New Research Showcases Duke’s Commitment to Global Cardiology and Care in Kenya

Rheumatic heart disease (RHD) is endemic in East Africa, but the extent and severity of cardiac disease in pregnancy has been unknown.

Rebecca Lumsden, MD, of Duke, presented findings from a retrospective study at the Moi Teaching and Referral Hospital (MTRH) in western Kenya at the American Heart Association's Scientific Sessions held November 11 to 15, 2017, in Anaheim, CA. As a Doris Duke clinical research fellow, Lumsden spent 1 year in Kenya and performed research at MTRH for 9 months while working on the study presented at this annual meeting. MTRH is 1 of 2 facilities in Kenya that offer specialized cardiovascular (CV) and obstetric care. Other researchers involved in the study included Lawrence P. Park, PhD, and Gerald S. Bloomfield, MD, MPH, of Duke. Bloomfield was a Fogarty fellow at the National Institutes of Health (NIH); he performed his fellowship in Kenya from 2009 to 2010.

Lumsden and colleagues sought to define the spectrum and severity of CV disease in pregnancy in western Kenya. They performed a retrospective chart review of all pregnant women who were admitted to MTRH between January 2011 and April 2015. Risk assessment of adverse outcomes was based on the Cardiac in Pregnancy (CARPREG) prediction score, which is a validated model that estimates the risk of a maternal cardiac event during pregnancy. The researchers used a modified CARPREG score that assigned 1 point for each of the following:

  • Prior cardiac event or arrhythmia
  • New York Heart Association (NYHA) class III or IV heart failure (HF) or blood oxygen saturation of less than 80% on admission
  • Mitral valve area of less than 2 cm2 or aortic valve area of less than 1.5 cm2
  • Left ventricular ejection fraction of less than 45%

A total of 97 cases of cardiac disease were noted in pregnant women, 43% of whom were diagnosed during pregnancy or post partum. RHD affected 75% of pregnant women with cardiac disease; of these, 63% had mitral regurgitation and 56% had mitral stenosis. More than 60% of patients had NYHA class III or IV HF. Additionally, nearly one-half of patients had pulmonary hypertension, and approximately 10% of women had had a prior CV event.

"Roughly 75% of the women in the study had been admitted to MTRH for antenatal care at least once. We found that most pregnant women had advanced cardiac disease, with high rates of severe valvular stenosis and pulmonary hypertension," remarked Lumsden. "Cardiac disease in pregnancy in Kenya is associated with a high rate of maternal mortality and serious maternal and neonatal morbidity."

The maternal mortality rate in the study was 9%, with a neonatal fatality/fetal demise rate of 14%. The majority of pregnancies (80%) had at least 1 maternal cardiac event or neonatal adverse event.

TABLE 1. Adverse Events Among Pregnant Women With Cardiac Disease

Maternal mortality 9.3
Any maternal adverse event 79.8
Any cardiac event 55.6
Cardiac arrest 3.1
Pulmonary edema 19.6
Arrhythmia 7.2
Heart failure 38.1
Stroke 2.1
CCU/ICU admission 21.7
Any neonatal event 58.8
Intrauterine fetal demisea 11.2
Neonatal deathb 5.1
Premature birthc 23.4
Low birth weightd 30.4

CCU = critical care unit, ICU = intensive care unit.
aDeath > 28 weeks' gestational age to birth.
bDeath at birth through 30 days of life.
cBirth < 37 weeks' gestational age.
dBirth weight < 2,500 g.

The modified CARPREG risk score did not predict this high degree of mortality and morbidity. For example, pregnant women with cardiac disease and a modified CARPREG risk score of 0 had an adverse event rate of 61%, a cardiac event rate of 30%, and a mortality rate of 5%. In addition, the neonatal event rate for women with a "low" modified CARPREG risk score was 42%. All women with a modified CARPREG risk score of 2 or higher experienced an adverse event, with 82% having a cardiac event. The mortality rate for women with these scores was 11%, with the neonatal event rate approaching 60%.

Lumsden summarized the findings: "This is the first look at cardiac disease and pregnancy in East Africa. Now we know that risk tools developed for high-income nations are not applicable to developing countries. As we move toward a global approach to cardiology, we need to better understand the unique risk factors and barriers to care for this high-risk population in order to design strategies for coordinated, multidisciplinary care to prevent and treat CV disease in resource-limited settings."

"This research builds on Duke's previous research projects in, and commitment to, Kenya," Lumsden added. For instance, Duke's Cardiac Intensive Care Unit at the MTRH hospital is an NIH Center of Excellence that serves more than 8 million people. In addition, Lumsden is currently participating in Duke's Global Health Pathway for residents and fellows.

Source: Lumsden RH, Barasa F, Park LP, et al. Defining the burden and predicting the risk of cardiac disease in pregnancy in Kenya. Presented at: American Heart Association Scientific Sessions 2017; November 11-15, 2017; Anaheim, CA. Abstract M2188.