Maternal mortality due to cardiovascular disease in the Netherlands: a 21-year experience

Twenty-seven women died due to unexplained or arrhythmic sudden death, SUDS (n = 19) or SADS (n = 8).

Of the women who died due to SADS, two were known to have cardiac rhythm disorders pre-pregnancy. In both women contributing care factors were identified. One woman with a long QT syndrome had discontinued her beta-blocker therapy against medical advice. The other woman was known to have Wolf-Parkinson-White syndrome. She had been arrhythmia-free without medication for 6 years. There was a delay in referral to a cardiologist when she complained about palpitations during pregnancy. The other six women who died due to SADS were either not diagnosed or not known to have been diagnosed with arrhythmias pre-pregnancy.

Autopsy was definitely not performed in 12 of 19 women with SUDS and not performed or not reported to the MMMC in another four cases. Five women had cardiac or non-cardiac underlying diseases that may possibly but not definitely have been related to the sudden death (myocardial infarction and heart failure, aortic valve surgery, left ventricular hypertrophy and anomalous left coronary artery, hyperthyroid disease, hyperemesis with dehydration).

Time of death was during pregnancy in eight and during the postpartum period in 16 women. Maternal death accompanied intrauterine and early foetal death in seven cases. Delay in hospital referral was reported in one woman. Information about possible contributing suboptimal care factors was missing in 22 cases.

Twenty women (21%) died from aortic dissection. Most dissections occurred in the third trimester (40%) and postpartum (35%). Pre-existing relevant morbidity was present in eight women (40%): three had a connective tissue disease (18% compared to 0% in women who died due to other CVD, p < 0.01) and five had pre-pregnancy hypertension (25%). Hypertensive obstetric complications occurred in four women (pregnancy-induced hypertension (n = 3) and pre-eclampsia (n = 1)). Overall, seven women had hypertension, pre-existent or pregnancy-induced (41% in women with aortic dissection vs 19% in women who died due to other CVD, p = 0.052). Maternal death accompanied perinatal death in 11 cases. In eight women (40%) suboptimal care factors were identified. In seven cases these involved the presence of signs and symptoms of aortic dissection (including chest or back pain, dyspnoea and circulatory failure) without this leading to proper diagnostic tests and correct diagnosis. Two of these women were misdiagnosed with a psychiatric disorder. In one case the diagnosis was rejected after false-negative sonography.

Seventeen women (18%) died from ischaemic heart disease (IHD). None of these women were known to have coronary artery disease. In three women the diagnosis of IHD was made at autopsy; two of them presented with heart failure. Fourteen women (13% of the total number of maternal deaths) presented with acute myocardial infarction (AMI) during pregnancy. All but one woman presented during the third trimester or postpartum period (1 case of missing data). AMI was caused by coronary dissection (n = 4), coronary sclerosis/thrombosis (n = 2) or other/unspecified causes. Risk factors for IHD were reported in four women (33%). Obstetric events included postpartum haemorrhage (n = 3, not related to maternal death) and CS because of pre-eclampsia (n = 2). Maternal death accompanied neonatal death (n = 5) in four cases. In five cases underestimation of the complaints was reported (four times by the doctor, once by the patient).

Thirteen women (14%) died from cardiomyopathy. In 43% potentially avoidable contributing care factors were identified.

One woman died from CHD and one from pulmonary hypertension. A woman with a Fontan circulation died due to cardiovascular collapse postpartum after a CS for a maternal non-cardiac indication. She had been advised against pregnancy.

A 28-year-old primiparous woman with systemic lupus erythematosus died 3 months postpartum due to heart failure caused by previously unrecognised pulmonary hypertension.

As in other recent reports, the largest group of contemporary maternal cardiac deaths were unexplained sudden deaths (SUDS and SADS) [1]. Whether or not these deaths are definitely from cardiac causes is therefore not sure. In our study autopsy was frequently not performed in those women, which is unfavourable with regard to medical knowledge and science and may in some cases even be legally disputable in the Netherlands. Failure to perform autopsy was not considered by the MMMC to represent substandard care during the period studied, since it did not influence maternal death. However, it may contribute to suboptimal care for the families of these deceased mothers because positive findings at autopsy facilitate family screening and may eventually contribute to preventing further deaths in these families. Four women with SUDS had conditions which may have induced cardiac arrhythmias. Women with known cardiac rhythm disorders should be under cardiological supervision before and during pregnancy. In some women, a switch in anti-arrhythmic medication to avoid foetal risk associated with some anti-arrhythmic medication may be necessary [18]. When a woman with a cardiac rhythm disorder experiences palpitations, dizziness, dyspnoea or chest pain during pregnancy, she should be referred to a cardiologist without delay for further evaluation.

Aortic dissection during pregnancy is rare but has a high case fatality rate of 83% [19]. The diagnosis can be challenging and it should be considered as a possible cause of thoracic pain, especially when risk factors for aortic dissection (e.g. connective tissue disease or hypertension) are present, as in 50% of our population [20]. Despite the presence of symptoms indicative of aortic dissection the diagnosis has frequently not been made [21]. Interestingly, two women were misdiagnosed with a neuropsychiatric disorder, which has been described as a pitfall in the presentation of aortic dissection [22‐24]. Sonography alone is insufficient to exclude aortic dissection according to current guidelines, as is illustrated by one of our cases [21]. We recommend that aortic dissection should be considered in every pregnant woman with risk factors for aortic dissection and unexplained chest pain, back pain, also when combined with unexplained neuropsychiatric symptoms. Electrocardiography (ECG)-gated computed tomography of the chest is indicated when there is a reasonable suspicion of aortic dissection.

IHD was the second most frequent cause of maternal cardiac death, in line with other reports [1]. Pregnancy is known to increase the risk of manifestations of IHD 3‑ to 4‑fold [25]. Acute coronary syndromes are known to be more prevalent during late pregnancy [26]. The maternal mortality ratio is 3–11% and maternal morbidity, foetal mortality and morbidity rates are also high [19, 26, 27]. Consistent with other reports we found coronary dissection to be a frequent cause of IHD [26]. Risk factors for coronary artery disease are often present, especially in women with atherosclerotic disease [26]. Proper recognition of the diagnosis appeared difficult and the severity of complaints was often underestimated. In pregnant women with chest pain it is not adequate only to rule out the differential diagnosis of pulmonary embolism and aortic dissection. IHD should be ruled out by ECG and biomarker tests. It is important to realise that IHD can also have an atypical presentation with symptoms such as syncope or dyspnoea.

A recent national study found a high incidence of cardiovascular morbidity due to cardiomyopathy during pregnancy but with a relatively low case fatality rate [19]. In our study cardiomyopathy was the third most frequent cause of maternal cardiac death. Four women died due to peripartum cardiomyopathy, which is known for its high mortality risk up to 30% [28]. Care factors related to the adverse outcome were mostly patient-related. Interestingly, one woman died from Takotsubo cardiomyopathy, which is a rare diagnosis during pregnancy. It is a stress-related cardiomyopathy presenting with severe heart failure and chest pain, with typical wall motion abnormalities of the apical region (apical ballooning). During pregnancy it has been mainly related to CS or pre-eclampsia. However, in our patient the Takotsubo cardiomyopathy occurred after a difficult vaginal delivery [29]. One woman died due to hypertrophic cardiomyopathy, a condition usually well tolerated during pregnancy with low mortality rates (0.5%) [30]. The number of patients with myocarditis is remarkable. In two cases the differential diagnosis of peripartum cardiomyopathy was not definitely ruled out.

Half of the women who were not known to have (non-pregnancy-related) cardiomyopathy before pregnancy had a positive family history for cardiomyopathy or early sudden death. These are the women who may benefit from pre-pregnancy screening and cardiac evaluation including echocardiography. Physicians should always consider family screening in first-degree relatives of patients with cardiomyopathy or early sudden death. Furthermore, in newly pregnant women a family history should be taken and women with a positive family history for cardiomyopathy and/or acute death should be evaluated by a cardiologist early during the pregnancy if screening has not been performed pre-pregnancy. Women with known cardiomyopathy should be under tight medical supervision during pregnancy as well as during the postpartum period.

Valvular heart disease is a frequent cause of cardiovascular morbidity during pregnancy in the Netherlands but has relatively low case fatality rates [19]. However, the cases described illustrate that the risk of severe native valvular disease should not be underestimated. Most women with valvular disease who died had been advised against pregnancy. Women who are advised against pregnancy for medical reasons may need intensive psychological and medical support and should be advised about safe and effective contraception. In women with a mechanical PHV, the main concern during pregnancy is balancing the risk of PHV thrombosis (which is relatively high with LMWH therapy) with the risk of bleeding and the risk of embryopathy (which exists only with warfarin therapy during the first trimester but not with LMWH therapy). Thrombosis and bleeding were causes of death in our population. All women who died due to PHV thrombosis were anticoagulated with LMWH. A high maternal death rate (9%) in women with a mechanical PHV has recently been described in the UK and seemed to be related to LMWH [31]. Given the higher risk of valve thrombosis, current European and American guidelines advise that LMWH therapy should be limited to the first trimester and the last month of pregnancy, and should be used only when frequent anti-Xa monitoring with dose-adjusting is available. In women needing a low dose of a VKA, continuing with the antagonist throughout pregnancy should be considered [32, 33]. Data concerning anti-Xa monitoring were not available to the MMMC and not taken into account in the assignment of substandard care. It is important to realise that the anticoagulation dose can change during pregnancy and therefore monitoring of the anticoagulation effect is more frequently necessary than pre-pregnancy or during the postpartum period [32, 33].

Though CHD is the most frequent underlying maternal heart disease in the Western world, only one death occurred, in a woman advised against pregnancy. The woman who died due to unrecognised pulmonary hypertension was known to have a condition associated with pulmonary hypertension (systemic lupus erythematosus). Pre-pregnancy evaluation in women with these conditions may influence pre-pregnancy advice, while current guidelines still advise against pregnancy in women with pulmonary hypertension because of high maternal mortality and morbidity rates [33].