Report of the ISTH registry on pregnancy and COVID-19-associated coagulopathy (COV-PREG-COAG)

Abstract

Background

Concerns about COVID-19-associated coagulopathy (CAC) in pregnant individuals were raised in early pandemic.

Methods

An ISTH-sponsored COVID-19 coagulopathy in pregnancy (COV-PREG-COAG) international registry was developed to describe incidence of coagulopathy, VTE, and anticoagulation in this group.

Results

All pregnant patients with COVID-19 from participating centers were entered, providing 430 pregnancies for the first pandemic wave. Isolated abnormal coagulation parameters were seen in 20%; more often with moderate/severe disease than asymptomatic/mild disease (49% vs 15%; p < 0.0001). No one met the ISTH criteria for disseminated intravascular coagulopathy (DIC), though 5/21 (24%) met the pregnancy DIC score. There was no difference in antepartum hemorrhage (APH) with asymptomatic/mild disease versus moderate/severe disease (3.4% vs 7.7%; p = 0.135). More individuals with moderate/severe disease experienced postpartum hemorrhage (PPH) (22.4% vs 9.3%; p = 0.006). There were no arterial thrombotic events. Only one COVID-associated venous thromboembolism (VTE) was reported.

Conclusions

Low rates of coagulopathy, bleeding, and thrombosis were observed among pregnant people in the first pandemic wave.

Background

COVID-19 infection remains a major public health challenge, with considerable morbidity and mortality, and continuing threat from new variants. ¹ It became apparent early in the pandemic that patients hospitalized with COVID-19 had an increased risk of thromboembolic events.^2–5 Subsequent observations linked these events to a newly defined entity termed COVID-19-associated coagulopathy (CAC). ⁶ CAC comprises microvascular and macrovascular thrombosis, initially within the pulmonary vasculature but with a risk of progression to life-threatening acute lung injury, and multiorgan dysfunction. ⁷ Initially, it was anticipated that due to its inherent hypercoagulable state, with higher levels of fibrinogen, von Willebrand factor, factor VII, and factor VIII and with increased levels of D-dimers, ⁸ pregnancy with concurrent COVID-19 infection would be more susceptible to CAC.^7,9

With a sharp increase in reports of CAC, including suggestion of disseminated intravascular coagulopathy and venous thromboembolism, in the non-pregnant population, and concerning early case reports of CAC in pregnant individuals,^10–13 a communication from the International Society on Thrombosis and Hemostasis (ISTH)'s Scientific Standardization Committee (SSC) for Women's Health outlined preliminary recommendations for management of CAC in pregnancy. ¹⁴ Owing to the paucity of data regarding CAC in this population, establishment of an international registry was recommended. The objective of this report was to leverage data from the ISTH-sponsored, COVID-19 coagulopathy in pregnancy (COV-PREG-COAG) registry, to describe (a) the incidence of coagulopathy, (b) the incidence of VTE, and (c) use of anticoagulation in COVID-19-affected pregnancies.

Methods

Registry questionnaire

The COV-PREG-COAG international registry is administered in the form of an online case report form (CRF), consisting of 569 elements (https://redcap.isth.org/surveys/?s=4JPX9W98RH).

Based on known differences in management for these populations, COVID-19 severity was defined as asymptomatic (SARS-CoV-2 positive, without clinical symptoms), mild (outpatient management), moderate (hospitalization, without critical care), or severe (intensive care unit admission).^15,16

Postpartum hemorrhage was defined as estimated blood loss of greater than 1000 mL, regardless of the mode of delivery. ¹⁷ Preterm birth was defined as birth before 37 weeks’ gestation, and early preterm birth as birth before 34 weeks’ gestation. ¹⁷ Small for gestational age (SGA) was defined as birthweight below 10th centile for gestational age. ¹⁸ Determination of the presence of other obstetric complications, such as gestational hypertension, pre-eclampsia, HELLP, gestational diabetes, cholestasis, and antepartum bleeding (APH), was left up to the individual centers, based on documentation of these conditions in their patient records. Pregnancy- and trimester-specific laboratory parameter ranges were used to determine laboratory abnormalities, ¹⁹ with values below the lower normal range considered as “cytopenias” (e.g., thrombocytopenia, leukopenia, etc.) and those above the upper normal range considered as “cytosis” (e.g., thrombocytosis, leukocytosis, etc.).

Documentation of various biochemical parameters was sought, based on their previously reported prognostic value in COVID-19.^20,21 Because CAC was an emerging concept and lacked an accepted definition at the time of the registry's inception, the survey definition encompassed “any coagulation parameter abnormality, bleeding or thrombotic event.” Questions regarding anticoagulation focused on low-molecular-weight heparin (LMWH), given its role as the preferred agent in pregnancy.^22,23 Registry entries were based on data provided by the respondent and did not involve a formal chart review. The study received ethics approval at each participating institution. In the UK, this included Health Research Authority and Health and Care Research Wales Approval. Data were handled in line with information governance regulations. Ethics approval without requirement for consent was likewise granted in the USA and Canada. Data entry proceeded under a study number with removal of any identifying information such as medical record number, date of birth, or hospital site.

An invitation to participate in the registry, with a web-link for access, was distributed through email via specialist societies, including the ISTH, Canadian Society of Maternal–Fetal Medicine, and British Maternal–Fetal Medicine Society. Social media platforms were also utilized to raise awareness of the registry and encourage participation.

Clinical data for each case were abstracted from relevant clinical charts and hospital records by local registry contributors, for entry directly into a Research Electronic Data Capture (REDCap) database, hosted on the ISTH website. A downloadable form of the questionnaire was also included on the website as an alternative where REDCap entry was not feasible.

Data analysis

Data were downloaded into an SPSS database that included all records and fields contained within the questionnaire, and descriptive statistics were carried out. Frequencies and percentages were used to summarize categorical variables. Depending on normality of distribution, means (standard deviations) or medians (interquartile ranges) were used for continuous variables, as appropriate. Group comparisons for categorical variables were made using chi-squared tests or Fisher's exact tests, as appropriate.

Results

Representatives of several institutions contacted the project leads (AKM, SK, MO) and entered data into the registry, which included 430 pregnancies for the first wave of the pandemic, with variable degree of detail recorded. Consecutive cases of COVID-19 in pregnancy from the participating institutions were captured.

Maternal demographic details are shown in Table 1, as are maternal co-morbidities and risk factors for venous thromboembolism (VTE).

Table 1.

Maternal demographics.

	N*	N† (%)‡	Mean (SD) Median (25–75 IQR; range)
Maternal age (years)	415		30 (26–34; 18–45)
≤30	415	208 (50.1%)
31–35	415	132 (31.8%)
36–40	415	58 (14%)
>40	415	17 (4.1%)
Ethnicity
White/Caucasian	413	176 (42.6%)
African/Caribbean/Black	413	72 (17.4%)
South Asian	413	61 (14.8%)
Hispanic/Latinx	413	55 (13.3%)
Middle Eastern	413	14 (3.4%)
East Asian	413	5 (1.2%)
Southeast Asian	413	1 (0.2%)
Unknown	413	21 (5.1%)
Other	413	8 (1.9%)
Body mass index (kg/m2)	414		27.1 (23.7–32; 16–56)
Underweight (<18.5)	414	6 (1.4%)
Healthy weight (18.5–24.9)	414	118 (28.5%)
Overweight (25.0–29.9)	414	136 (32.9%)
Obese (≥30)	414	154 (37.2%)
Overweight/obese (BMI ≥ 25)	414	290 (70.0%)
Type of conception	411
Spontaneous	411	405 (98.5%)
Assisted (non-IVF)	411	1 (0.2%)
IVF	411	5 (1.2%)
Multipara	406	248 (61.1%)
Maternal underlying conditions
Any underlying conditions§	385	299 (77.7%)
Chronic hypertension	241	6 (2.5%)
Diabetes mellitus	242	4 (1.7%)
Cardiovascular disease	241	1 (0.4%)
Antiphospholipid syndrome	242	1 (0.4%)
Respiratory disease	240	16 (6.7%)
Asthma	16	14 (88%)
Gastrointestinal disease	241	8 (3.3%)
Immunological disorder	162	2 (1.2%)
VTE risk factors aside from pregnancy
Immobility	141	8 (5.7%)
Surgery	158	51 (32.3%)
Overweight/obesity	414	290 (70.0%)
Thrombophilia¶	139	4 (2.9%)

* “N” refers to the total number of responses received for a particular field (denominator).

^† “n” refers to the indication of the presence of a particular finding (nominator).

^‡ “%” refers to the proportion of each finding as a function of the total responses.

^§ Overweight/obesity is included in maternal underlying conditions; however, it is reported in the table under VTE risk factors.

^¶ Factor V Leiden (FVL) (n = 2—heterozygous in 1; unspecified in 1); antiphospholipid antibodies (n = 1); unspecified (n = 1).

Table 2 presents characteristics of COVID-19 infection, including details of testing, symptomatology, and disease course. Diagnosis was most frequent in the third trimester (321, 75%), and the majority of cases (255, 60%) were asymptomatic. Hospitalization was required in 53 (31%) of those who were symptomatic, with intensive care unit (ICU) admission in 10 (19%) of hospitalized individuals. Oxygenation via high-flow nasal cannula was necessary in seven (2.1%) individuals, non-invasive mechanical ventilation in six (1.8%), and invasive mechanical ventilation (IMV) in six (1.8%). No one received extracorporeal membrane oxygenation (ECMO), and there were no maternal deaths.

Table 2.

Clinical characteristics of COVID-19 infection course.

	N	n (%)	Mean (SD) Median (25–75 IQR; range)
Timing of COVID-19 diagnosis
First trimester	429	15 (3.5%)
Second trimester	429	50 (11.7%)
Third trimester	429	321 (74.8%)
Postpartum	429	43 (10%)
Gestational age at diagnosis (wks)	368		37.4 (31.0–39.4; 4.4–42.3)
Method of testing
Naso-pharyngeal swab	427	408 (95.6%)
Throat swab	265	178 (67.2%)
Serology—IgM	364	27 (7.1%)
Serology—IgG	349	8 (2.3%)
Symptomatic
Any symptoms	425	170 (40.0%)
Cough	146	92 (63.0%)
Fever	144	89 (61.8%)
Shortness of breath	131	41 (31.3%)
Anosmia	131	31 (23.7%)
Pneumonia	125	28 (22.4%)
Fatigue/malaise	131	26 (19.8%)
Headache	131	21 (16.0%)
Flu-like illness/myalgia	131	13 (9.9%)
Chest pain/tightness	131	12 (9.2%)
Sore throat	131	11 (8.4%)
Nasal congestion/coryza	131	9 (6.9%)
Tachycardia	131	9 (6.9%)
Diarrhea	131	7 (5.3%)
Arthralgia	131	5 (3.8%)
Tachypnea	131	3 (2.3%)
Illness severity overall
Asymptomatic	429	255 (59.3%)
Mild (common COVID-19 symptoms)	429	121 (28.1%)
Moderate (hypoxia, hospitalization)	429	43 (10.0%)
Severe (ICU admission)	429	10 (2.3%)
Illness severity in those with symptoms
Mild (common COVID-19 symptoms)	174	121 (69.5%)
Moderate (hypoxia, hospitalization)	174	43 (24.7%)
Severe (ICU admission)	174	10 (5.7%)
Hospitalization if symptomatic	174	53 (30.5%)
Hospitalization LOS (days)	30		4 (3–10; 1–74)
ICU admission if hospitalized	53	10 (18.9%)
ICU LOS (days)	5		11 (3.5–38.5; 1–65)
Maternal death	162	0
Oxygen support
High-flow nasal cannula	339	7 (2.1%)
Non-invasive mechanical ventilation	337	6 (1.8%)
Invasive mechanical ventilation	338	6 (1.8%)
Extracorporeal membrane oxygenation	336	0

ICU: intensive care unit; LOS: length of stay.

Details of general hematological and coagulation-specific parameters are summarized in Table 3. There were no illness severity-based differences for most of the general laboratory parameters. Conversely, significant differences for asymptomatic/mild illness and moderate/severe illness were observed for leukopenia and hyperferritinemia, both of which were higher in the moderate/severe group. Hypoferritinemia was observed more frequently in the asymptomatic/mild illness group. Isolated abnormal coagulation parameters were seen in 70 (20%), according to pregnancy- and trimester-specific values. ¹⁹ These included thrombocytopenia (13%), thrombocytosis (3%), prolonged APTT (16%), hypofibrinogenemia (30%), hyperfibrinogenemia (13%), and elevated D-dimer (67%). More individuals with moderate/severe disease experienced isolated coagulation abnormalities in comparison to those with asymptomatic/mild disease (49% vs 15%; p < 0.0001). Prolonged APTT was noted only in individuals with moderate/severe disease. None of the participants met the criteria for disseminated intravascular coagulopathy (DIC) per the ISTH definition, ²⁴ though 5/21 (24%) of those for whom all hematological parameters were available met the pregnancy DIC score. ²⁵

Table 3.

Pregnancy- and trimester-specific laboratory parameters ¹⁹ according to severity of COVID-19 infection.

	Overall severity			Asymptomatic/Mild		Moderate/Severe
	Mean (SD) Median (25–75 IQR; range)	N*	n† (%)‡	N	n (%)	N	n (%)	p-Value§
General hematological parameters
WBC (×10(9)/L)	9.3 (7.1–11.2; 2.1–25.1)	200
Leukopenia (<5.7 in T1; <5.6 in T2; <5.9 in T3/PP)		200	20 (10%)	160	12 (7.5%)	40	8 (20%)	0.034
Leukocytosis (>13.6 in T1; >14.8 in T2; >16.9 in T3/PP)		200	15 (7.5%)	160	13 (8%)	40	2 (5%)	0.740
Neutrophils (×10(9)/L)	6.5 (5.4–8.6; 1.0–17.2)	67
Neutropenia (<3.6 in T1; <3.8 in T2; <3.9 in T3/PP)		67	3 (4.5%)	52	1 (1.9%)	15	2 (13.3%)	0.123
Neutrophilia (>10.1 in T1; >12.3 in T2; >13.1 in T3/PP)		67	8 (11.9%)	52	7 (13.5%)	15	1 (6.7%)	0.672
Lymphocytes (×10(9)/L)	1.5 (0.6)	90
Lymphopenia (<1.1 in T1; <0.9 in T2; <1.0 in T3/PP)		90	19 (21.1%)	63	11 (17.5%)	27	8 (29.6%)	0.195
Hematocrit (%)	36 (33.7–39; 22.4–88.0)	144
C-reactive protein (CRP) (mg/L)	28.5 (12.5–74.0; 2–256)	66
Elevated CRP (>20.3 in T1/T2; >8.1 in T3/PP)		66	55 (83.3%)	35	8 (22.9%)	30	3 (10%)	0.201
Ferritin (µg/L)	27 (13.5–91; 8–865)	37
Hypoferritinemia (<50¶)		37	21 (56.8%)	25	20 (80%)	12	1 (8.3%)	<0.01
Hyperferritinemia (>130 in T1; >230 in T2; >116 in T3/PP)		37	6 (16.2%)	25	0	12	6 (50%)	<0.001
Albumin (g/L)	28.5 (5.4)	21
Hypoalbuminemia (<31 in T1; <26 in T2; <23 in T3/PP)		21	3 (14.3%)	9	1 (11.1%)	12	2 (16.7%)	1.000
Aspartame transaminase (AST) (IU/L)	42.2 (23.6)	13
Elevated AST (>23 in T1; >33 in T2; >32 in T3)		13	7 (53.8%)	2	2 (100%)	11	5 (45.5%)	0.462
Alanine transferase (ALT) (IU/L)	26 (18–40; 8–123)	47
Elevated ALT (>30 in T1; >33 in T2; >25 in T3)		47	22 (46.8%)	29	14 (48.3%)	18	8 (44.4%)	0.798
Lactate dehydrogenase (LDH) (IU/L)	285 (123)	11
Creatinine (µmol/L)	49 (41–59; 23–113)	58
Elevated creatinine (>62 in T1; >71 in T2; >80 in T3/PP)		58	4 (6.9%)	38	3 (7.9%)	20	1 (5%)	1.000
Abnormal urine P:C ratio (g/mol) (>30 g/mol)		5	4 (80%)	0	—	5	4 (80%)	—
Coagulation parameters
Isolated abnormal coagulation Parameters		354	70 (19.8%)	306	46 (15%)	47	23 (48.9%)	<0.001
Hemoglobin (g/L)	118 (13.9)	201
Anemia (<110 g/L)		201	47 (23.4%)	161	34 (21.1%)	40	13 (32.5%)	0.128
Platelet count (×10(9)/L)	231 (177–276; 56–570)	352
Thrombocytopenia (<150)		352	45 (12.8%)	304	37 (12.2%)	47	8 (17%)	0.355
Thrombocytosis (>391 in T1; >409 in T2; >429 in T3/PP)		352	10 (2.8%)	304	7 (2.3%)	47	3 (6.4%)	0.137
Prothrombin time (sec)	10.4 (10.0–11.0; 9.0–12.5)	59
Prolonged PT (>13.5 in T1; >13.4 in T2; and >12.9 in T3/PP)		59	0	—	—	—	—	—
Activated partial thromboplastin time (sec)	30.6 (5.2)	55
Prolonged APTT (>38.9 in T1; >38.1 in T2; >35.0 in T3/PP)		55	9 (16.4%)	27	0 –	28	9 (32.1%)	0.002
Fibrinogen (g/L)	4.3 (1.3)	23
Hypofibrinogenemia (<2.4 in T1; 2.9 in T2; 3.7 in T3/PP)		23	7 (30.4%)	7	1 (14.3%)	16	6 (37.5%)	0.366
Hyperfibrinogenemia (>5.1 in T1; 5.4 in T2; 6.2 in T3/PP)		23	3 (13%)	7	1 (14.3%)	16	2 (12.5%)	1.000
D-dimer (µg/mL)	7.2 (3.6–13.9; 0.7–35.2)	21
Elevated D-dimer (above ULN for each institution)		21	14 (66.7%)	4	3 (75%)	16	10 (62.5%)	1.000
DIC - ISTH criteria		12	0	—	—	—	—	—
DIC - Pregnancy score (25)		21	5 (23.8%)**	6	1 (16.7%)	15	4 (26.7%)	1.000

* “N” refers to the total number of responses received for a particular field (denominator).

^† “n” refers to the indication of the presence of a particular finding (nominator).

^‡ “%” refers to the proportion of each finding as a function of the total responses.

^§ Chi-square test reported in regular font; Fisher's exact test reported in italics.

^¶ Hypoferritinemia definition according to Tang et al. ⁵⁶

**4.9% of total sample.

DIC: disseminated intravascular coagulation; ISTH: International Society on Thrombosis and Hemostasis; PP: postpartum; T1: first trimester; T2: second trimester; T3: third trimester.

Data on bleeding and thrombosis is presented in Table 4. There was no difference in APH between those with asymptomatic/mild disease and those with moderate/severe disease (3.4% vs 7.7%; p = 0.135). Conversely, more individuals with moderate/severe disease experienced PPH compared to those with asymptomatic/mild disease (22.4% vs 9.3%; p = 0.006), with more urgent or emergent Caesarean deliveries in the moderate/severe group (44% vs 18.8%; p < 0.001). There were no reported arterial thrombotic events. Only one VTE was observed. This was in the form of a pulmonary embolism on post-operative day 5 after an urgent Caesarean delivery, despite thromboprophylaxis with LMWH.

Table 4.

Pregnancy outcomes.

	Overall cohort			Asymptomatic/mild		Moderate/severe
	Mean (SD) Median (25–75 IQR; range)	N*	n† (%)‡	N	n (%)	N	n (%)	p-Value§
Pregnancy complications
Gestational hypertension		407	34 (8.4%)	354	31 (8.8%)	52	3 (5.8%)	0.599
Pre-eclampsia		409	33 (8.1%)	356	32 (9%)	52	1 (1.9%)	0.101
HELLP syndrome		409	3 (0.7%)	357	3 (0.8%)	51	0 –	1.000
Gestational diabetes		407	35 (8.6%)	354	30 (8.5%)	52	5 (9.6%)	0.791
Cholestasis		406	14 (3.4%)	353	11 (3.1%)	52	3 (5.8%)	0.404
PPROM		145	7 (4.8%)	116	7 (6%)	29	0 –	0.345
Thrombotic or bleeding complications
Any thrombotic or bleeding event		410	63 (15.4%)	357	47 (13.2%)	52	15 (28.8%)	0.003
APH ≥14 wks		407	17 (4.2%)	354	12 (3.4%)	52	4 (7.7%)	0.135
GA at bleeding event ≥14 wks	35.6 (29.0–38.1; 21.9–38.6)	407	17 (4.2%)
PPH		410	44 (10.7%)	354	33 (9.3%)	49	11 (22.4%)	0.006
APH ≥14 wks + PPH		410	8 (2.0%)	358	38 (10.6%)	52	14 (26.9%)	<0.001
Venous thromboembolism¶		338	1 (0.3%)	294	0 –	43	1 (2.3%)	0.128
Arterial thrombosis		130	0	—	—	—	—	—
Thromboprophylaxis
VTE prophylaxis (overall)		418	273 (65.3%)	363	232 (63.9%)	52	46 (88.5%)	<0.001
LMWH** Duration (days)	10 (10–10; 1–133)	279	270 (96.8%)	232	223 (96.1%)	46	46 (100%)	0.364
Obstetric outcomes
Location of birth
Hospital		411	406 (98.8%)	359	355 (98.9%)	51	50 (98%)	0.487
Birthing center		411	4 (1.0%)	359	3 (0.8%)	51	1 (2.0%)	0.413
Home		411	1 (0.2%)	359	1 (0.3%)	51	0 –	1.000
Type of birth
Spontaneous vaginal delivery		406	225 (55.4%)	355	203 (57.2%)	50	21 (42%)	0.043
Assisted vaginal delivery		406	37 (9.1%)	355	34 (9.6%)	50	3 (6%)	0.600
Elective Caesarean delivery		406	55 (13.5%)	355	51 (14.4%)	50	4 (8%)	0.219
Urgent Caesarean delivery		406	61 (15%)	355	97 (27.3%)	50	19 (38%)	0.118
Emergent Caesarean delivery		406	28 (6.9%)	355	72 (20.3%)	50	11 (22.2%)	0.778
COVID-19 as sole indication for Caesarean delivery		116	13 (11.2%)	93	2 (2.2%)	23	11 (47.8%)	<0.001
Birth outcome
Livebirth		410	400 (97.6%)	358	350 (97.8%)	51	50 (98%)	1.000
Miscarriage (loss <20 w or <500 g)		410	3 (0.7%)	358	3 (0.8%)	51	0 –	1.000
Stillbirth (loss ≥20 w or ≥500 g)††		410	6 (1.5%)	358	5 (1.4%)	51	0 –	1.000
Preterm birth <34 weeks		146	14 (9.5%)	116	12 (10.3%)	30	1 (3.3%)	0.305
Preterm birth <37 weeks		146	39 (26.7%)	116	30 (25.9%)	30	9 (30%)	0.648
Gestational age at birth (weeks)	39.3 (38.1–40.1; 19.7–42.4)	405
Small for gestational age		393	52 (13.2%)	344	45 (13.1%)	48	7 (14.6%)	0.774
Birthweight (g)	3275 (2940–3556; 200–4585)	395

* “N” refers to the total number of responses received for a particular field (denominator); the sum of N for “asymptomatic/mild” and N for “moderate/severe” may not total N for overall cohort, as severity of illness was not specified in several cases.

^† “n” refers to the indication of the presence of a particular finding (nominator).

^‡ “%” refers to the proportion of each finding as a function of the total responses.

^§ Chi-square test reported in regular font; Fisher's exact test reported in italics.

^¶ This was a COVID-19-associated PE on POD#5 after an urgent C-section; several VTEs were also reported, pre-dating COVID-19 diagnosis: one PE in the first trimester, pre-dating COVID-19 diagnosed postpartum; one upper extremity DVT pre-dated third trimester COVID-19 diagnosis; one further note was made of “thrombotic at 7 weeks,” but VTE was coded as no in context of mild illness diagnosed in the third trimester; one further DVT was suspected, but ultrasound compression venography was negative for DVT and the patient received only LMWH prophylaxis.

**Details regarding LMWH thromboprophylaxis were provided for 84 cases, as follows: enoxaparin 40 mg ×67; enoxaparin 60 mg ×6 (for weight between 90 and 110 kg); enoxaparin 80 mg ×3 (for weight of 97 kg on one case and unspecified in two others); enoxaparin unspecified dose ×5; tinzaparin 4500 IU ×1; dalteparin 5000 IU ×1; dalteparin 2500 IU ×1 (for a weight of 42 kg).

^†† One stillbirth resulted from TOP for cardiac anomaly.

The incidence of PPH did not differ between those who did and those who did not meet the criteria (1/4 with PPH vs 3/4 without PPH, p = 1.000), although the numbers are likely too small to appreciate a difference. The sole person who experienced a VTE did not meet the criteria for the pregnancy DIC score.

Thromboprophylaxis was provided more frequently in those with moderate/severe illness than with asymptomatic/mild illness (64% vs 89%; p < 0.001). Postpartum thromboprophylaxis was recorded in 90% of cases, with data available for 27% of the sample. Details of agent and dose were available in 84 cases, with standard prophylactic doses in the majority, and select cases of weight-based thromboprophylaxis (enoxaparin 40 mg in 67 cases, enoxaparin 60 mg in six cases (weight 90–110 kg), enoxaparin 80 mg in three cases (weight 97 kg in one and unspecified in two others), enoxaparin at unspecified dose in five cases, and in one case each tinzaparin 4500 IU, dalteparin 5000 IU, and dalteparin 2500 IU (weight 42 kg)).

Data on pregnancy complications and obstetric outcomes are shown in Table 4. There were no differences between illness severity groups for the majority of pregnancy complications and obstetric outcomes. Vaginal delivery was achieved more often in those with asymptomatic/mild illness compared with moderate/severe illness. COVID-19 as a sole indication for Caesarean delivery was more frequent in those with moderate/severe illness than in those with asymptomatic/mild illness (47.8% vs 2.2%; p < 0.001).

Conclusions

Principal findings

This report of the ISTH-sponsored COV-PREG-COAG international registry includes data from 430 individuals in the first wave of COVID-19 infection. Moderate or severe COVID-19 disease was recorded in 30% of individuals. Isolated abnormal coagulation parameters were seen in 20%, with a higher number of individuals with moderate or severe disease experiencing isolated abnormal coagulation parameters in comparison to those who were asymptomatic or had mild disease. None of the participants met the criteria for DIC per the ISTH, though 24% of those for whom all hematological parameters were available met the pregnancy DIC score. There was no difference in APH between those who were asymptomatic or had mild disease and those with moderate or severe disease; however, a higher number of individuals with moderate or severe disease experienced PPH in comparison to those who were asymptomatic or had mild disease. There were no reported arterial thrombotic events. Only one COVID-associated VTE was reported. Thromboprophylaxis was provided more frequently in those with moderate or severe illness than in those who were asymptomatic or had mild illness.

Results in context of literature

Clinical outcomes

Consistent with reported literature, our findings suggest a higher risk of COVID-19 in those of non-Caucasian ethnicity, multiparous patients, and those with underlying medical conditions, most notably obesity. ²⁶ Likewise, in keeping with published reports is the mild nature of the illness in the majority^27,28 and predominance of cough, fever, and shortness of breath in those who were symptomatic, as in the non-pregnant population. ²⁹

Of those who were symptomatic, 31% were hospitalized, and of those hospitalized, 19% required ICU admission. Comparison to the initial reports from China concerning the general population demonstrates slightly higher rates of hospitalization in our symptomatic pregnant patients (20% vs 31%) and slightly higher ICU admission in those who were hospitalized (5–15% vs 19%). ²⁹

COVID-19 diagnosis was most often made in the third trimester (75%) at a median gestational age of 37.4 weeks; potentially explained by universal screening adopted by countries like the USA and the UK for any individual attending hospital for any reason, including labor and birth,^30,31 although the practice of universal screening was more variable in Canadian centers, despite similar timing of diagnosis. ³²

General hematological and coagulation parameters

Anemia (hemoglobin below 100 g/L) was observed in 23% of our sample, compared to 5% observed with COVID-19 outside of pregnancy. ³² The higher rate of anemia in pregnancy may be partially related to the physiologic hemodilution of pregnancy, alongside a high prevalence of iron deficiency.^19,33

In our sample, elevated CRP was seen in 83%, elevated AST in 54%, elevated ALT in 47%, and high creatinine in 7%. Abnormalities in these parameters have been shown to be predictive of severe illness, hospitalization, or worse outcomes in the non-pregnant population.^34,35,36,37

Conversely, hyperferritinemia was seen in only 16% of pregnant patients, while hypoferritinemia was noted in 57%. Whereas hyperferritinemia in the non-pregnant population with COVID-19 has been associated with illness severity, its relative infrequency in this pregnant population may be explained by the prevalence of iron deficiency in this cohort. ³³ Furthermore, the relatively higher rate of hypoferritinemia in those with asymptomatic/mild illness and relatively higher rate of hyperferritinemia with moderate/severe disease are likely attributable to ferritin's role as an acute-phase reactant.

As with the non-pregnant population, ³⁸ we noted approximately half of those with moderate/severe disease exhibited an abnormality in at least one coagulation parameter (vs 15% with mild disease). Gestational thrombocytopenia can accompany 7–12% of otherwise uncomplicated pregnancies; ³⁹ thus we are unable to comment whether thrombocytopenia observed herein was directly related to COVID-19. Reassuringly, thrombocytopenia was mild, with only two cases of severe thrombocytopenia, one in moderate/severe disease, and one owing to chronic liver disease.

DIC involves a dysfunctional activation of the coagulation cascade, which can be incited by a number of catalysts including injury, sepsis, or pregnancy-related events such as placental abruption, stillbirth, HELLP syndrome, or obstetric hemorrhage. ⁴⁰ In 2001, the ISTH proposed a definition of DIC, which incorporated assessment of the platelet count, fibrin-related markers (D-dimer or fibrin degradation products), PT, and fibrinogen; the committee further acknowledged that DIC can be overt or non-overt, with the diagnosis of the latter hampered by the insensitivity of these global coagulation parameters. ⁴¹ Using the ISTH definition, there were no cases of DIC in our study. However, the physiologic changes of pregnancy leading to an increase in pro-coagulant activity, designed to anticipate the potential peripartum blood loss, render the application of the ISTH DIC definition problematic, as it reflects the non-pregnant state. ⁴⁰ To account for these physiologic adaptations, the pregnancy DIC score was developed and validated. ²⁵ It includes modified criteria for the platelet count, fibrinogen, as well as the assessment of the difference in PT (calculated from the patient's result as compared to the average control for a given laboratory), while it excludes D-dimer, the values of which are typically increased in pregnancy. ⁴⁰ Given these stipulations, we calculated the pregnancy DIC score, which met the criteria in 24% of patients in our study for whom all hematological parameters were available. ²⁵ However, the incidence of PPH did not differ between those who did and those who did not meet the criteria for the pregnancy DIC score. The sole person who experienced a VTE likewise did not meet criteria for the pregnancy DIC score. Given the small number of patients for whom all necessary parameters were available to calculate the pregnancy DIC score, its association with clinical bleeding or thrombotic outcomes remains unclear and requires further study.

It is worth noting that the pattern of anemia, low-platelet counts, and elevated liver enzymes encountered in non-pregnant patients with COVID-19, in pregnancy may be mistaken for HELLP syndrome, a condition warranting delivery to achieve resolution. ⁴² However, while delivery is the cure for HELLP syndrome, it will not alter the course of COVID-19 and may add unnecessary risks: (a) to the mother through exposure to the physiologic and potentially surgical stresses of birth when medically unwell and (b) to the infant if iatrogenically premature.^13,43,44 To add complexity, it is worth noting the increased risk ratio for pre-eclampsia in nulliparous individuals with COVID-19 in pregnancy (RR 1.89; 95% CI 1.17–3.05). ⁴⁵ Careful differentiation of whether the clinical and laboratory alterations represent a hypertensive disorder of pregnancy or whether they are purely reflective of the underlying COVID-19 is vital and will help guide the most appropriate management. The following parameters can be helpful in considering pre-eclampsia/HELLP as the inciting event: elevated blood pressure (if present), inappropriate adaptation of uterine artery blood flow, and elevated soluble fms-like tyrosine kinase-1 or decreased placental growth factor where their measurements are feasible. ⁴⁴

Bleeding and thrombosis

Bleeding complications were noted in 15% of our sample, with APH after the first trimester in 4% and PPH in 11%. While disease severity did not influence APH rates, individuals with moderate/severe disease had higher rates of PPH, observed in nearly 25%. The higher incidence of PPH in those with more severe illness may have been influenced by the higher rate of urgent or emergent Caesarean delivery within this group.

Non-pregnant patients hospitalized with COVID-19 have been shown to have a higher risk of thromboembolic events,^2,3,4,5 later linked with the presence of CAC. ⁶ With these observations, there was a concern that the additional hypercoagulable state of pregnancy will potentiate thrombotic risk. Interestingly, only one COVID-associated VTE was reported in our first-wave sample, which occurred despite LMWH thromboprophylaxis. This lower-than-expected incidence of thrombosis in the pregnant population may have been mitigated by thromboprophylaxis in two-thirds of the sample for whom data was available and in 90% postpartum. This perhaps underscores the importance of continuing this practice, which has now been supported by a number of studies outside of pregnancy,^46,47,48 although it does require further scrutiny in the pregnant population.

Pregnancy complications and obstetric outcomes

Pregnancy has now been reported as an independent risk factor for adverse outcomes in individuals with COVID-19 infection, especially with co-existence of co-morbidities, such as elevated BMI, diabetes, or hypertension. ²⁶ However, even in the absence of co-morbidities, COVID-19 infection itself has been established as a risk factor for higher rates of adverse pregnancy outcomes, including increased risk of thromboembolic disease, hypertensive disorders, preterm birth, SGA size, and Caesarean delivery.^49,50,51,52 Our data is similar to that of the Canadian national registry showing livebirths in 1774 (97.4%), stillbirths in 19 (1.0%), prematurity in 228 (12.9%), and SGA size in 127 (10.8%). ⁵² Thus, ongoing fetal and maternal surveillance is warranted in COVID-recovered pregnant people.

In contrast to earlier reports, vaginal delivery was achieved in most pregnancies, with COVID-19 as a sole indication for Caesarean delivery noted in 11%. It is likely that the higher Caesarean delivery rates in earlier reports represented a cautious approach reflecting a novel pathogen for which outcomes were unknown. With evolving familiarity, it became apparent that vaginal delivery did not confer higher rates of adverse maternal or fetal outcomes than Caesarean delivery, removing COVID-19 infection as a sole indication for Caesarean delivery.^53,54

Clinical implications

Despite isolated coagulation abnormalities, the thrombotic nature of COVID-19, and thrombophilic physiologic state of pregnancy, the rates of thrombosis and bleeding among pregnant people from the first wave of the pandemic are low, even in the context of inconsistent thromboprophylaxis.

Research implications

Our report includes pregnancies from the first wave of the pandemic. It is possible that findings from pregnancies complicated by subsequent variants of SARS-CoV2 will differ and comparisons based on data from later waves of the pandemic will be informative.

Strengths and limitations

Our study adds to the sparse reports describing the state of coagulopathy in pregnancies affected by COVID-19. Our definitions of severity reflect the state of knowledge at the time of the first wave of the pandemic, as such severity of illness may be under-represented, as space and resource constraints led to care of sicker patients on lower-acuity wards as the pandemic advanced. Our findings are derived from a retrospective international registry and as such are vulnerable to the limitations inherent in this type of research design, ⁵⁵ which include lack of random allocation, potentially less robust patient follow-up, and selection bias introduced by missing or incomplete data and by the fact that blood draws are more likely to be completed in sicker individuals. While efforts have been made to include consecutive patients from each participating institution, untested individuals with asymptomatic or mild illness may have been missed. Within those limitations, consideration must be given to the fact that in some conditions, contexts, or populations, studies employing higher-order research designs may not be feasible and that in such circumstances registries serves a valuable function ⁵⁵ by gathering much-needed data to fill critical knowledge gaps that would otherwise remain unexplored.

Interpretation

Our findings suggest relatively low rates of severe COVID-19 infection and hematological complications such as thrombosis and bleeding among pregnant people. However, we have only included pregnancies from the first wave of the COVID-19 pandemic, and therefore it is possible that pregnancies complicated by subsequent variants of SARS-CoV2 will differ in this regard. The use and duration of thromboprophylaxis on admission and after discharge was inconsistent, though with studies in non-pregnant individuals demonstrating improved outcomes with the use of anticoagulation, there may have been a change in practice in the later waves of the pandemic, which would be worth exploring.