Evaluating the Association Between Fibrinogen and Rotational Thromboelastometry and the Progression to Severe Obstetric Hemorrhage

Abstract

Aim

This study aims to investigate the ability of fibrinogen and rotational thromboelastometry (ROTEM) parameters measured at obstetric hemorrhage protocol initiation to predict severe hemorrhage.

Methods

In this retrospective study we included patients whose hemorrhage was managed with an obstetric massive transfusion protocol. Fibrinogen and ROTEM parameters EXTEM clotting time (CT), clot formation time (CFT), alpha angle, A10, A20, lysis index 30 min after CT (LI30), FIBTEM A10, A20, were measured at initiation of the protocol with transfusion based on a predefined algorithm. Patients were grouped into either severe or nonsevere hemorrhage based on: peripartum fall in hemoglobin ≥4 g/dL, transfusion of ≥4 units of blood product, invasive procedures for hemorrhage control, intensive care unit admission, or death.

Results

Of the 155 patients included, 108 (70%) progressed to severe hemorrhage. Fibrinogen, EXTEM alpha angle, A10, A20, FIBTEM A10, A20 were significantly lower in the severe hemorrhage group while the CFT was significantly prolonged in the severe hemorrhage group. In univariate analysis, predicted progression to severe hemorrhage yielded areas under the receiver operating characteristic curve (95% confidence interval [CI]) of: fibrinogen: 0.683 (0.591-0.776), CFT: 0.671 (0.553, 0.789), EXTEM alpha angle: 0.690 (0.577-0.803), A10: 0.693 (0.570-0.815), A20: 0.678 (0.563-0.793), FIBTEM A10: 0.726 (0.605-0.847), and A20: 0.709 (0.594-0.824). In a multivariable model, fibrinogen was independently associated with severe hemorrhage (odds ratio [95% CI] = 1.037 [1.009-1.066]) for every 50 mg/dL decrease in fibrinogen drawn at obstetric hemorrhage massive transfusion protocol initiation.

Conclusion

Both fibrinogen and ROTEM parameters measured at the initiation of an obstetric hemorrhage protocol are useful parameters for predicting severe hemorrhage.

Introduction

Obstetric hemorrhage still contributes significantly to global maternal morbidity and mortality. ¹ In the United States, obstetric hemorrhage occurring from the day of delivery to 6 days postpartum, is both a leading cause of pregnancy-related deaths ² and its associated morbidity is one of the leading causes of intensive care unit (ICU) admission.^3,4 An in-depth review of maternal mortality has demonstrated that the majority of maternal deaths from obstetric hemorrhage are preventable. ⁵ One strategy aimed at preventing maternal mortality and morbidity resulting from major obstetric hemorrhage involves early identification of patients who are at risk of progressing to severe hemorrhage. Early identification allows for earlier implementation of transfusion strategies, and medical and surgical interventions that may prevent this progression to severe obstetric hemorrhage and improve maternal outcomes.

Early measurement of plasma fibrinogen levels may serve as a biomarker which predicts progression to severe hemorrhage.^6,8 Plasma fibrinogen levels fall to critically low levels early in obstetric hemorrhage which may be associated with progression to severe obstetric hemorrhage. This fall in fibrinogen may result from a combination of increased fibrinogen consumption, increased fibrinolysis and fibrinogenolyis, and hemodilution from resuscitation.^9,10 Charbit et al ⁶ demonstrated that fibrinogen levels below 200 mg/dL at the start of a hemorrhage have a 100% positive predictive value of progression to severe hemorrhage. However higher fibrinogen levels may also be associated with progression to severe hemorrhage. ¹¹ This was demonstrated in previous work in which even patients with fibrinogen concentrations between 200 and 300 mg/dL were at risk of progression to severe hemorrhage.^8,11 Unfortunately, the laboratory turnaround time needed for fibrinogen measurements using the Clauss method could significantly delay therapeutic interventions that may reduce maternal morbidity and mortality. Rapid fibrinogen kits based on lateral flow technology in which fibrinogen is determined by distance traveled before flow cessation from clotting can be used to measure fibrinogen in whole blood and plasma and are now commercially available. These kits use a very small sample volume (15 μL), can provide results in <5 min, and are capable of determining fibrinogen concentration in the range of 1 to 7 g/L. ¹² However these kits provide qualitative rather than quantitative data, so an exact value cannot be determined.

An alternative method for monitoring coagulation involves quantifying the viscoelastic properties of whole blood coagulation using real-time techniques such as rotational thromboelastometry (ROTEM). ROTEM provides clinically relevant parameters of hemostasis quicker than standard coagulation tests particularly when used as point of care testing.^7,13 Furthermore the turnaround times of ROTEM have been greatly improved by the ability of clinical care teams to remotely view the temogram. Both EXTEM, which measures both fibrin and platelet-based clot strength, and FIBTEM which measures fibrin-based clot strength following platelet inhibition in whole blood, have emerged as clinically relevant parameters in the obstetric population. A recent study demonstrated that the earlier reported EXTEM alpha angle (α-EXTEM), EXTEM clot amplitude at 10 or 20 min (EXTEM A10 or A20), and FIBTEM clot amplitude at 10 or 20 min (A10 or A20) correlated well with fibrinogen levels in women with a postpartum hemorrhage. ¹³ Another study in patients with obstetric hemorrhage also found that plasma fibrinogen levels were strongly correlated with EXTEM alpha angle, A10 and A20. ⁸ While the early measurement of FIBTEM A5 has been shown to independently predict the progression of postpartum hemorrhage, the roles of the other ROTEM parameters in predicting the progression of obstetric hemorrhage are less clearly defined. ⁷

The aim of this study was to predict severe hemorrhage based on both fibrinogen and ROTEM parameters measured at the initiation of an obstetric hemorrhage massive transfusion protocol (OB-MTP) in a cohort of obstetric patients. With this information, our ultimate goal was to identify clinically relevant parameters that could be used to stratify patients at risk for hemorrhage-related adverse outcomes.

Methods

This study was approved by the Duke University School of Medicine Institutional Review Board. In this retrospective observational study, we included as a convenience sample all patients who had a clinically significant obstetric hemorrhage managed with an OB-MTP from September 2012 to May 2018 at the Duke Birthing Center, a tertiary delivery unit. As a tertiary delivery unit and a regional placental accreta spectrum center, the Duke Birthing Center manages patients with obstetric comorbities that increase the risk of hemorrhage. We excluded patients who were transferred to our institution who had their hemorrhage management initiated at an outside hospital or cases where we were unable to extract laboratory data due to a change in the electronic medical records (EMR) system used at our institution. The obstetric massive transfusion protocol was activated when patients developed a clinically significant hemorrhage loosely defined as estimated blood loss of >1000 to 1500 mL, ongoing obstetric bleeding with hemodynamic instability, initial clinical evidence of a coagulopathy or based on the clinical judgment of the healthcare team. Following OB-MTP activation, blood samples were drawn at OB-MTP activation and at 30min intervals for monitoring of complete blood count, prothrombin time, activated partial thromboplastin time, fibrinogen (measured using a modification of the Clauss method) and arterial or venous shock panel. Beginning in September 2014 the protocol was modified to include the ROTEM, EXTEM, and FIBTEM laboratory results used for monitoring at similar time intervals. Initially, the only parameter of clot firmness that was reported by our laboratory was the A20 for both the EXTEM and FIBTEM, followed a few months later by the A10. Patients were transfused based on a laboratory-based algorithm or a fixed ratio transfusion regime in the case of a massive hemorrhage (Figure 1). The ROTEM values highlighted in the lab-based algorithm were derived from the normal ranges in our local population, projected changes with blood loss and observed changes in parameters in response to transfusion of blood products. Obstetric and surgical management was at the discretion of the managing clinical team. A quality improvement (QI) database was created for all patients who had an obstetric hemorrhage managed by the OB-MTP.

Figure 1.

Massive transfusion protocol algorithm.

Data Collected

From both the EMRs and the QI database we extracted the following data:

• demographic data;

• relevant obstetric history;

• obstetric comorbidities;

• mode of delivery;

• hemorrhage etiology (documented by the obstetrician managing the hemorrhage);

• oxytocic agents used;

• estimated blood loss;

• volume of intravenous fluids administered during the resuscitation period;

• tranexamic acid (TXA) use;

• fibrinogen concentrate use;

• blood products transfused during OB-MTP and during the remainder of the hospitalization period;

• interventions for hemorrhage control (B-Lynch suture, hysterectomy, arterial embolization);

• ICU admission;

• maternal death.

The date of the hemorrhage was extracted from the QI database. The approximate time of OB-MTP activation was determined either by a time stamped documentation by a healthcare provider or the time of the first OB-MTP laboratory panel blood draws documented in the EMR. Blood loss was primarily estimated by visual estimation and included the cumulative blood loss from delivery until the end of resuscitation when bleeding was controlled. We extracted baseline laboratory data (24 h before date of hemorrhage) and all OB-MTP laboratory data for 48 h starting on the date of the hemorrhage.

Stratification Into Severe Versus Nonsevere Hemorrhage

Included patients were stratified into severe or nonsevere obstetric hemorrhage where a severe hemorrhage, based on previously published criteria,^6,11,14 was defined as:

• Peripartum fall in hemoglobin ≥4 g/dL (determined by the difference in baseline hemoglobin and the nadir hemoglobin measured 48 h from the date of the hemorrhage);

• Transfusion of ≥4 units of blood products ¹⁴ ;

• Need for invasive procedures for hemorrhage control (B-Lynch suture, arterial embolization, hysterectomy);

• ICU admission;

• Death.

Statistical Analysis

The primary aim of our study was to predict severe hemorrhage using fibrinogen levels measured at the activation of the OB-MTP. Secondarily we aimed to determine the ability of EXTEM clotting time (CT), clot formation time (CFT), A10, A20, LI30, α-EXTEM, FIBTEM A10, and A20 also drawn at OB-MTP activation to predict severe hemorrhage. For all the laboratory parameters drawn at the initiation of the OB-MTP we also planned to determine optimal cutoff values that could be used to predict progression to severe obstetric hemorrhage along with the negative (NPV) and positive predictive values (PPV) for each parameter. Patient demographics and characteristics were stratified by severe hemorrhage status and group comparisons were made using the t-tests or Mann-Whitney U tests for continuous data and Chi-squared tests or Fischer's exact tests for categorical data. Univariate logistic regression models were fit to determine the association between fibrinogen levels and ROTEM parameters with severe hemorrhage. Additionally, we planned to fit multivariable logistic models adjusting for initial fibrinogen levels measured at OB-MTP activation, age, vaginal delivery versus C-section, TXA administration, and presence of uterine atony. The adjustment parameters included in the multivariable analysis were specified a priori based on risk factors associated with obstetric hemorrhage or in the case of TXA a therapeutic intervention which may reduce hemorrhage severity.^15,18 The odds ratios (ORs) for severe hemorrhage were calculated for every 50 mg/dL decrease in fibrinogen, every mm decrease in the EXTEM A10, A20, FIBTEM A10, and A20, every degree decrease in α-EXTEM, every percentage decrease in LI30 and every second decrease in CFT and CT measured at OB-MTP activation where appropriate. Models were assessed for validity and predictive performance using Hosmer-Lemeshow tests and by examining the area under the receiver operating curves.

Results

For the study period we initially identified 165 eligible patients. However, in 9 patients we were unable to access any laboratory data due to a healthcare system wide change in the EMR that limited access to archived data and 1 patient had her hemorrhage management initiated at an outside hospital. The remaining 155 patients were retrospectively classified into 108 patients with a severe hemorrhage and 47 patients with a nonsevere hemorrhage (Figure 2). All 155 patients had fibrinogen levels measured at OB-MTP activation. A subgroup of 112 patients had ROTEM parameters measured at OB-MTP activation reflecting the later modification to the protocol which introduced ROTEM measurements in the OB-MTP laboratory tests. Transfusion of ≥4 units of blood products was the most common classification variable for which women were stratified into the severe hemorrhage group (Figure 2). Patient demographics and obstetric outcomes were similar between the severe and nonsevere hemorrhage group, as shown in Table 1. Overall uterine atony was the commonest cause for an obstetric hemorrhage in this cohort, but the incidence was not different between the groups; and, both abnormal placentation and coagulopathies were more frequently observed in the severe hemorrhage group (Figure 3).

Figure 2.

Patients included in the study were stratified into severe or nonsevere hemorrhage based on prespecified criteria.

Table 1.

Subject Characteristics Stratified by Hemorrhage Severity and Displayed as n (%) or Median [Q1, Q3].

	Nonsevere hemorrhage (N = 47)	Severe hemorrhage (N = 108)	P value
Age (years)	31.0 [28.0, 34.0]	32.0 [28.0, 36.0]	.284
BMI (kg/m2)	33.7 [27.8, 40.2]	32.6 [27.9, 37.8]	.417
Race/ethnicity			.444
Hispanic	9 (19.1)	27 (25.0)
Non-Hispanic Black	17 (36.2)	35 (32.4)
Non-Hispanic White	18 (38.3)	32 (29.6)
Other	3 (6.4)	14 (13.0)
Insurance status			.856
Private	19 (40.4)	47 (43.5)
Public/nonprivate	28 (59.6)	61 (56.5)
Gestational age (weeks)	38.6 [36.0, 39.5]	38.0 [34.6, 39.9]	.538
Gravidity	3.0 [1.0, 5.00]	3.0 [2.0, 4.0]	.797
Parity	1.0 [0.0, 3.0]	1.0 [0.0, 2.0]	.470
Multiple gestation	7 (14.9)	12 (11.1)	.694
History of previous cesarean section	12 (25.5)	43 (39.8)	.127
Augmented/induced delivery	22 (46.8)	41 (38.0)	.394
Mode of delivery			.070
Scheduled Cesarean section	9 (19.1)	21 (19.4)
Unscheduled Cesarean section	17 (36.2)	58 (53.7)
Vaginal delivery	21 (44.7)	29 (26.9)
Additional oxytocic agents administered	39 (83.0)	80 (74.1)	.228

Comparisons between hemorrhage groups were evaluated with either Chi-square or Fischer's exact tests for categorical variables and t-tests or Mann-Whitney U tests for continuous variables.

Abbreviation: BMI, body mass index.

Figure 3.

The causes of hemorrhage grouped by severity status. The height of the bars indicates the number of cases, with the percentage of each hemorrhage group by cause indicated on the y-axis.

Compared to the non-severe hemorrhage group, the following measurements pertaining to blood products transfused, intravenous fluids, and hemostatic adjuncts administered were higher in the severe hemorrhage group: median estimated blood loss, administered volumes of intravenous crystalloids and colloids, the proportion of women transfused packed red blood cells (PRBCs), thawed plasma (TP), platelets and cryoprecipitate, and the median number of units of PRBCs, TP, and cryoprecipitate transfused (Table 2). Additionally, of the women who experienced a severe hemorrhage, 62% received tranexamic acid and 15% received fibrinogen concentrate compared to only 36% and 6%, respectively, among the women in the nonsevere hemorrhage group (Table 2).

Table 2.

Blood Products Transfused, Intravenous Fluids and Adjuncts Administered Stratified by Hemorrhage Severity and Displayed as n (%) or Median [Q1, Q3].

	Nonsevere hemorrhage (N = 47)	Severe hemorrhage (N = 108)	P value
Estimated blood loss (mL)	2000 [1500, 2300]	2675 [1975, 3500]	<.001
Total volume of crystalloids (mL) a	2400 [1450, 3000]	3000 [2225, 4200]	<.001
Total volume of colloids (mL) a	0 [0, 250]	250 [0, 500]	<.001
Received PRBC	25 (53.2)	95 (88.0)	<.001
Units of PRBC transfused	1 [0, 2]	3 [2, 4]	<.001
Received TP	2 (4.3)	66 (61.1)	<.001
Units of TP transfused	0 [0, 0]	1 [0, 2]	<.001
Received platelets	2 (4.3)	32 (29.6)	.001
Units of platelets transfused	0 [0, 0]	0 [0, 1]	<.001
Received cryoprecipitate	4 (8.5)	63 (58.3)	<.001
Units of cryoprecipitate transfused b	0 [0, 0]	1 [0, 2]	<.001
Received tranexamic acid	17 (36.2)	67 (62.0)	.005
Received fibrinogen concentrate	3 (6.4)	16 (14.8)	.228

Comparisons between hemorrhage groups are evaluated with either Chi-square or Fischer's exact tests for categorical variables and t-tests or Mann-Whitney U tests for continuous variables.

^a Volume administered during the initial resuscitation only.

^b Refers to pooled units.

Abbreviation: PRBC, packed red blood cells; TP, thawed plasma.

The laboratory parameters of interest stratified by hemorrhage group are summarized in Table 3. In the severe hemorrhage group, fibrinogen level, hemoglobin, and platelet count at OB-MTP activation were all lower compared to the nonsevere hemorrhage group. In the subgroup of patients with ROTEM data, the EXTEM A10, A20, α-EXTEM, FIBTEM A10, and A20 were also lower in the severe hemorrhage group compared to the nonsevere hemorrhage group. The EXTEM CFT was prolonged in the severe hemorrhage group when compared with the nonsevere group.

Table 3.

Laboratory Parameters Stratified by Hemorrhage Severity Displayed as Median [Q1, Q3].

Parameters	Nonsevere hemorrhage (N = 47)	Severe hemorrhage (N = 108)	P value
Fibrinogen (mg/dL)	310 [268, 426]	264 [177, 324]	< .001
Hemoglobin (g/dL)	9.5 [9.0, 10.6]	8.5 [7.1, 10.2]	= .007
Platelet count (× 109/L)	179 [145, 247]	164 [130, 210]	= .037
	Nonsevere hemorrhage (N = 25)	Severe hemorrhage (N = 72)
FIBTEM A10 (mm)	20 [15, 26]	15 [11, 20]	< .001
EXTEM A10 (mm)	61 [53, 67]	56 [46, 61]	< .001
	Nonsevere hemorrhage (N = 29)	Severe hemorrhage (N = 83)
EXTEM alpha angle (°)	74 [71, 80]	72 [65, 75]	< .001
FIBTEM A20 (mm)	22 [16, 28]	16 [11, 21]	< .001
EXTEM A20 (mm)	67 [60, 72]	62 [54, 67]	< .001
Clotting time (s)	55 [49, 62]	57 [52, 65]	= .117
Clot formation time (s)	82 [55, 104]	93 [76, 136.5]	= .048
	Nonsevere hemorrhage (N = 27)	Severe hemorrhage (N = 79)
EXTEM LI30 (%)	100 [100, 100]	100 [100, 100]	= .176

In the univariate analysis of laboratory parameters measured at the activation of the OB-MTP, fibrinogen was found to be significantly associated with severe obstetric hemorrhage [OR (95% confidence interval [CI])  = 1.278 (1.102, 1.500)], along with CFT, EXTEM A10, A20, α-EXTEM, FIBTEM A10, and A20. Additionally, the area under the receiver operating characteristic (ROC) for fibrinogen, CFT, EXTEM A10, A20, α-EXTEM, FIBTEM A10, and A20 demonstrated that all these parameters performed similarly in predicting severe hemorrhage (Table 4).

Table 4.

Results of Univariate Logistic Regression and Performance Characteristics of the Model.

	Progression to severe hemorrhage			ROC
Predictor	OR	95% CI of OR	P value	AUC	95% CI of AUC
Fibrinogen (mg/dL) a	1.278	(1.102, 1.500)	.002	0.684	(0.591, 0.777)
FIBTEM amplitude at 10 min (mm) b	1.162	(1.070, 1.271)	<.001	0.726	(0.605, 0.847)
FIBTEM amplitude at 20 min (mm) b	1.140	(1.062 1.238)	<.001	0.709	(0.594, 0.824)
EXTEM alpha angle (°) b	1.126	(1.044, 1.242)	.008	0.690	(0.577, 0.804)
EXTEM amplitude at 10 min (mm) b	1.086	(1.029, 1.161)	.007	0.693	(0.570, 0.815)
EXTEM amplitude at 20 min (mm) b	1.097	(1.034, 1.180)	.006	0.678	(0.563, 0.793)
EXTEM LI30 (%) b	0.767	(0.349, 1.166)	.282	0.494	(0.432, 0.555)
Clotting time (s) c	1.018	(1.000, 1.052)	.174	0.579	(0.459, 0.699)
Clot formation time (s) c	1.017	(1.004, 1.034)	.027	0.671	(0.553, 0.789)

^a Per 50 mg/dL decrease.

^b Per unit decrease.

^c Per unit increase.

Abbreviations: AUC, area under the receiver operating curve; CI, confidence interval; OR, odds ratio; ROC, receiver operating characteristic.

Multivariable analysis adjusting for age, mode of delivery, TXA use and uterine atony as the cause of the hemorrhage, revealed fibrinogen measured at the initiation of OB-MTP was still associated with severe hemorrhage [OR (95% CI) = 1.037 (1.009-1.066)] for every 50 mg/dL decrease in fibrinogen measured at OB-MTP activation. Due to the larger than anticipated decrease in the sample size when we restricted the analysis to patients with complete ROTEM measurements, a multivariable analysis was not performed for those laboratory parameters.

The optimal cutoff values (PPV and NPV) for predicting severe obstetric hemorrhage for each of the laboratory parameters that were significantly different between the severe and nonsevere hemorrhage groups were: fibrinogen = 289 mg/dL (0.835 and 0.403), CFT = 62 s (0.811 and 0.647), FIBTEM A10 = 19 mm (0.836 and 0.393), FIBTEM A20 = 17 mm (0.855 and 0.352), EXTEM alpha angle = 72° (0.821 and 0.339), EXTEM A10 = 57mm (0.830 and 0.363), and EXTEM A20 = 65 mm (0.826 and 0.395).

Discussion

Our findings demonstrate that the maternal fibrinogen level measured at the activation of the OB-MTP was independently associated with progression to severe hemorrhage, further confirming its role as a useful biomarker for predicting severe hemorrhage and validating the findings of previous studies, albeit with a smaller effect size. Interestingly, the ROTEM parameters EXTEM CFT, A10, A20 and α-EXTEM and FIBTEM A10 and A20, primarily markers of clot strength and to a lesser extent clot kinetics, were comparable predictors of progression to severe hemorrhage (Table 4). Since these parameters may be reported earlier than fibrinogen results, they too may have a role in predicting the severity of hemorrhage and marshaling resources in a timely manner to successfully manage these high-risk patients. While recent evidence has focused on the use of FIBTEM, our data suggests that the EXTEM, which has a contribution from fibrinogen, may also provide useful parameters associated with progression to severe hemorrhage.

In line with previously published studies, we demonstrated that fibrinogen measurements were lower in patients who progressed to severe hemorrhage.^6,8,11,19,20 In fact, the ROC analysis for the ability of fibrinogen to predict hemorrhage also demonstrated that in our cohort the predictive ability of fibrinogen was similar to those in previously published cohorts despite differences in the definition of severe hemorrhage, lack of standardization of protocol activation, and the heterogenous etiology contributing to the development of hemorrhage.^6,7,11,19,21 While the article by Charbit et al highlighted a fibrinogen level of 200 mg/dL as predictive of progression to severe hemorrhage, Cortet et al demonstrated that fibrinogen levels between 200 to 300 mg/dL were also predictive of progression to severe obstetric hemorrhage.^6,11 In our study, a fibrinogen value of 289 mg/dL and below was associated with a PPV of progressing to severe hemorrhage of 83% with a low NPV of 40%.With median fibrinogen levels at term reported to be 500 mg/dL, fibrinogen levels between 200 and 300 mg/dL still represent a significant deficit. ²² These findings provide further evidence that higher fibrinogen levels in the early stages of a hemorrhage, even though they maybe higher than have been previously reported, may be useful clinical predictors of adverse clinical outcomes.

Another interesting finding from our study was that several ROTEM parameters investigated were also significant predictors of progression to severe hemorrhage. The optimal cutoff points for all 6 parameters had comparable PPV to those of fibrinogen and interestingly the optimal cutoff values for all the ROTEM parameters were below the published median baseline values in peripartum term laboring and immediate postpartum patients.^23,24 However these ROTEM parameters also fell within recently published peripartum reference ranges, highlighting the importance of interpreting these laboratory parameters in a clinical context.^23,24 Our findings demonstrate that, rather than using ROTEM parameters solely as triggers for transfusion, they may also play a role in identifying patients at risk for hemorrhage-related morbidity and can then be used by clinicians in conjunction with other tools to provide a much more targeted approach to obstetric hemorrhage management.

Prior work has focused on the role of FIBTEM since this parameter looks specifically at the contribution of fibrinogen to clot strength, with fibrinogen levels correlating strongly with FIBTEM A5, A10, A15, and maximum clot firmness (MCF) in patients with postpartum hemorrhage.^13,25 In our cohort both the FIBTEM A10 and A20 were significantly lower in the patients with severe hemorrhage and both were predictors of progression to severe hemorrhage in the univariate analysis. This provides further evidence that FIBTEM may be useful in identifying obstetric patients at risk for adverse outcomes after the onset of obstetric hemorrhage. ⁷ Interestingly, we also demonstrated that the CFT, α-EXTEM, EXTEM A10, and A20 were useful predictors and comparable to the FIBTEM A10 and A20 in predicting hemorrhage severity. Both the EXTEM CFT and clot firmness are based on both platelet and fibrinogen contribution and would be impacted by lower levels of fibrinogen. ²⁶ In fact, both α-EXTEM and EXTEM A10 show strong correlation with fibrinogen levels in obstetric patients with postpartum hemorrhage. ¹³ Because the α-EXTEM result is available within 1 min after clot formation and also correlates strongly with the FIBTEM A10, it may be one of the earliest indicators of hypofibrinogenemia, making it potentially clinically relevant. ¹³ The α-EXTEM and A10 are also affected by platelet number but given that profound thrombocytopenia (platelet count <50 × 10⁹/L) is not commonly observed early in severe hemorrhage, ²⁷ and platelet counts measured at the start of a hemorrhage are not predictive of severe hemorrhage, the α-EXTEM could be used as a surrogate marker of hypofibrinogenemia.^6,13,20,28

While several studies have identified hypofibrinogenemia as a predictor for progression to severe hemorrhage, little evidence exists to determine whether early replacement of fibrinogen might improve clinical outcomes. In one randomized controlled clinical study aggressive fibrinogen replacement with fibrinogen concentrate in women with moderate to severe postpartum hemorrhage and associated hypofibrinogenemia (FIBTEM A5 ≤ 15 mm) did not significantly reduce allogeneic blood product transfusion requirements. ²⁹ However, the authors could not rule out a significant treatment effect in patients with a fibrinogen of <250 mg/dL and a FIBTEM A5 ≤ 12 mm at enrollment. ²⁹ In another prospective observational study, a ROTEM-based transfusion algorithm aimed at identifying patients with an acquired hypofibrinogenemia (FIBTEM <12 mm) corrected with fibrinogen concentrate replacement were compared with a formulaic transfusion regime. ³⁰ Patients treated with the ROTEM-based algorithm required significantly less blood products and were at lower risk of developing transfusion associated circulatory overload. ³⁰ These findings suggest that a subset of patients with severe hypofibrinogenemia identified early by ROTEM testing might benefit from fibrinogen replacement therapy which could have an impact on the clinical outcome.

Interestingly we did not observe any significant differences in clot lysis (LI30) between the severe and nonsevere groups and LI30 was not a significant predictor of severe hemorrhage in this cohort. However, the sensitivity of viscoelastic testing in detecting pregnancy-related hyperfibrinolysis or hypofibrinolysis in vitro is debatable.^31,32 In fact, current normal ranges for the LI30 of 94% to 100% make it difficult to differentiate between normal lysis and hyperfibrinolysis during pregnancy, and it is unclear what degree of fibrinolysis in obstetric hemorrhage is associated with adverse clinical outcomes. ³¹ The relative insentivity of viscoelastic testing to detect fibrinolysis may be partly due to an unclear understanding of fibrinolytic mechanisms in obstetric hemorrhage and localization of fibrinolysis to the placental bed.^31,32

Our study has several limitations. Firstly, the timing of initiation of the OB-MTP was not standardized and was based on both loosely defined criteria and the clinical judgment of providers to determine when activation was appropriate. As a result, initiation of the OB-MTP might represent different points during the early stages of a hemorrhage with the OB-MTP activated later in patients with lower fibrinogen levels. Our findings may be confounded by the fact that low fibrinogen levels and lower EXTEM and FIBTEM A10 levels may have influenced the management of the patients. While the OB-MTP primarily described the transfusion of cryoprecipitate and platelets, the transfusion of other blood products and obstetric and surgical management of patients were at the discretion of the clinical team. Due to the overall small study sample size and the use of ROTEM in only a subset of the patients, we were unable to develop a comprehensive clinical model for the prediction of severe hemorrhage for all the laboratory parameters. Our results should also be interpreted in the context that our cohort had a high proportion of patients who experienced a severe hemorrhage, which would indicate that the OB-MTP was not activated as frequently for patients experiencing less severe hemorrhages potentially leading to selection bias. Furthermore, the cohort used for this study was sampled from a tertiary delivery unit with a patient population having a higher prevalence of hemorrhage than the general obstetric population. As a result the PPV is likely increased while the NPV is likely decreased for predicting severe hemorrhage in our population for both fibrinogen and ROTEM and our findings may not be generalizable to obstetric populations with a lower hemorrhage prevalence. Future studies focused on the development of risk prediction models which include both laboratory parameters and clinical risk factors would be of significant benefit to clinicians involved in the management of obstetric hemorrhage.

Conclusion

Fibrinogen and to a lesser extent ROTEM parameters which evaluate primarily clot strength measured at the initiation of obstetric hemorrhage protocol are associated with the development of a severe obstetric hemorrhage. ROTEM parameters may be more rapidly available and therefore allow for earlier identification of at-risk patients for severe hemorrhage-related morbidity and mortality. Earlier risk stratification of patients in a severe obstetric hemorrhage would allow for the development and implementation of targeted transfusion practices, thus decreasing morbidity and mortality in this patient population.