From the Journals

Severe maternal morbidity climbs; racial disparities persist

 

Key clinical point: The rate of severe maternal morbidity rose 45% from 2006 to 2015.

Major finding: Black women were over 100% more likely to experience severe maternal morbidity, compared with white women.

Study details: Statistical analysis of Healthcare Cost and Utilization Project (H-CUP) data.

Disclosures: The study was funded by the Agency for Healthcare Research and Quality. The authors reported no conflicts of interest.

Source: Fingar K et al. Agency for Healthcare Research and Quality Statistical Brief #243.

View on the News

Obstetricians can’t flinch from their own biases

The data here confirm, again, that we have a crisis in maternal health care in the United States. Not only is severe maternal morbidity increasing, but disparities in outcomes between white and black women persist, such that severe maternal morbidity is twice as common among black women as it is among white women.

Dr. Alison Stuebe, associate professor of maternal-fetal medicine at the University of North Carolina in Chapel Hill

Dr. Alison Stuebe

This gap reflects multiple issues. First, we have to consider how health care systems aggravate disparities. There are higher rates of severe maternal morbidity among women covered by Medicaid, which suggests that publicly insured women need specialized care to address their higher-risk status. And yet, Medicaid reimbursement for maternity care is much lower than private reimbursement – in North Carolina, Medicaid reimburses $1,327.53 for global obstetric care (CPT 54900), whereas private, in-network insurance reimburses more than $2,800. This payment difference, in turn, drives private practices to turn away Medicaid patients, such that poor women receive different care than well-to-do women. This payment difference also means that safety net hospitals operate with fewer dollars to deliver services to low-income families. Essentially, this payment scheme values the lives of Medicaid-insured women half as much as the lives of privately insured women. If we want to eliminate disparities in outcome, we have to invest in the lives of women facing those disparities.

It’s tempting to attribute the gap to social determinants of health that are outside the control of the obstetrician-gynecologist. But education doesn’t protect black women: the New York City Department of Health and Mental Hygiene found that non-Hispanic black women with a college degree have higher rates of severe maternal morbidity than women of other race/ethnicities who never graduated high school. It’s a stark finding that’s illustrated by the heartbreaking story of Shalon Irving, a PhD epidemiologist at the Centers for Disease Control and Prevention, whose postpartum death was reported by ProPublica last winter.

One contributing factor is likely to be implicit biases that affect how well we see and hear patients who are different from us. I was mortified to see my own implicit biases when I took the Project Implicit Social Attitudes test (https://implicit.harvard.edu/implicit/). These aren’t the beliefs that we choose – they are the patterns that we absorb from our day-to-day lives. If you search Google images for “pregnancy,” “motherhood,” or “breastfeeding,” the overwhelming majority of images feature white women, whereas the phrase “welfare queen” is used to denigrate the reproduction of women of color. That’s the world we live in, and these patterns reinforce mental short cuts that are more likely to drive our behavior when we’re pressed for time, as we are every day in clinical practice. In one study in a pediatric emergency department, the busier the shift, the higher the providers’ implicit bias at the end of the shift.

We might start by looking at the images on our web sites and hanging in our offices and hospital corridors – do they look like the women we care for? Do they celebrate pregnancy and parenthood for diverse families? And how can we make sure that we pause, before we walk into a patient’s room, so that we can be fully present to her individual strengths and vulnerabilities, and tailor our care to her as an individual?

Alison Stuebe, MD, of the University of North Carolina, Chapel Hill, is medical director of lactation services and on the steering committee for Moms Rising North Carolina, the Breastfeeding Expert Work Group for the American College of Obstetricians and Gynecologists, and a board member of the Society for Maternal-Fetal Medicine. She was asked to comment on the AHRQ report by this newspaper.


 

FROM AN AHRQ STATISTICAL BRIEF

The rate of severe maternal morbidity in the United States has climbed steadily since 2006, increasing 45% overall in the decade ending in 2015, according to a new report from the Agency for Healthcare Research and Quality that also found large ethnic and racial, geographic, and socioeconomic variation in rates of severe maternal morbidity.

The longstanding increased risk for severe maternal morbidity for black women, compared with white women, continued essentially unchanged, with black women 112%-115% more likely to experience any of 21 conditions and procedures that defined severe maternal morbidity in the report. Disparities also existed between white women and those of Hispanic or Asian/Pacific Islander origin, but those gaps are narrowing, according to the report.

“Black women, Hispanic women, and women of other races/ethnicities were overrepresented among deliveries involving severe maternal morbidity, as compared with white women,” wrote Kathryn Fingar, PhD, and her coauthors. “White women constituted a lower percentage of deliveries with any severe maternal morbidity than they did other deliveries” – 23% lower.

The 21 indicators, developed by the Centers for Disease Control and Prevention, range from conditions such as renal failure and sepsis to in-hospital procedures such as blood transfusion and hysterectomy. Women were considered to have severe maternal morbidity if any of the indicators were present, regardless of whether in-hospital death occurred.

Deliveries involving severe maternal morbidity, 2006 and 2015

Dr. Fingar of IBM Watson Health, Cambridge, Mass., and her collaborators summarized data from AHRQ’s Healthcare Cost and Utilization Project (H-CUP) in the statistical brief. The two most common procedures that indicate severe maternal morbidity are blood transfusion and hysterectomy, and these indicators were tracked in deliveries where women had a condition that served as one of the severe maternal morbidity indicators.

Dr. Fingar and her colleagues noted that they excluded data for the final quarter of 2015, because that is when the transition from the 9th to the 10th edition of the International Classifications of Diseases was made.

In addition to the overall increase from 101.3 to 146.6 women per 10,000 experiencing any severe maternal morbidity, the incidence of blood transfusion during a delivery hospitalization, either alone or in conjunction with other indicators, rose from 78.9 to 121.1 women per 10,000, an increase of 54%.

Showing that increased blood transfusions were a major driver of the jump in severe maternal morbidity, the composite increase in the rates of all other indicators went from 34 to 42 per delivery hospitalization, an increase of just 24%, or less than half the increase in blood transfusions.

There was significant variation in trends over time for the rates of the other indicators: acute renal failure, shock, the need for mechanical ventilation, and sepsis all increased by at least 100% (range, 104%-134%), and the rates of aneurysm increased by 99%. Rates of other indicators fell; pulmonary edema, embolism, eclampsia, myocardial infarction, cerebrovascular disorders, serious anesthesia complications, and intraoperative heart failure and arrest all declined by 25%-53% during the data-gathering period.

Looking at the data another way, 78% of cases of severe maternal morbidity in 2006 and 83% in 2015 involved a blood transfusion, making it the most common indicator. Far fewer delivery hospitalizations involved disseminated intravascular coagulation (DIC) and hysterectomy, the next most common indicators, which were seen in just 8% of cases.

Blood transfusions were most likely in women with shock (72%), amniotic fluid embolism (63%), sickle cell disease with crisis (54%), and DIC (51%).

One in three women who experienced shock during delivery had a hysterectomy, as did more than 20% of women who experienced adult respiratory distress syndrome or cardiac arrest (or ventricular fibrillation).

Women delivering at the youngest and oldest ends of the age spectrum were more likely to experience severe maternal morbidity. For women younger than 20 years of age, the rate was 206 per 10,000. In the group of women aged 40 years and over, the rate of any severe maternal morbidity was 248 per 10,000 delivery hospitalizations. Women in their 20s and 30s had rates of 136 and 143 per 10,000, respectively.

Besides race/ethnicity and age, a variety of other patient characteristics were associated with increased rates of severe maternal morbidity and mortality, with higher rates seen in women from the poorest quartile than the wealthiest (177 versus 122 per 10,000, respectively). Women with Medicaid were more likely than those with private insurance to experience severe maternal morbidity during a delivery hospitalization (175 versus 121 per 10,000, respectively).

Similarly, severe maternal morbidity was more common in safety net hospitals and in minority-serving hospitals (182 versus 128 and 176 versus 123 per 10,000, respectively) than other hospitals.

Regionally, severe maternal morbidity was most common in the Northeast and the South, at 165 and 164 per 10,000, respectively, compared with rates of 132 and 116 in the West and Midwest, respectively.

Hospital deaths per 100,000 delivery hospitalizations were 248% higher for black than white women in 2006. By the end of 2015, that figure declined modestly to 193%, with absolute rates of 19 versus 5.5 deaths per 100,000 delivery hospitalizations for black versus white women in 2006. In 2015, the absolute rates were 11 versus 4 per 100,000.

The study was conducted by the Agency for Healthcare Research and Quality. There were no reported conflicts of interest.

koakes@mdedge.com

SOURCE: Fingar K et al. Agency for Healthcare Research and Quality Statistical Brief #243.

Next Article: