JAK, MAK, NFkb etc. These mediators activate various early activation gene.(18).
Fig: 1: Demostrates the various signalling pathways involved in sepsis
(12)
b)Early activation gene
Early activation genes are interleukins, IL-1, IL12 and interferon alpha that are produced in response to nuclear translocation of NF kb. These inflammatory mediators release a cascade of inflammatory mediators like IL-6,IL-8 and suppression of components of adaptive immunity (18).
c)Microvascular dysfunction
Another mechanism of sepsis is the alteration of the microcirculation which consists of an area of 1000m2 (17). Microcirculation which is embedded in various organs consists of pre capillary arterioles, arterioles and post capillary venules. These vessels help in regulation of blood flow, oxygenation and tissue perfusion and blood pressure (4). In fact in the year 2010 sepsis was redefined as ‘severe endothelial dysfunction syndrome leading to reversible or irreversible injury to micro circulation and thereby multiple organ dysfunction syndrome’ Hence any injury to the microcirculation can cause serious injury in various organs leading to multi organ dysfunction syndrome (19).
d)Endothelial barrier dysfunction
Endothelial barrier dysfunction is fundamental event in sepsis. In normal state endothelium acts as anticoagulant surface in the blood vessels and it maintains normal flow of various molecules. Its integrity is maintained by actin, intercellular adhesion molecules which form tight junctions and various support proteins(18).
Microvascular leak in the endothelium is produced by various inflammatory mediators like cytokines, chemokines and direct action of microbial virulence factors like LPS and staphylococcal toxin. Similarly host endogenous products released from the vascular cells cause a deadly blow to microvascular endothelial cells (4).
In addition, endothelial cells are covered by a glycoprotein polysaccharide layer called glycocalyx. It maintains the tight junctions and supports the anticoagulant state. In sepsis, various inflammatory mediators target the glycococalyx and disrupt the barrier. Due to leaky capillaries large amount of protein is lost into the extra vascular compartment. There is diffuse vasodilatation leading to alteration in microvascular blood flow and there by poor tissue perfusion and shock (18).
Fig 2: Demonstrates endothelial dysfunction in sepsis
(20)
Prognostic markers in sepsis
As sepsis involves high incidence of mortality, early recognition of patients who are at risk of rapid progression of sepsis is imperative. Early goal directed therapy is imperative for better outcomes of patients (21). Much attention has been given to the inflammatory markers in sepsis that precedes organ dysfunction and mortality in sepsis. The inflammatory response may provide protective effect against pathogenic organism, but when there is an exaggerated response it becomes counterproductive and harmful, leading to multi-organ dysfunction(22).
Different parameters have been studied in the past as prognostic markers in sepsis. In a study conducted in Iran it was found that age, sex, low albumin and diastolic blood pressure <52 mm of Hg were indicators of rapid progression of sepsis (23). Another prospective study done over a period of 6 years showed older age, rise in serum lactate and elevated procalcitonin was associated with increase in mortality(21).
As sepsis is known to be associated with coagulation abnormalities, studies were done to assess causality between various determinants of coagulation and sepsis.
Sepsis and coagulation system
Coagulation abnormalities are almost universal in sepsis. Inflammation and coagulation abnormalities are tightly linked together that, each sends a positive feedback to the other.
The coagulopathy in acute sepsis can range from florid thrombo-embolic disease to microvascular fibrin deposition (22).
The mechanism of coagulation can be explained by the conventional coagulalation pathways and new coagulation pathways.
Conventional coagulation pathway
Fig 3: Demonstrates the conventional method of coagulation
(24)
Mechanism of coagulation involves primary haemostasis followed by activation of the intrinsic and extrinsic coagulation pathways. Primary haemostasis is involved in platelet plug formation secondary to complex interactions between platelets, vessel wall and complex proteins as illustrated above.(25)
New coagulation pathway
Current evidence supports that extrinsic pathway initiates thrombin formation and the intrinsic pathway augments the process. This is in contrast to the classical teaching where extrinsic and intrinsic pathways were thought to get activated in a parallel manner. Various steps involved in the process are initiation, amplification, propagation and stabilisation.(26)
Initiation phase
The tissue factor in damaged vessel bind to factor VIIa which in turn activates factor IX and factor X simultaneously.
Fig 4: Demostrates the initiation phase
Amplification phase
Trace amount of thrombin that is generated in the initiation phase is inadequate for clot formation. Various positive feedback loops exist that bind thrombin and platelets.
Thrombin activates factor VIII and V which in turn activates factor II.
Propagation phase
Enzyme complexes like tenase complex and prothrombinase complex on platelet surface causes platelet activation and thrombin formation. This ensures generation of sufficient amount of thrombin , fibrin and later blood clot.
Stabilisation phase
Thrombin stabilises the clot through 2 mechanisms
1. It activates the clot stabilizing factor(factor XIII) . It covalently links fibrin polymers and provides stability to the fibrin.
2. It stimulates the thrombin activable fibrinolysis inhibitor (TAFI) which inhibits fibrinolysis.(25)
The above mentioned pathways are deranged in sepsis. The pathogenesis of coagulopathy in sepsis is due to activation of the procoagulant pathway, downregulation of anticoagulantion mechanisms and impraired fibrinolysis as mentioned below.
A)Activation of the coagulation
In sepsis there is direct activation of the coagulation pathway via both extrinsic and intrinsic pathway. Various toxins cause upregulation of tissue factor. The tissue factor directly activates factor VII of the extrinsic pathway leading to thrombin formation and formation of fibrin clots. Factor VIIa activates factor IX of the intrinsic pathway and hence formation of thrombin (27).
Coagulation system is also activated by activation of the complement pathway.
Complement mediated shedding of the microvescicles releases various tissue factors that are prothrombotic (18).
A) Downregulation of anticoagulation mechanisms
Various molecules that promote anticoagulation are embedded in the endothelial surface.
Various inflammatory mediators released in sepsis leads to endothelial injury (28). This leads to breech in the endothelium. The damaged glycocalyx layer exposes the collagen underneath which activates the von willibrand factor and in turn the platelets. Platelets activate coagulation cascade by activation of thrombin. Thrombin converts insoluble firbrinogen in to fibrin. Platelets and strands of fibrin form the backbone of clot which gets deposited in microcirculation leading to reduction in perfusion and multi organ dysfunction syndrome.(18)
C)Reduction in fibrinolysis
Thrombin releases thrombin–activable fibrinolysis inhibitor (TAFI) which helps in reduction of fibrinolysis. Thus thrombin not only causes clot formation but also inhibits its removal.(27)
The activation of coagulation system, down regulation of anticoagulant mechanisms and inhibition of fibrinolysis ultimately leads to disseminated intravascular coagulation (29).
Fig 5: Illustrates pathophysiology of DIC
(30)
Conventional tests for coagulation in sepsis
In sepsis there is widespread activation of the coagulation cascade and inhibition of the fibrinolytic pathway leading to widespread thrombosis and MODS. Haemostatic factors like fibrinogen, factors V and VIII, platelets, protein C and anti thrombin III are
Consumption of platelets
consumed as part of widespread thrombin formation in sepsis. Compensatory thrombolysis is caused by tissue plasmin activator which is released by endothelial cells of occluded vessels. As a result conventional tests of coagulation may show thrombocytopenia and prolongation of prothrombin time and activated partial thromboplastin time. The byproducts of thrombolysis such as fibrin degradation products or D-dimer are usually elevated in severe sepsis(31). The diagnosis of DIC by conventional methods is based on 3 different criteria, International Socitey on Thrombosis and Haemolysis (ISTH), Japanese Ministry Health, Labour and Welfare (JMHLW) and Japanese Association of Acute Medicine (JAAM ).
Table1: Demonstrates the ISTH score for DIC
Parameters Score
Platelets x 109/L >100 0
≤100 1
<50 2
PT (s) <3 0
≥3 and <6 1
≥6 2
Fibrin related marker, mg/L D dimer <1 0 1.0 ≤D dimer<5 1
D dimer ≥5 2
Fibrinogen , g/L >1.0 0
≤1.0 1
(7)
A score ≥5 is suggestive of overt DIC. However a score ≤ 5 is indicative of DIC but not confirmatory.
Traditional tests of coagulation depend on extrinsic and intrinsic pathway and the number of platelets. However clot formation involves interaction between myriad other factors like vascular endothelium, von willibrand factor, pro-coagulant and anti coagulant factors and blood flow. For effective haemostasis there should be adequate thrombin generation (platelets and coagulation factors), enough fibrin and clot stability. Only <5% of the thrombin that would have been generated during anticoagulation is formed at the time of testing in conventional tests and it does not represent the overall haemostasis (32).Conventional tests for coagulation tests various parts of the coagulation cascade in isolation and it may not reflect the in vivo coagulopathy (10).
The various fallacies in conventional tests of coagulation are
1. PT, APTT reflects overall decrease in coagulation factors but it does not reflect the nature of the coagulation defect.
2. State of hypercoagulability cannot be assessed
3. Spurious values in conditions like anti phospholipid antibody syndrome(33).
4. Conventional tests do not account endogenous anticoagulant factors and cellular elements.
5. Moreover it does not account for different thrombin dependent reactions like activation of the platelets(32).
Muzaffar et al in his study compared conventional tests with TEG. It was found that among patients with INR >1.6, 60% had a normocoagulable state where as 20% had hypercoagulable state. Similarly normocoagulable state was found in 81% of the patients with platelets <1,00,000 (34). This reiterates the need for alternative tests to assess the
coagulation. HheseMoreover these abnormalities are manifested in conventional tests only after occurrence of DIC by then it is too lat
Thromboelastography a novel diagnostic tool
TEG was first conceived by Helmut Hartert in Germany in the year 1948 at the Heidelberg University of Medicine (35). TEG is a real time visco-elastic assay which sums the effect of various factors involved in coagulation and provides a global assessment of haemostatic function (31,36).
TEG is performed by placing the 340µl whole blood sample in a cup which oscillates 4°45′
every 5 seconds at 37o C. A stationary pin is suspended from torsion wire in to the cup. The oscillations in the cup mimic the sluggish flow in the venous circulation. The coagulation process is activated leading to the formation of fibrin strands. These strands couple the motion of cup and pin together. The electromagnetic transducer detects increased tension in the wire and the electrical signal is amplified to create a trace in the computer screen (37).
Free rotation corresponds to an amplitude of 0 mm and no rotation corresponds to an amplitude of 100 mm (38).
ROTEM (Rotational thromboelastometry) is a modification of TEG in which the cup containing 340µL of blood remains static while pin oscillates on a ball bearing mechanism.
As the fibrin forms on the pin, the oscillation of the pin is impeded .This is detected on an optical based system (38).
Fig 6: Demonstrates principle of TEG
(39)
Fig 7: Demonstrates normal TEG tracing
(40)
There are 4 important variables in TEG:
R time: It is the time from the initiation of the study to initiation of clot formation. It is analogous to PT/APTT.
K time: Denotes R time till the time till enough fibrin cross linkages occur to produce amplitude of 20mm. Clotting at this time is dependent on fibrin cleavage and fibrin polymerisation.
Alpha angle: It is the angle between the baseline and a tangent to the curve of the TEG. It expresses the kinetics of clot formation. Higher angle represents greater rate of clot strength formation and vice versa. Alpha angle depends upon fibrinogen concentration and function.
Maximum amplitude: It is the width in millimetres of the widest gap in the tracing and it indicates maximum strength of clot formation. MA depends upon platelet number and function as well as on the fibrinogen. Around 20% of the clot strength depends on fibrinogen and 80% depends on platelets.
Lysis 30: It is the standard measure of fibrinolysis. It denotes percentage reduction in clot strength 30 min after achieving MA(35).
TEG gives holistic picture of clot tensile strength over time: clot formation, clot strength and clot stability(32).
Table: 2 Illustrates definition of hyper and hypocoagulability based on TEG parameters
R time K time Alpha angle Maximum
amplitude Hypercoagulable
state
Hypocoagulable state
(37) .
Fig 8: Demonstrates TEG tracing in various clinical conditions
(41)
Table 3: Demonstrates various TEG parameters in different clinical situitations Haemostatic abnormality R time Alpha angle MA
Clotting factor deficiency High Low / normal Low/normal
Fibrinogen deficiency Normal Low Low/normal
Low platelet counts Normal Normal Low
Primary hyperfibrinolysis Normal Normal Low Hypercoagulability with late
fibrinolysis: Early DIC
Low High High
DIC late High Low Low
(42).
Thromboelastogram and its uses Diagnostic purposes
TEG has been used to prognosticate cerebrovascular events. In 246 patients with acute ischemic stroke it was noted that hypercoagulability as demonstrated with reduced R time was associated with early neurological deterioration (43). In a study conducted in China among post-operative patients it was noted that R value of TEG was significantly different in patients who had venous thromboembolism and patients who had bleeding. Hence thromboelastogram could be used as a guide to predict post-operative complications (44).
TEG was used to analyse coagulopathy in patients with dengue. It was noted that factor deficiency followed by platelet dysfunction was the most common coagulopathy(45). In a study in paediatric patients with snake bite and it was noted that an abnormal TEG was a
better predictor for clinical bleeding than INR with a sensitivity of 94.4% and specificity of 45.5% (46). However TEG did not predict mortality, risk of bleeding or thrombosis or liver transplantation in patients with liver cirrhosis (47).
Another area visco elastic tests may of use is in the monitoring of newer oral anticoagulant agents. In a small study done in patients on dabigatran showed strong correlation ROTEM values (activated by reagents of EXTEM and FIBTEM) with dabigatran levels. However these tests did not show any correlation with vitamin K antagonists or with other direct oral anticoagulant drugs like rivaroxaban (38).
Therapeutic purposes
a)TEG and its use to tailor transfusions
Thromboelastogram has been studied to as a tool to determine transfusion requirements.
1. Prolonged R time indicates clotting factor deficiency.
2. Prolonged K time and low alpha angle indicates fibrinogen deficiency and /or fibrin production or function.
3. Low MA indicates lack of clot formation by fibrin platelet mesh demonstrating the need for platelet transfusions (48).
Table: 4: Illustrates use of TEG in transfusion of blood products
Potential therapeutic intervention Finding on TEG
Plasma or prothrombin concentrate Prolonged R value (>7 minutes) Cryoprecipitate / fibrinogen concentrate Low or flat angle(<45)
Platelets+/cryoprecipitate/fibrinogen concentrate
Narrow MA (<48mm)
Anti fibrinolytic agent Increased LY 30 (>7.5%)
(48)
TEG has been used in various clinical settings as a test for coagulation and as a guide to transfusion of blood products. In a systematic review that included 17 randomised control trials, TEG guided therapy was found to reduce need for transfusion of all types of blood products. However it was noted that the overall quality of these studies were low(49).
TEG has been studied in various populations.
-In patients undergoing liver transplantation, assessing TEG to monitor and treat hyperfibrinolysis in patients helped to reduce blood transfusions (50).
- In cardiac surgery transfusion tailored according to TEG results resulted in significantly lower incidence of blood transfusion in comparison to transfusions based on standard tests for coagulation (78.5% versus 86.5%) (51). However a systematic review in patients post cardiac surgery did not show reduction in all cause mortality or reduction in transfusion of platelets with TEG guided transfusions(52).
-It has been studied in trauma induced coagulopathy. It was noted that in these group of patients 28 day mortality was better in patients who received TEG guided therapy (53).
-However various studies that looked into the role of TEG in predicting bleeding in patients with aspirin and LMWH, found that TEG cannot be used as a guide to titration of dose of LMWH (54).
Thromboelastography as a prognostic factor in sepsis
Thromboelastography has been studied as a predictor of mortality. However the quality of evidence supporting TEG as prognostic marker in sepsis is low to moderate.
(22). There has been wide variation in TEG parameters in patients with sepsis (34). The reasons for this wide variation would be
1. The time when TEG was done. Initial phase of sepsis is characterised by formation of micro thombi in small vessels and later hypocoagulability due to consumptive coagulopathy.
2. There is lack of universal reference ranges for TEG.
3. Differences in disease severity in different study groups (42).
One of the important uses of TEG was in prediction of neonatal sepsis. In a prospective trial that included 103 children TEG was found to have 96 % sensitivity and specificity and positive predictive value of 88% and negative predictive value of 98% (55).
Adamzik et al compared TEG with SOFA score and SAPS II. There was 4.1-fold increase in 30-day mortality in patients, when there was at least one pathological
thromboelastometry finding. SAPS II and SOFA score were not different among survivors and non survivors where as mean CFT, MCF (parameters in ROTEM) and alpha angle was significantly different between the 2 groups. Hence TEG could be used as a prognostic factor in sepsis (56).
As mentioned earlier TEG can differentiate between hypocoagulable state and hypercoagulable state. Hence it is important to know what state of coagulopathy is associated with adverse prognosis. In a study by Brenner et al, TEG demonstrated a hypocoagulable state in patients with DIC and hypercoagulable state in non-overt DIC (9).
Similar results were shown in a study by Sivula et al and it was postulated that hypercoagulable state in non overt DIC was due to increase in fibrinogen level as an acute phase reactant (57).
Early hypocoagulability was found to be an independent risk factor for 28 day mortality and higher risk for MODS. Patients with normal coagulation at admission who progressed to hypocoagulation later had a mortality rate of 80% (22) .
Limitations of thromboelastogram
As with any other tests thromboelastogram is associated with its limitations
1. The method is not standardized and hence there is lot of inter laboratory variability in the results. It involves manual pipetting and mixing of reagents which are operator dependent.
2. Vibration on the surface where the TEG machine is kept may introduce artifacts.
3. Citrated samples needs to be rested for fixed amount of time before analysis and hence it may be time consuming.
4. The results of the TEG may vary depend on multiple factors like haematocrit, platelet count, infusion of colloids or crystalloids etc.
5. There are no reference values for different populations like males, females, parturients.
6. The test involves cost and expertise which precludes multiple analysis for a single patient (58).
Modifications of TEG to overcome the limitations
Various modifications have been made to TEG for the ease of analysis. Addition of various re-agents aid in overcoming the limitations involved in TEG.
Table 5: Various modifications in TE
Reagent
Outcome
Citrate Enables prolonged storage of samples before analysis
Heparin Inhibits thrombin, allowing the contribution of fibrin and platelets to be studied
Heparinase Reverses the effects of heparin
Activators (celite, kaolin, tissue factor)
Increase speed of result acquisition
Glycoprotein IIb/IIIa inhibitors
Inhibits platelet function
allowing the contribution of fibrinogen to be assessed
Antifibrinolytic drugs (tranexamic acid)
Reverses fibrinolysis
Reptillase and factor XIIIIa
Activates fibrin formation without affecting platelets
Arachadonic acid Activates platelets via production of thromboxane A2
ADP Activates platelets via P2Y1 and P2Y12 receptors (59)
As mentioned in the table addition of citrate reduces the need for immediate analysis of the sample which may of use in transport of the sample to referral centres. Similarly addition of heparinise allows use of the test in patients who are on prophylaxis of deep vein thrombosis.
Mean platelet volume and sepsis
Platelets play an important role in the pathophysiology of sepsis. Platelets cause activation of the innate immune system as mentioned in figure below. Hence thrombocytopenia may be associated with dysregulated host response. (60).
Fig 9: Demonstrates the role of platelets in activation of immune system in sepsis.
(60)
Even though low platelet count is common in sepsis, cause of death is rarely due to bleeding. Platelets release various cytokines that interact with leukocytes and endothelium to cause micro thrombi formation. These mechanisms are useful in containing sepsis in localised infection whereas in severe infections it becomes maladaptive and dysregulated leading to multiorgan dysfunction. (61)
Thrombocytopenia in sepsis is caused by decreased production or increased destruction of platelets. The etiology of thrombocytopenia is multi-factorial. However the most common cause of thrombocytopenia is platelet activation followed by platelet consumption, platelet destruction or thrombus formation(62). The bone marrow tries to compensate this by
increase in platelet production which leads to release of immature platelets into peripheral circulation (36).Young platelets have large surface area. Moreover platelets change their shape from discoid to spherical along with formation of pseudopods during platelet activation to achieve a large surface area (63).
Fig 10: Shows a resting platelet and an activated platelet.
(64) Thrombocytopenia is known to be an adverse prognostic factor in sepsis. However it is unclear whether thrombocytopenia causes adverse outcome or whether it is a biomarker of disease severity (60). Various studies have been done that looked at various platelet indices and its association with mortality and multi-organ dysfunction as mentioned in the table below.
Table 6: Platelet parameters as markers of disease severity.
Biomarker Outcome
Thrombocytopenia Mortality
Impaired platelet function MOF
Impaired platelet aggregation MOF and mortality Immature platelet fraction Sepsis progression
TPO MOF
Platelet neutrophil aggregates MOF (62)
Mean platelet volume is a simple and cost effective platelet index that measures the average volume of platelets. The vertical volume of the platelet is analysed by deformation of the electric field based on impedence technology (64). Destructive thrombocytopenia is characterised by high MPV levels where as hypoproliferative thrombocytopenia is characterised by low MPV levels. High mean platelet volume was associated with statistically significant difference between survivors and non survivors in a study from China among patients with septic shock (36).
Zang et al in their study found that patients with high MPV had high APACHE and SOFA scores. High MPV was useful in predicting mortality with a sensitivity of 58.2% and specificity of 90.2% (65). In another study done among patients admitted with nosocomial sepsis it was noted that patients high MPV was associated with gram positive bacteraemia for the first 3 days and gram negative sepsis for the first 5 days . Fungal septicaemia had a strong association with MPV(66).
TEG and MPV are automated tests. If abnormal TEG and MPV predict poor outcomes in sepsis, incorporation of these in evaluation of sepsis in secondary care centres may facilitate early detection of DIC and MOF and referral to higher centres. Similarly if TEG can predict requirement of blood transfusions, blood products may be kept arranged before clinical bleeding occurs in centres with no facilities for blood products. Hence it is worthwhile to study correlation between these tests and mortality in sepsis.
Materials and
Methods
Study design
This was a hospital-based prospective study.
Patient Selection
Patients with a diagnosis of severe sepsis who were admitted under the Department of General Medicine in the medical ICU/HDU of Christian Medical College, Vellore, from April 2017 to June 2018, were screened for eligibility for enrolment in the study.
Inclusion Criteria
All patients with severe sepsis admitted in to the medical/HDU with no clinical bleeding were screened.
Exclusion criteria - Age under 18 years
-Usage of oral anticoagulants, -Treatment with platelet inhibitors
- Pre existent haematological or hepatic disorders -Patients with malignancy
-Patients who received transfusions
-Patients who received haemodialysis prior to sampling -Patients not recruited for >24 hours
-Patients with any clinical bleeding
-Envenomation
-Patients who received heparin for prophylaxis of deep vein thrombosis Definition of terms
Severe sepsis was defined as suspected or confirmed infection and fulfilling ≥ 1 of the following criteria for sepsis with atleast 1 organ dysfunction.The criteria for sepsis was (67)
General variables
-Fever: Core temperature >38o C
-Hypothermia: Core temperature<36 OC -Elevated heart rate >90 bpm
-Tachypnoea
-Altered mental status
-Substantial edema/positive fluid balance (>20ml/kg body weight over 24 hours) -Hyperglycemia (plasma glucose> 120 mg/dl in the absence of diabetes)
Inflammatory variables
-Leukocytosis (white cell count >12,000 /mm3) -Leukopenia (white cell count <4,000/mm3) -Normal white cell count >10% immature forms
-Elevated C reactive protein (>2 SD above the upper limit of normal) -Elevated procalcitonin (2 SD above the upper limit of normal) Tissue perfusion variables
Elevated lactates>1mmol/litre
Decreased capillary refill or mottling
Organ dysfunction was defined as presence of one or more of the following features.
-Acute oliguria< 0.5mL/kg/hr for more than 2 hrs despite adequate fluid resuscitation -Arterial hypoxemia Pao2/Fio2 <300
-Increase in creatinine> 0.5mg/dL from the baseline -Total bilirubin > 4mg/dL
-Platelet count < 100,000 µL
- Coagulopathy (international normalized ratio > 1.5 or APTT>60 sec -Paralytic ileus
Community acquired pneumonia:
Patient with shortness of breath, cough with or without expectoration, pleuritic chest pain with atleast one of the following symptoms(viz malaise, fever, chills and rigors) with new chest signs not explained by any other illness(68).
Nosocomial pneumonia:
When there were new opacities on the chest radiograph with 2 of the following , new onset or worsening cough or dyspnoea or altered mental status in the elderly, new onset purulent sputum or change in sputum character, worsening gas exchange, fever, leukocytosis or leucopenia and no other cause identified.
Acute central nervous system (CNS) infection:
Patient with two of the following symptoms (viz, headache, fever, dizziness, change in level of consciousness or confusion , localizing neurological signs AND one of the following (organism identified on microscopic examination of cerebrospinal fluid, brain tissue, abscess OR imaging suggestive of acute CNS infection OR diagnostic antibodies for an organism)(69).
Scrub typhus :Positive scrub IgM ELISA and defervesence within 48hrs of initiation of Doxycycline or Azithromycin therapy or with the presence of an eschar and a positive IgM ELISA(70). Other infections will be diagnosed based on standard guidelines.
Among nosocomial infections that develop in ICU, only patients who developed catheter related blood stream infections were included in the study. A diagnosis of catheter related blood stream infection(CRBSI) was made when there was systemic inflammatory
response attributed to an intravascular catheter with differences in growth between catheter and peripheral blood cultures or by quantitative culture of the catheter tip(71).
However as most of these patients had received heparin for deep vein thrombosis, such patients could not be included.
Normal values in TEG
The variables considered in TEG were r time, k time, alpha angle, maximum angle and lysis.
Table 7: Normal values of TEG parameters as on our accredited lab
r time: 3-8 minutes,
k time 1-3 minutes,
alpha angle: 55-78 degrees,
maximum amplitude 51-69 mm,
ly30 0-8%.
The normal value of mean platelet volume of our laboratory is 5-10 fl.
Hypercoagulable state was defined as TEG values with reduced k and r times and increased maximum amplitude and alpha angle .Hypocoagulable state was defined as prolonged r and k times and reduced MA and alpha angle(37) .
It was a prospective observational study which was conducted among patients admitted under the department of General medicine in the medical ICU/HDU from April 2017 to June 2018 after obtaining approval of the Institutional review board.
All patients with a diagnosis of severe sepsis without clinical bleeding were included in the study. An informed consent was taken from the patients prior to enrolment. Data collection was done using a clinical research form. The form included baseline characters and details for calculation of SOFA score. 2.8ml of blood was collected in 2 citrated tubes for
assessment of TEG and 2 ml of blood was drawn in 1 EDTA tube for measurement of mean platelet volume through vacutainer. These were sent immediately to the clinical pathology lab for doing TEG and MPV analysis. These patients were followed up till discharge from the hospital. The duration of ICU stay and duration of hospital stay were noted. The number of blood products transfused was also noted. Data was collected from ICU records and from the electronic workstation.
Fig 11: Shows picture of TEG machine
Fig 12: Demonstrates tracing of TEG on the system
Fig 13: Demonstrates analysis of TEG being done
Sample size calculation
The sample size to study the abnormal TEG in patients with severe sepsis was found to be 85 patients, as per the formula given below:
𝑛 =𝑍𝛼/22 𝑃𝑄 𝑑2 P – prevalence of severe sepsis
Q – 100 – P d – precision
Table 8: Estimation of sample size
Standard Deviation
1.65 1.65 1.65 1.65
Absolute Precision 0.2 0.3 0.35 0.4
Desired confidence level (%) 95 95 95 95
Required sample size 261 116 85 65
Based on KILIC et al (2012), the mean(standard deviation) of thromboelastography in sepsis was 3.37(1.65). With 0.35% precision level and 95% confidence level, the required sample size was 85.
Statistical analysis
Data was screened for outliers and extreme values using Box-Cox plot and histogram (for shape of the distribution). Summary statistics was used for reporting demographic and clinical characteristics.
t-test was use for analysis of continuous data with Normal distribution and Mann- Whitney U test for data with non- Normal distribution. Chi-square test was performed for categorical variables and the outcome variable. Differences were considered significant at p<0.05.
Univariate analysis was performed on the pre hoc likely to be associated with a poor outcome (mortality). This would include age, gender, severity of illness, lag time to presentation and organ dysfunction. In addition TEG and MPA were assessed on univariate analysis. All the statistical analysis was performed using SPSS 25
Fig 14: Detailed Diagrammatic Algorithm of the Study
Medical ICU/HDU from April 2017 to June 2018 : All patients with severe sepsis: 241
A
Excluded: 154 Clinical bleeding:26 Transfusions:8
Use of Heparin/aspirin:91 Others:29
87 patients recruited in to the study
• Informed consent
• Data collection using clinical research form
• Analysis for TEG and MPV
Analysis
TEG with outcome TEG with SOFA
TEG with conventional tests TEG with transfusion of blood products
Results
Demographic characters
A total of 241 patients were screened and 87 patients were included in the study. 91
patients were excluded in view of being administration of oral on parenteral anticoagulants .26 patients were excluded in view of clinical bleeding.
Age distribution
The age of the patients in the study group ranged from16 to 85 years. Majority of patients were aged between 56-65 years.
Table 9: Age distribution of patients with severe sepsis
Age group (years) No: of patients (%)
16-25 9 (10.3%)
26-35 9(10.3%)
36-45 15(17.2%)
46-55 17(19.5%)
56-65 24(27.5%)
66-75 12(13.7%)
76-85 1(1.1%)
Gender distribution
There were 49 males (56.3%) and 38 females (43.7%) in the study.
Fig 15: Pie diagram demonstrates gender distribution of patients
Comorbidities of the patients
Table 10: Demonstrates the comorbidties of the patients
Baseline characters Number
Charlson comorbidity index 1
Diabetes 35(40.2%)
Hypertension 32(36.7%)
SOFA score at admission(median and IQR) 8 (6-11)
Males Females
49(56.3%) 38(43.7%)
Gender distribution
Etiology of sepsis
Fig 16: Demonstrates the etiology of sepsis in our patients
0 5 10 15 20 25
Series1 Series2
Etiology of sepsis
Frequency of different types of infection
Table 11: Shows the spectrum of patients with bacteraemia
Blood culture positivity 25(28.7%)
Gram negative organisms 20(76%)
Multidrug Resistant organisms 10(38.4%)
Extended spectrum beta lactamase inhibitors(ESBL) 7
Carbapenem resistant organisms(CRO) 1
Colistin resistant organism 1
Methicillin resistant staphylococcus aureus(MRSA) 1
Baseline blood results
Table 12: Demonstrates various coagulation parameters
Baseline characters Mean
Hemoglobin 11.03mg/dL
Platelets (median and IQR) 1,43,000
(71,0000-2,41,000)
Prothrombin time 14.18
Activated prothrombin time 43
R time 6.9
K time 1.9 (1.2-3.4)
Alpha angle 58.9
Maximum amplitude 59.45
Mean platelet volume 10.7 fL
Primary outcome
Fig 17: Demonstrates the primary outcome
Mortality in bacteraemic patients
Tables 13: Demonstrates compares the mortality in bacteraemic and non bacteraemic patients
Mortality No mortality P value
Bacteraemia(n=25) 12 13 0.029
No bacteraemia(n=62) 15 47
Outcome
Alive Died DAMA
21(24.1%)
5(5.7%)
61(70.1%)
Coagulation profile of patients based on R time
Tables 14: Comparison of mortality in patients with hypercoagulable, normocoagulable and hypocoagulable state based on R time
R time Total Death Alive A vs B B vs C
A Hypercoagulable state (0-2.99)
3(3.4%) 1(33.3%) 2(66.7%) 0.718
B Normal(3-8) 58(66.7%) 14(24.1%) 44(75.9%) C Hypocoagulable
state(>8)
26(29.9%) 12(46.2%) 14(53.8%) 0.04
Tables 15: Comparison of mortality in patients with hypercoagulable, normocoagulable and hypocoagulable state based on K time
K time(n=87) Total Death Alive A vs B B vs C
A
Hypercoagulable state (0-0.9)
8(9.3%) 3(37.5%) 5(62.5%) 0.437
B
Normal(1-3) 53(61.6%) 13(24.5%) 40(75.5%)
C
Hypocoagulable state(>3)
25(29.1%) 10(40%) 15(65%) 0.074
Tables 16:Comparison of mortality in patients with hypercoagulable, normocoagulable and hypocoagulable state based on alpha angle.
Alpha angle(n=86)
Total Death Alive A vs B B vs C
A Hypercoagulable state(>78)
3(3.4%) 2(66.7%) 1(33.3%) 0.074
B
Normal(55-78) 60(69%) 13(21.7%) 47(78.3%)
C
Hypocoagulable state (0-54)
23 (26.4%)
11(47.8%) 12(52.2%) 0.019
Tables 17: Comparison of mortality in patients with hypercoagulable, normocoagulable and hypocoagulable state based on maximum amplitude.
Maximum amplitude(n=85)
Total Death Alive A vs B B vs C
A Hypercoagulable state(>69)
27(31.8%) 8 (29.6%))
19(70.4%) 0.769
B
Normal(51-69)) 38(44.7%) 10(26.3%) 28(73.7%)
C
Hypocoagulable state (0-50)
20 (23.5%) 8 (40%) 12(60%) .284
Fig 18: Summarises coagulation abnormalities in patients with severe sepsis based on various parameters in TEG
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
R time K time Alpha angle Maximum
amplitude
Hypercoagulable state Normal coagulation Hypocoagulable state 20
38
27
3 60
23
8 53
25
3 58
26
Secondary outcomes
Table: 18: Demonstrates various secondary outcomes Duration of ICU stay (median and IQR) 7(4-10) Duration of hospital stay 11.5(7-17.5)
Transfusions 28(32.1%)
Platelets 7
Fresh frozen plasma 9
Cryoprecipitates 3
Packed red cells 19
Primary outcome and TEG parameters
Table: 19: Demonstrates correlation between TEG parameters and mortality
No Mortality Mortality p value
R time 6.4253 8.0015 0.032
K value 2.5983 3.9726 0.090
Alpha angle 62.245 51.754 0.014
Maximum amplitude 61.260 55.465 0.099
Lysis 30 1.082 1.560 0.339
TEG parameters and transfusion of blood products
Table 20: Demonstrates association of TEG parameters and requirement of transfusion Transfusions No transfusions p value
R time 7.3620 6.6789 0.344
K value 4.2743 2.3672 0.015
Alpha angle 54.444 61.040 0.182
Maximum amplitude 53.174 62.386 0.026
Lysis 30 1.807 .913 0.126
Transfusion of FFPs
Table 21: Demonstrates association of TEG parameters and requirement of transfusion of fresh frozen plasma
Transfusions No transfusions p value
R time 9.2180 6.6153 0.014
K value 5.5300 2.6995 0.015
Alpha angle 41.240 61.420 0.001
Mean amplitude 45.220 61.388 0.001
Lysis index 2.300 1.079 .082
Mean platelet volume 10.490 10.793 .517
Transfusion of platelets
Table 22: Demonstrates association of TEG parameters and requirement of transfusion of platelets
Transfusions No transfusions p value
R time 8.8333 6.6931 0.055
K value 8.5389 2.3886 0.000
Alpha angle 29.533 62.526 0.000
Mean amplitude 30.389 62.912 0.000
Lysis .667 1.288 .402
Mean platelet volume 10.356 10.805 .359
Transfusion of cryoprecipitates
Table 23: Demonstrates association of TEG parameters and requirement of transfusion of cryoprecipitates
Transfusions No transfusions p value
R time 9.8000 6.7754 .063
K value 12.0000 2.5923 0.000
Alpha angle 23.475 60.810 0.000
Mean amplitude 22.775 61.300 0.000
Lysis index .250 1.270 .343
Mean platelet volume 9.275 10.830 .027
Correlation between conventional tests of coagulation with TEG parameters Table 24: Shows correlation between PT and TEG parameters
PT R time K time Alpha
angle
Maximum amplitude
Lysis index
MPV
Pearson correlation
0.207 0.103 -0.103 -0.209 0.95 -0.153
p Value .059 .353 .098 .058 .395 .168
Table 25: Shows correlation between APTT and TEG parameters
APTT R time K time Alpha
angle
Maximum amplitude
Lysis index
MPV
Pearson correlation
0.26 0.18 -0.24 -0.35 0.32 -0.05
p Value 0.016 0.102 0.027 0.001 0.003 0.626
Table 26: Shows correlation between Platelets and TEG parameters Platelets R time K time Alpha
angle
Maximum amplitude
Lysis index
MPV
Pearson
correlation -0.266 -0.271 0.29 0.51 -0.12 -0.25
p value 0.013 0.011 0.005 0.000 .910 0.018
Correlation between TEG parameters and duration of hospital stay
Table 27 : Demonstrates correlation between various TEG parameters and ICU stay and hospital stay
ICU stay Hospital stay
Pearson Correlation
p value Pearson Correlation
p value
R time -0.205 0.068 -0.175 0.104
K value -0.020 0.860 -0.077 0.479
Alpha angle 0.023 0.842 0.059 0.590
Maximum amplitude 0.078 0.497 0.048 0.658
Lysis index -0.034 0.768 -0.085 0.438
Mean platelet volume -0.083 0.465 -0.005 0.964
TEG parameters and its correlation with SOFA score
Table 28: Demonstrates correlation between various TEG parameters and SOFA score SOFA score
Pearson Correlation p value
R time 0.222 0.070
K value 0.370 0.002
Alpha angle -0.389 0.001
Maximum amplitude -0.441 0.000
Lysis index -0.084 0.500
Mean platelet volume -0.025 0.840
Scrub typhus and TEG
Survivors Non survivors P Value
R time 7.09 7.75 0.749
K time 2.77 4.37 0.408
Alpha angle 59.35 54.90 0.696
Maximum amplitude
62.20 57.72 0.585
Lysis index 0.81 1.80 0.354
MPV 11.48 10.94 0.464
Discussion
Epidemiology of sepsis
Sepsis is defined as life threatening organ dysfunction in response to dysregulated host response to infection(13). Even though it is difficult to ascertain, the global burden of sepsis is estimated to be around 30 million (72). There has been a trend towards increase in the reported incidence of sepsis due to various factors like increase in the elderly population, better identification and reporting of sepsis (13).
In our study 56% of patients were males which were similar to many studies conducted in India. 40% of the patients had diabetes and 36% of the patients had hypertension. The most common cause of sepsis in a study in India was respiratory tract infection(73). In our study the most common cause of sepsis was scrub typhus as the study was conducted during an outbreak of Scrub typhus followed by lower respiratory tract infection.
Bacteraemia was seen in 28.7% of the patients out of which 73% of the patient had gram negative bacteraemia. This was similar to a study by Chatterjee et al in India where 73.4%
of the patients had gram negative sepsis (73).
Sepsis and coagulopathy: hypercoagulable state versus hypocoagulable state
Coagulation abnormalities play a pivotal role in sepsis. It can cause disseminated intravascular coagulation characterised by widespread thrombosis and major bleeding on the other side. In a study by Muzaffar et al patients with severe sepsis were either normocoagulable (56%), hypercoagulable (27%) or hypocoagulable(17%). It was found that patients without shock had a trend towards hypercoagulant state whereas patients with
shock showed a trend towards hypocoagulant state. Coagulation abnormality in these patients were assessed using the coagulation index. Coagulation index <3 was considered to be hypocoagulable and a coagulation index >3 was considered to be hypercoagulable state (34). Coagulation index was not available in our centre and hence statistical analysis was performed with individual parameters of TEG and the outcome. The results of our study showed that based on R, K, alpha angle between 61%-69% of the patients are normocoagulable where as only 44.7% had a normal coagulation based on MA. Even though these patients had no clinical bleeding at the time of recruitment 23-29% of the patients had a hypocoagulable state.
Muller et al in a systematic review found that patients with severe sepsis can have a hyper or hypocoagulable state and that patients with hypocoagulable profile showed a trend towards mortality which was similar in our study(74). The wide variability in TEG results can be explained by the pathophysiology of DIC in sepsis. The course of DIC has 3
consecutive stages.
1. The hypercoagulable state: The stage of active coagulation 2. The stage of coagulation exhaustion
3. The stage of hyper fibrinolysis characterised by clinical bleeding(75).
The coagulation abnormality in patients with severe sepsis may depend upon the time at which TEG was done. Serial assessment of TEG would give a holistic picture of the coagulation abnormality. Due to logistic reasons TEG was done only once within 24 hours of ICU admission.
