Cardiac and Obstetric Complications of Pregnant Women with Rheumatic Heart Disease in Sub-Saharan Africa: A Systematic Review

Study objectives

Our study aimed to address several questions about the cardiovascular and obstetric risks surrounding pregnancy in women with RHD in SSA. We studied the rates of cardiac and obstetric complications in pregnant or postpartum women with RHD. We also investigated how many pregnant women with severe RHD received cardiac intervention during pregnancy or within 12 months of delivery.

Study protocol and registration

This review followed the PRISMA guideline for systematic reviews (28) and was registered in the International Prospective Register of Systematic Reviews (PROSPERO) database under the registration number CRD42024628121.

Search strategy and information sources

The search was developed and conducted by a professional medical librarian in consultation with the author team and included a mix of keywords and subject headings representing pregnancy, RHD, and LMICs. Search hedges or database filters were used to remove publication types such as editorials, letters, notes, and conference abstracts as was appropriate for each database. Searches were conducted in MEDLINE via PubMed, Embase via Elsevier, Web of Science via Clarivate, Global Health via EBSCOhost, and Global Index Medicus via the World Health Organization (WHO). The searches were executed on December 16, 2024, and found 1,478 unique citations. Complete reproducible search strategies, including search filters, for all databases are detailed in the Supplemental Materials (Supplemental Table 1). All citations were imported into Covidence, a systematic review screening software, which also deduplicates the results.

Inclusion/exclusion criteria

Inclusion criteria included studies of women of child-bearing age (15–49) with a diagnosis of RHD living in SSA; study types included were randomized controlled trials (RCTs), non-randomized observational studies including retrospective and prospective cohort studies, case-control studies, case reports, and case series. English-language studies in the year 2000 or later were included. Case reports were included in the initial search to obtain useful references; however, data was not extracted from case reports, and these studies were ultimately excluded. We included studies that reported combined outcomes for pregnant women with cardiac disease if at least 60% of the study population consisted of participants with RHD, and we excluded studies in which less than 60% of participants had RHD. We excluded studies focused solely on congenital or degenerative heart disease.

Study selection

Abstract review of records was performed by two investigators (EA, GSB). Each author reviewed the titles and abstracts independently to determine if they should be included. In cases of disagreement, the record was revisited to reach a consensus.

Data extraction

A Google Form was developed for authors to use as a tool for data extraction. Data collection was performed by two authors (EA, DN). The key data were predetermined and included title, first author’s name, year of publication, study design, country of origin, study setting (hospital-based, community-based, or combined), total sample size, number of women of reproductive age (15–49) with RHD, type of valvular lesion and if severe disease is present, and number of women with RHD currently pregnant or who delivered within the last year. Of this sample of pregnant or recently pregnant women with RHD, we collected: total number of any cardiac complication as determined by the individual study, hospitalization for heart failure, arrhythmia (atrial or ventricular), infective endocarditis, stroke, venous or arterial thromboembolism, valvular complication (complication following valvular intervention), acute coronary syndrome or death due to cardiovascular disease, total number of any obstetric complication as determined by the individual study, spontaneous pregnancy loss (including miscarriage, first and second trimester losses, and intra-uterine fetal demise and excluding terminations), newborn/neonatal death, maternal death, preterm delivery, preeclampsia, intrauterine growth restriction (IUGR) or low birth weight, postpartum hemorrhage, rates of cardiac intervention including valve replacement, valve repair/valvuloplasty/commissurotomy, valve myotomy, left atria reduction, left atrial appendage exclusion (clip/stich), and Maze procedure. Definitions for early pregnancy loss came from the American College of Obstetricians and Gynecologists. Definitions for newborn/neonatal death, maternal death, preterm delivery, and intrauterine growth restriction were obtained from the WHO. Maternal death was defined by the WHO as the death of a woman while pregnant or within 42 days of termination of pregnancy, irrespective of the duration and site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management but not from accidental or incidental causes (29). This is an aggregate number that included death due to maternal cardiac causes.

Quality assessment

Two authors (EA, DN) independently assessed the quality of the studies using a Joanna Briggs Institute (JBI) critical appraisal tool (30, 31). Consensus was reached for each checklist item. The appraisal tools were specific to the study type (i.e., case series, cohort study, and case-control study).

Data synthesis

Data was combined in a narrative format due to a low volume of studies of varying quality and to mitigate potential issues of data completeness. There was a large amount of variability between the specific data points collected by each study.

Study characteristics

Study characteristics are summarized in Table 1. These studies were published between 2000 and 2022 and represented 4,477 individuals. Seven case series, one cohort study, and one case-control study were included. One of the initial objectives of this systematic review was to compare pregnant and non-pregnant age-matched women with RHD; however, we did not find any studies that compared these two groups. All the studies were of pregnant women with cardiac disease, mostly RHD. We included two studies that reported on procedural interventions for pregnant women with RHD (32, 33).

Study sample sizes ranged from 21 to over 3,000 participants with a median of 95 study participants [Q1, Q3: 54, 264] (Table 1). Our nine studies included a total of 787 pregnant women with cardiac disease, of whom the majority or all (63.5% to 100%) study participants were pregnant women with RHD. All studies reported a mean or median age of study participants between 26 and 30 years old. Most (N = 8) studies were hospital-based; one was a compiled study of regional health centers and one hospital (34).

Detailed reasons for exclusion can be found in Supplemental Table 2. The most common reasons for exclusion were study location (many were outside of the study region), study type (review or case report), and reported outcomes that were not included in our review protocol. Many excluded studies were conducted in India, Australia, and Southeast Asia or did not include details for participants specifically from SSA in the case of international cohort studies (16, 35). We excluded one study that treated patients from Portuguese-speaking African nations, as these patients decided to permanently relocate and receive their care in Portugal (36). We included seven studies in SSA that presented combined outcomes for patients with cardiac disease with RHD present in ≥60% of the study population; other cardiac diseases included congenital disease and peripartum cardiomyopathy (32, 34, 37, 38, 39, 40, 41). Of the seven studies with combined outcomes, five studies reported RHD in >80% of study participants (32, 34, 38, 39, 40). We excluded six studies with outcomes of interest in which pregnant women with RHD were <60% of the total study population (27, 42, 43, 44, 45, 46).

Included studies were conducted in South Africa (3), Kenya (2), Ethiopia (2), Senegal (1), and Uganda (1). Geographic distribution and number of the included studies are depicted in Figure 2. Eight of the studies collected information about the specific valve lesions of patients, two of which were studies of pregnant women with rheumatic MS (33, 47). We found variations in the classification of MS across studies. MS was reported in 357 participants across seven studies; severe MS was reported in 116. In one of these studies, approximately half of participants had an MVA of less than 1.3 cm² and 29% had an MVA of ≤1 cm (33). Joint guidelines from the American Society of Echocardiography and the European Association of Echocardiography classify moderate MS as MVA between 1 and 1.5 cm² and severe MS as MVA < 1 cm² (48). In the other study, 55.2% (n = 32) of participants had severe MS, defined as MVA ≤ 1 cm² (47). One study did not give specific measurements but stated that 27.5% of participants had mitral regurgitation and 20.8% had MS (41). Among the other five studies, there were a total of 327 participants with RHD, of which mitral valve lesions were predominant. Aortic lesions were much less common; aortic stenosis was reported in 14 participants, and AR was reported in 49 participants across these studies.

Figure 2

Cardiac complications of pregnant women with RHD in the included studies are shown in Table 2. No participants had a known diagnosis of acute coronary syndrome. Only one valvular complication was captured, a thrombosed mechanical valve (32). Heart failure was the most common cardiac complication across all included studies, occurring in 12.9% to 100% of study participants. Arrhythmia was the second most common cardiac complication, reported in eight studies, ranging from 1.1% to 36% of study participants. One study did not have any deaths in their sample, whereas the remaining eight studies reported deaths due to cardiac causes (median: six deaths due to cardiac disease; total number of deaths: 56). Infective endocarditis and stroke were uncommon, each occurring in a total of three study participants.

Obstetric complications of pregnant women with RHD in the included studies are shown in Table 3. Pregnancy loss (encompassing early pregnancy loss and intrauterine fetal demise (IUFD) or stillbirth) was recorded in all nine studies, ranging from 5.5 to 22.4% of participants. In one study, early pregnancy losses made up most pregnancy losses (47); in several other studies, IUFD/stillbirth comprised more of the cases (33, 34, 37). Neonatal death was recorded in four of the studies, occurring in <5%. Preterm delivery was common and captured in all the included studies, ranging from 5.2% to 35.2%. Though low birth weight had variable definitions across studies, it was reported in four studies; up to one-quarter of neonates were born at low birth weight. Maternal death occurred in 44 individuals across seven studies. Maternal death rates, most of them due to cardiac disease, ranged from 1.7% to 34% in our studies. Seventeen individuals died in one study, representing 34% of the study participants (38).

A total of 28 patients from two studies received mitral valvulotomy during pregnancy for MS, and one patient received a mitral valve replacement at 32 weeks gestation (32, 33). The remaining studies did not report on interventions for RHD in pregnancy.

Quality assessment

Full assessments of risk of bias according to JBI are presented in Supplemental Tables 3–5 for case series, cohort, and case-control studies, respectively. Many studies were case series assessing outcomes of pregnant women with cardiac disease; therefore, we were cautious in interpreting results lacking control groups and using mostly descriptive analysis. Bias in the measurement of outcomes may be introduced in some of the case series; several did not use clear definitions of outcomes in the methods section. The case-control study and cohort study had stronger definitions and adjusted for confounding variables. Overall, studies maintained a level of quality sufficient for inclusion in this systematic review.

Cardiac complications

Heart failure was also the most common cause of hospital admission in pregnant women with rheumatic mitral valve disease in the ROPAC study, a large multicenter prospective cohort study located in countries with emerging economies, including South Africa. ROPAC reported combined outcomes for emerging countries without country-specific data and was therefore not included in our formal analysis. Heart failure occurred in 26.7% of the ROPAC cohort, which is the lower end of reported heart failure in our nine studies (16). In a Brazilian cohort of pregnant women with RHD, 21.4% experienced a cardiac complication, of which the majority (71.4%) had heart failure or pulmonary congestion, 16.8% had arrhythmia, and 1.7% had infective endocarditis (51). Heart failure/pulmonary congestion only occurred in 15.3% of the total sample with RHD, in contrast to our studies, in which the incidence of heart failure trended higher. In a Canadian cohort of women with rheumatic MS, 35% of pregnancies were associated with maternal cardiac complications, which were significantly higher in women with severe MS compared to moderate and mild MS, and no maternal deaths occurred (52). Predictors of cardiac complications included moderate or severe MS, pulmonary hypertension, and a history of valve repair (ibid.). These global trends are consistent with the higher rates of cardiac complications we observed in our studies. Overall, maternal mortality ranged from 1.7% to 34% in our studies; a systematic review focused on South Asia found a pooled maternal mortality rate of 26.1 per 1,000 women with cardiac disease (49).

Obstetric complications

Obstetric complications were frequently observed across the reviewed studies, though not comprehensively captured in most studies. Rates of obstetric (8%) and fetal events (30%) were lower in the study of Canadian women than in our qualitative estimates (52). Among these, preterm labor/delivery was the most frequently reported obstetric event in our review, with incidence ranging from 5.2% to 35.2%. This aligns with an international systematic review and meta-analysis (with two included studies from South Africa) in which preterm birth was the most reported obstetric adverse event in pregnancies with RHD (50). Hameed et al., in a study from the US of a majority Hispanic population, found a statistically significant decrease in the duration of pregnancy, an increase in the incidence of premature deliveries, and an increase in the incidence of IUGR among patients with valve disease (53).

Spontaneous pregnancy loss was another recurring outcome reported in all our included studies (incidence ranged from 5.5% to 22.4%). The highest incidence of IUFD or stillbirth was 10.3%. Neonatal death was less common or less often reported. The ROPAC study noted miscarriage or fetal death in 5.1% of women with MS and 3.4% of women with MR (16).

IUGR, or low birth weight, was a common problem occurring in neonates of mothers with RHD, occurring in up to one-quarter of infants in our studies, which is similar to estimates from multiple studies performed in India (54). Prior surgical correction of valvular lesions may protect against lower neonatal birth weights (55). There is not a pathological link between preeclampsia and RHD that has been directly studied, though it is known that low-income individuals are at higher risk and rates are higher in women with rheumatic diseases such as systemic lupus erythematosus (56, 57). We noted preeclampsia in 1.3–23.8% of the pregnant women in our studies. This observation should be further elucidated to determine if there is a link between RHD and preeclampsia. Pregnancies impacted by RHD are associated with a higher risk of adverse outcomes for both mother and fetus in LMICs compared to high-income nations and in SSA particularly.

Valve interventions

Though pregnant women in SSA have high rates of severe disease, particularly MS, and are at high risk for complications, valve interventions during pregnancy are rarely performed. The low (n = 29) number of patients receiving valve interventions during pregnancy in our included studies is consistent with findings from other studies and registries in LMICs. The VALVAFRIC study, a multicenter registry of over 3,000 patients with RHD in eight Western and Central African countries that characterizes echocardiographic features, symptoms, interventions, and sociodemographic characteristics, reported that only 2.2% of patients qualifying for surgical intervention were operated on; all of these patients received surgical care abroad (58). In the ROPAC registry, less than 5% of patients received an intervention during pregnancy (14 balloon mitral commissurotomy and two surgical mitral valve replacements) (16). A retrospective study in India reported that 29 women received balloon mitral valvotomy during pregnancy (10.6%) (54).

If valve replacement surgery is pursued, the type of valve implanted—mechanical or bioprosthetic—has lasting implications. Mechanical valve replacement requires lifelong anticoagulation with a vitamin K antagonist, typically warfarin (20). Warfarin is a known teratogen; levels may be decreased during pregnancy or alternative anticoagulants may be offered that are associated with higher maternal complications (20). Bioprosthetic valves do not require lifelong anticoagulation; however, in young patients, the risk for degeneration of the valve is high.

In resource-limited settings, the optimal management for advanced RHD in pregnancy is unclear, with each pathway posing risks. Severe native valve dysfunction during pregnancy, especially with heart failure or pulmonary hypertension, increases the odds for poor outcomes for mother and baby; however, valve intervention, cardiopulmonary bypass, and postoperative management must be navigated with caution.

Implications and areas for future research

Understanding the complications of RHD in pregnant women is key to counseling patients of reproductive age. Indeed, we know that current efforts to encourage safe pregnancies in women with RHD in SSA face challenges such as affordability and cultural acceptance of contraception, underdiagnosis of RHD, and patient education surrounding the disease. Raising public awareness and championing public health efforts in RHD screening are key tools for addressing the burden of RHD. Routine antenatal echocardiographic screening for RHD presents an important strategy in identifying previously undiagnosed RHD early in pregnancy (59, 60). Efforts to bring diagnostic echocardiography to high-risk populations have surged in recent years, a crucial step forward to diagnose RHD in pediatric populations (61, 62, 63, 64).

One Tanzanian study at a tertiary cardiac center found that the proportion of reproductive-age women with RHD using contraception was 7.1%, though the majority of participants in the study were high-risk according to the modified WHO maternal risk classification system for predicting adverse pregnancy outcomes among women with cardiac disease (65). A qualitative study from Uganda of postpartum women with RHD highlighted the central role of the spouse in medical decision-making (66). Educational programs for women with RHD are indispensable; future efforts may focus on support groups or patient-led initiatives to provide resources and teaching that can empower reproductive health. These efforts may involve the partners and spouses of women with RHD to facilitate shared understanding (67).

Another key component to enhance clinical care of women with RHD is implementing prospective registries to track contemporary outcomes of women during pregnancy. The VALVAFRIC and ROPAC studies are models for cohort studies (retrospective and prospective) that can be translated to future international efforts (58). Moi Teaching and Referral Hospital (MTRH) in Eldoret, Kenya, is aiming to address this gap by creating a prospective structural heart disease registry that is inclusive of obstetric history and complications. Future research can leverage the foundations of such registries for large-scale initiatives bridging multiple centers.

The 2015 Addis Ababa roadmap of actions to eradicate RHD in Africa recommends expanding access to reproductive health services for women with RHD and establishing prospective RHD registries (68). Another key component of their recommendations is establishing cardiac surgery centers of excellence. Valve repairs and replacements are uncommon in SSA due to a paucity of resources, though it is estimated that the need for cardiac surgery to address RHD in SSA is 200–250 per million (22, 23, 24). Suggestions for surgical capacity building in SSA include mentorship models between HICs and LMICs, outcome tracking, and value-based care (69). Such efforts require strong government backing and medical insurance systems that will support surgical interventions (70).

Risk calculators for pregnancies in patients with cardiovascular disease that are commonly utilized in high-income countries, such as the Cardiac Disease in Pregnancy Study (CARPREG II) and modified WHO prediction tools, do not accurately predict risk outside of high-income countries (37, 71). The growing field of cardio-obstetrics offers a pathway to patients receiving specialty care for RHD in pregnancy, though has yet to be formally established in many settings. Joint cardio-obstetrics clinics offer the opportunity to validate and integrate pregnancy risk classifications into routine RHD care and serve as a medical home for the management of RHD in pregnant women.

Strengths and Limitations

The main strengths of our review are the focus on SSA, consideration of both cardiac and obstetrics outcomes, and adherence to standard review protocols. There are also limitations to consider. This review should not be taken to summarize the issue of RHD in pregnancy for much of a continent but instead offers a high-level overview of the published literature on complications that pregnant women with RHD and their neonates may encounter in a vast region that faces historical and current social, economic, and political challenges. Definitions on obstetric and cardiac complications in pregnant women with RHD were not collected systematically across studies, though we applied standardized definitions in our protocol. Many articles presented combined clinical data for patients with RHD and patients with other cardiac conditions; therefore, our review of the impact of RHD on outcomes was less precise from those studies. We attempted to mitigate this limitation by only including studies in which most participants had RHD. Most included studies were performed in tertiary referral hospitals with small sample sizes of individuals with cardiac disease. We did not conduct a meta-analysis or assess statistical effect heterogeneity; therefore, we cannot assume precision of our estimates. We chose to focus mainly on maternal outcomes within pregnancy and the immediate postpartum period; future studies are needed to assess the consequences of RHD on the children of affected mothers and follow fetal outcomes into childhood.