The correlation of Rotational Thromboelastometry (ROTEM) derived parameters with severe obstetric haemorrhage

Abstract

Background

Prediction of severe obstetric haemorrhage may facilitate timely interventions and reduce maternal morbidity and mortality. Rotational thromboelastometry (ROTEM) provides point-of-care coagulation parameters that may identify patients at risk of severe haemorrhage.

Purpose

Our aim was to determine the correlation of ROTEM-derived parameters at early stages of postpartum haemorrhage.

Research design

Retrospective cohort study.

Study Sample

Patients who utilised ROTEM during postpartum haemorrhage from 2013 to 2021.

Data Collection and Analysis

Severe harmorrhage was defined as: (1) decrease in haemogoblin concentration >4g.dL^-1 within 48 h; (2) packed red blood cell transfusion of >1000 mL; (3) haemostatic intervention such as angiographic embolisation, arterial ligration, or hysterectomy; or (4) death. ROTEM-derived parameters were collected accordingly.

Results

Data from 36 patients were analysed, of whom 29 (80.6%) developed severe haemorrhage. FIBTEM A5, A10, and A20 were highly correlated (r > +/−0.7) with severe haemorrhage. There were moderate correlations (r = +/− 0.5 to 0.7) between EXTEM A5, A10, A20, maximum clot firmness, maximum lysis, and lysis index-30 with severe haemorrhage.

Conclusions

These findings are consistent with prior studies reporting associations between fibrinogen concentration, EXTEM A10, and FIBTEM A10 with severe haemorrhage. In conjunction with these studies, our results support the use of FIBTEM A5 and A10 as early biomarkers to predict progression to severe haemorrhage and facilitate timely implementation of appropriate medical or surgical interventions.

Keywords

Caesarean massive haemorrhage delivery fibrinogen

Postpartum haemorrhage is a significant cause of maternal morbidity and mortality worldwide, and is one of the leading contributors to pregnancy-related deaths and intensive care unit (ICU) admission.¹ Given that the majority of deaths resulting from postpartum haemorrhage are preventable,² risk stratification to identify patients at risk of severe haemorrhage may facilitate timely implementation of appropriate medical or surgical interventions.

Although fibrinogen levels have been proposed as a biomarker to predict severe haemorrhage,^3,4 laboratory measurements using the Clauss method involves significant turnaround time which delay therapeutic interventions. Rotational thromboelastometry (ROTEM) provides point-of-care coagulation parameters that may guide clinical management and identify patients who are at increased risk of severe obstetric haemorrhage. Specifically, FIBTEM (measurement of fibrin clot strength) amplitude at 5 min (A5) independently predicts progression of obstetric haemorrhage, while FIBTEM and EXTEM (measurement of both fibrin and platelet-based clot strength) amplitude at 10 min (A10) correlate well with fibrinogen levels in mothers with obstetric haemorrhage.^5,6 Nonetheless, the roles of other ROTEM-derived parameters in predicting severe obstetric haemorrhage are unclear. Hence, our goal was to determine the correlation between ROTEM parameters (FIBTEM A5, A10, A20; EXTEM alpha angle, A5, A10, A20; maximum clot firmness (MCF); clotting time (CT); clot formation time (CFT); maximum lysis (ML); and lysis index at 30 min (LI 30)) measured in the early stages of postpartum haemorrhage, with severe haemorrhage (primary outcome measure) defined as: (1) decrease in haemoglobin concentration >4 g.dL⁻¹ within 48 h; (2) packed red blood cell transfusion of >1000 mL; (3) haemostatic intervention such as angiographic embolization, arterial ligation, or hysterectomy; or (4) death.³ We also investigated correlations between (1) ROTEM-derived parameters and (2) lowest fibrinogen concentration within 48 h, with lowest haemoglobin concentration within 48 h and estimated blood loss (secondary outcome measures).

After obtaining SingHealth Combined Institutional Review Board waiver of consent (2022/2244 on 6 June 2022) we retrospectively analysed a convenient sample of adult patients where ROTEM was utilised during postpartum haemorrhage management between 2013 and 2021 at KK Women’s and Children’s Hospital, Singapore. Patients receiving anticoagulant and antiplatelet medications were excluded. Although clinical management and ROTEM utilisation were at the discretion of the attending anaesthesiologist, institutional practice recommends activation of massive transfusion protocol and early use of ROTEM in patients with rapid blood loss, ongoing obstetric bleeding with haemodynamic instability, and clinical evidence of coagulopathy. The following data were extracted from the electronic medical record: patient characteristics, obstetric comorbidities, aetiology of haemorrhage, pharmacologic and surgical haemorrhage management (e.g. uterotonic agents, tranexamic acid, B-Lynch suture, hysterectomy, etc.), volume of intravenous fluids and blood products administered, estimated blood loss (visually estimated at end of resuscitation when haemostasis was achieved), blood investigations (e.g. haemoglobin concentration, prothrombin time (PT), activated partial thromboplastin time (aPTT), fibrinogen concentration, etc.), ICU admission, and maternal mortality. The first recorded ROTEM parameters were used if repeated tests were performed. In the case of obstetric haemorrhage during caesarean delivery, institutional practice was to administer oxytocin three units or carbetocin 100 µg after foetal delivery, with second-line uterotonic agents consisting of intravenous infusion of oxytocin 30 units over 4 h, intramuscular ergometrine, and intramuscular prostaglandin F2α.

Demographic and clinical characteristics were summarised based on severe haemorrhage status. Continuous characteristics were summarised as median [interquartile range] (range) while categorical characteristics were summarised as frequency (percentages). Primary outcome of severe haemorrhage was treated as binary data and secondary outcomes lowest haemoglobin concentration within 48 h (g.dL⁻¹) and estimated blood loss (mL) were treated as continuous data. Bi-serial correlation was calculated for severe haemorrhage and other continuous variables, while Pearson’s correlation was calculated between secondary outcomes and other continuous variables. All tests were two-sided and statistical significance was set at p-value <.05. All analyses were performed using SAS 9.4.

Data from 36 patients were analysed, of whom 29 (80.6%) developed severe haemorrhage. Patient characteristics and haemorrhage management are summarised in Table 1, and our results are summarised in Table 2. We found that FIBTEM A5, A10, and A20 were highly correlated (r > +/−0.7) with severe haemorrhage. There were moderate correlations (r = +/− 0.5 to 0.7) between EXTEM A5, A10, A20, MCF, ML, LI 30, and lowest fibrinogen concentration within 48 h with severe haemorrhage. However, all ROTEM parameters analysed were poorly correlated (r < +/−0.4) with our secondary outcomes of lowest haemoglobin concentrations within 48 h and estimated blood loss.

Table 1.

Patient or clinical characteristics and management of postpartum haemorrhage.

Variable	Severe haemorrhage (n = 29)	No severe haemorrhage (n = 7)
Age in years, median [IQR] (range)	36 [7] (25 to 45)	34 [9] (28 to 42)
Body mass index in kg.m⁻², median [IQR] (range)	25.0 [6.1] (20.0 to 50.0)	27.0 [15.7] (18 to 39)
Race/ethnicity, n (%)
Chinese	18 (62.1)	2 (28.6)
Malay	8 (27.6)	3 (42.9)
Indian	0	1 (14.3)
Other	3 (10.3)	1 (14.3)
Number of previous caesarean deliveries, median [IQR] (range)	1 [1] (0 to 3)	1 [1] (0 to 2)
Previous abdominal surgery, n (%)	1 (3.4)	2 (28.6)
Gravidity, median [IQR] (range)	3 [2] (1 to 6)	2 [2] (2 to 7)
Parity, median [IQR] (range)	1 [1] (0 to 4)	1 [1] (0 to 3)
Gestational age in weeks, median [IQR] (range)	34.0 [9.6] (17 to 40)	38.3 [3.0] (24 to 39)
Multiple gestations, n (%)	2 (6.9)	0
Comorbidity, n (%)^a
Hypertensive disease (existing or pregnancy-related)	2 (6.9)	3 (42.9)
Diabetes (existing or gestational)	2 (6.9)	1 (14.3
Gestational thrombocytopenia	0	0
Blood parameters prior to haemorrhage, median [IQR] (range)
Haemoglobin concentration in g.dL⁻¹	11.3 [2.4] (9.0 to 13.0)^b	10.8 [1.3] (9.0 to 12.0)
Platelet count in 109.L⁻¹	235 [114] (111 to 384)^b	227 [119] (139 to 473)
International normalised ratio	1.0 [0.1] (1.0 to 2.0)	1.2 [0.3] (1.0 to 2.0)
Activated partial thromboplastin time in seconds	34.9 [7.2] (30.0 to 51.0)^c	33.8 [2.5] (32.0 to 39.0)^d
Induction/augmentation of labour, n (%)^a	4 (13.8)	2 (28.6)
Mode of delivery, n (%)
Vaginal	9 (31.0)	0
Scheduled caesarean	3 (10.3)	4 (57.1)
Emergency caesarean	17 (58.6)	3 (42.9)
Instrumental delivery, n (%)	1 (3.4)	0
Mode of anaesthesia, n (%)
Neuraxial anaesthesia	3 (10.3)	2 (28.6)
Neuraxial converted to general anaesthesia	1 (3.4)	2 (28.6)
General anaesthesia	25 (86.2)	3 (42.9)
Uterotonics administered
First-line uterotonic agents, n (%)	18 (62.1)	7 (100)
Oxytocin bolus dose in IU, median [IQR] (range)	0 [3] (0 to 12)	3 [2] (3 to 20)
Carbetocin 100 µg, n (%)	18 (62.1)	7 (100)
Second-line uterotonic agents, n (%)	17 (58.6)	2 (29.6)
Oxytocin 40 IU over 4 h infusion, n (%)	12 (41.4)	6 (85.7)
Prostaglandin F2α in ug, median [IQR] (range)	0 [250] (0 to 1000)	250 [500] (0 to 500)
Ergometrine 500 µg with oxytocin 5 IU, n (%)	3 (10.3)	2 (28.6)
Surgical management of haemorrhage, n (%)^d
None	8 (27.6)	5 (71.4)
Uterine artery balloon/embolisation	5 (17.9)	0
B-Lynch suture	4 (13.8)	0
Bakri balloon	7 (24.1)	1 (14.3)
Hysterectomy	13 (44.8)	-
Intravenous fluids administered in mL, median [IQR] (range)
Crystalloids	2000 [1900] (0 to 4500)	1500 [2000] (1000 to 4000)
Colloids	500 [500] (0 to 1500)	0 [250] (0 to 500)
Blood products administered, median [IQR] (range), n (%)
Packed red cells in mL	1600 [800] (0 to 5000), 29 (100)	500 [300] (0 to 1000), 4 (57.1)
Fresh frozen plasma in mL	841 [589] (0 to 2642), 24 (82.8)	0 [500] (0 to 593), 3 (42.9)
Pooled platelets in mL	325 [68] (0 to 827), 23 (79.3)	0 [0] (0 to 335), 1 (14.3)
Cryoprecipitate in IU	0 [10] (0 to 66), 12 (41.4)	0 [0] (0 to 10), 1 (14.3)
Fibrinogen concentrate in g	0 [0] (0 to 4), 8 (27.6)	NIL
Recombinant factor 7 (Novoseven) in µg	0 [0] (0 to 8000), 2 (6.9)	NIL
Tranexamic acid dose in mg, median [IQR] (range)	1000 [500] (0 to 2000)	1000 [0] (0 to 1000)
Cell saver use, n (%)	3 (10.3)	1 (14.3)
Intensive care unit admission, n (%)	27 (93.1)	4 (57.1)
Estimated blood loss in mL, mean (SD)	2700 (1810)	1600 (880)
Cause of haemorrhage, n (%)^a
Placental abruption	8 (27.6)	2 (28.6)
Placental accreta spectrum	11 (37.9)	1 (14.3)
Amniotic fluid embolism	1 (3.4)	0
Vascular injury or significant tissue trauma	6 (20.7)	2 (28.6)
Intrauterine death	4 (13.8)	1 (14.3)
Uterine atony	2 (6.9)	3 (42.9)
Disseminated intravascular coagulation	3 (10.3)	1 (14.3)
ROTEM parameters, median [IQR] (range)
FIBTEM A5 in mm	7 [7] (3 to 14)^e	12.0 [11] (10 to 34)
FIBTEM A10 in mm	8 [7] (3 to 18)^f	14 [11] (10 to 36)
FIBTEM A20 in mm	8 [8] (3 to 16)^g	16 [11] (12 to 39)^d
EXTEM alpha angle in degrees	58 [30] (11 to 76)^d	75 [10] (60 to 77)^d
EXTEM A5 in mm	28 [23] (5 to 53)^h	48 [11] (28 to 54)^d
EXTEM A10 in mm	39 [22] (8 to 64)^d	58 [10] (40 to 64)^d
EXTEM A20 in mm	47 [19] (11 to 68)^b	64 [9] (49 to 67)^h
Clotting time (CT) in seconds	82 [74] (48 to 895)^d	55 [4] (51 to 60)^d
Maximum clot firmness (MCF) in mm	50 [20] (16 to 70)^d	64 [9] (50 to 69)^d
Clot formation time (CFT) in seconds	180 [227] (66 to 517)^b	76 [54] (63 to 193)^d
EXTEM maximum lysis (ML) in %	0 [4] (0 to 23)^h	9 [15] (0 to 23)^d
EXTEM lysis index at 30 min (LI 30)	100 [0] (100 to 100)ⁱ	100 [0] (93 to 100)^h
Blood parameters within 48 h, median [IQR] (range)
Lowest haemoglobin concentration in g.dL⁻¹	7.4 [2.2] (4.0 to 12.0)^h	7.9 [1.3] (7.0 to 10.0)
Lowest platelet count in 109.L⁻¹	106 [83] (32 to 201)^h	161 [69] (82 to 278)
Highest international normalised ratio	1.5 [1.0] (1.0 to 15.0)^j	1.2 [0.2] (1.0 to 1.5)^d
Highest activated partial thromboplastin time in seconds	46.0 [43.6] (32.0 to 180.0)^b	45.5 [10.9] (34.0 to 56.0)^d
Lowest fibrinogen in g.dL⁻¹	1.3 [1.3] (0.4 to 4.7)^g	3.4 [6.0] (0.4 to 6.4)^j

^aEach patient may have multiple variables present.

^bMissing data from three patients.

^cMissing data from 14 patients.

^dMissing data from one patient.

^eMissing data from eight patients.

^fMissing data from seven patients.

^gMissing data from nine patients.

^hMissing data from two patients.

ⁱMissing data from 11 patients.

^jMissing data from five patients.

Table 2.

Correlation of ROTEM parameters with postpartum haemorrhage outcomes.

ROTEM parameter	Correlation coefficient
ROTEM parameter	Severe haemorrhage	Lowest haemoglobin concentration within 48 h (g.dL⁻¹)	Estimated blood loss (mL)
FIBTEM A5 in mm	−0.827	0.036	0.055
FIBTEM A10 in mm	−0.774	0.040	0.046
FIBTEM A20 in mm	−0.855	0.031	0.051
EXTEM alpha angle in degrees	−0.535	0.263	0.000
EXTEM A5 in mm	−0.609	0.228	0.007
EXTEM A10 in mm	−0.591	0.248	0.012
EXTEM A20 in mm	−0.544	0.225	0.021
Clotting time (CT) in seconds	0.289	−0.400	0.188
Maximum clot firmness (MCF) in mm	−0.524	0.270	0.008
Clot formation time (CFT) in seconds	0.495	0.266	0.038
Maximum lysis (ML) in %	−0.593	−0.040	−0.122
Lysis index at 30 min (LI 30)	0.555	−0.125	−0.004
Lowest fibrinogen within 48 h in g.dL⁻¹	−0.577	0.223	0.089

In summary, our results show that FIBTEM A5, A10, and A20 were highly correlated with the development of severe haemorrhage. There was also moderate correlation between EXTEM parameters and lowest fibrinogen concentration within 48 h. Overall, these findings are consistent with prior studies reporting associations of fibrinogen, EXTEM A10, and FIBTEM A10 with severe haemorrhage.^3,4,6 Furthermore, in accordance with recommendations by Collins et al and Toffaletti et al.,^5,6 our results support the use of FIBTEM A5 or A10 as early biomarkers to predict progression to severe obstetric haemorrhage, especially given the long turnaround time associated with laboratory measurements of fibrinogen concentrations that may delay appropriate interventions.

We acknowledge several limitations such as the retrospective study design, small sample size, and missing ROTEM data including eight patients for FIBTEM A5 and seven patients for FIBTEM A10, which could have affected the accuracy of our results. Of note, however, our findings were consistent with prior studies and recommendations.^3,5,6 Second, our definition of severe haemorrhage was based on that used by Charbit et al.,³ which differs from the American College of Obstetricians and Gynecologists (ACOG) definition of postpartum haemorrhage as cumulative blood loss greater than 1000 mL or blood loss accompanied by signs or symptoms of hypovolemia within 24 h of delivery.⁷ Notably, our definition included clinically-significant morbidity and invasive procedures, and is objective and reproducible. This strengthens the association between FIBTEM parameters and poor haemorrhage-related outcomes, in addition to providing important generalisability of our findings to prior studies in this field.³ Third, rapid changes in patient condition, clinical management, and levels of coagulation factors during obstetric haemorrhage could have influenced our findings, especially given that ROTEM and other investigations were not performed at specific timepoints. Nonetheless, the points at which obstetric haemorrhage is recognised and clinical resuscitation instituted are hugely variable in clinical practice, and our study methodology reflects this variability. Finally, it is possible that the correlation between fibrinogen concentration and severe haemorrhage may have been affected by the administration of cryoprecipitate and fibrinogen concentrate during clinical resuscitation.

To conclude, FIBTEM A5, A10, and A20 were highly correlated with severe obstetric haemorrhage. Notwithstanding the limitations of this study, our findings suggest that future research should focus on elucidating the role of FIBTEM in risk stratification for severe obstetric haemorrhage, with the aim to facilitate timely implementation of appropriate medical or surgical interventions.

Footnotes

Acknowledgments

We would like to acknowledge the contribution of the members of the Department of Women’s Anaesthesia at KK Women’s and Children’s Hospital, Singapore for their assistance with this study.

Author contributions

H.S.T.: Conceptualisation, methodology, investigation, writing – original draft, review, and editing. J.J.L.Y.: Conceptualisation, methodology, investigation, writing – original draft, review, and editing. T.K.A.: Conceptualisation, methodology, investigation, writing – review and editing. R.S.: Methodology, formal analysis. C.W.T.: Conceptualisation, methodology, writing – review and editing. B.L.S.: Conceptualisation, methodology, investigation, writing – review and editing, supervision.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethics statement

ORCID iDs

Chin Wen Tan

Ban Leong Sng

Data availability statement

The datasets generated and/or analysed during this study are not publicly available but are available from the corresponding author on reasonable request.

References

Kassebaum

Barber

Bhutta

. Global, regional, and national levels of maternal mortality, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet 2016; 388: 1775–1812.

Berg

Harper

Atkinson

, et al. Preventability of pregnancy-related deaths: results of a state-wide review. Obstet Gynecol 2005; 106: 1228–1234.

Charbit

Mandelbrot

Samain

, et al. The decrease of fibrinogen is an early predictor of the severity of postpartum hemorrhage. J Thromb Haemost 2007; 5: 266–273.

Collins

Solomon

Sutor

, et al. Theoretical modelling of fibrinogen supplementation with therapeutic plasma, cryoprecipitate, or fibrinogen concentrate. Br J Anaesth 2014; 113: 585–595.

Collins

Lilley

Bruynseels

, et al. Fibrin-based clot formation as an early and rapid biomarker for progression of postpartum hemorrhage: a prospective study. Blood 2014; 124: 1236–1727.

Toffaletti

Buckner

. Use of earlier-reported rotational thromboelastometry parameters to evaluate clotting status, fibrinogen, and platelet activities in postpartum hemorrhage compared to surgery and intensive care patients. Anesth Analg 2019; 128: 414–423.

Menard

Main

Currigan

. Executive summary of the reVITALize initiative: standardizing obstetric data definitions. Obstet Gynecol 2014; 124: 150–153.