Addressing Preventable Deaths in Maternal Care

By Matt Phillion

A recent CDC report found that four out of five pregnancy-related deaths are preventable. Key findings in the report include details about the underlying causes of these deaths:

  • Mental health conditions (23%)
  • Excessive bleeding/hemorrhage (14%)
  • Cardiac or coronary conditions (13%)
  • Infection (9%)
  • Thrombotic embolism (9%)
  • Cardiomyopathy (9%)
  • Hypertensive disorders of pregnancy (7%)

The report also noted that the leading underlying cause of death varied by race and ethnicity, with cardiac and coronary conditions the leading cause among non-Hispanic Black patients, mental health among Hispanic and non-Hispanic white patients, and hemorrhage among non-Hispanic Asian patients. The numbers clearly point to a need to reassess and improve outcomes across the board. But where do we start?

Wildflower Health is a value-based maternal health platform looking to address these issues. “We started working on this problem 10 years ago. It can be demoralizing at times that it’s getting worse,” says Leah Sparks, CEO and founder.

Sparks says a number of interwoven issues converge to lead to the greater problem of poor maternal health: provider burnout, staffing shortages, patient complication rates, and more. “We’re thinking about it in three broad buckets,” she says.

The first bucket is providers. “On the provider side, there’s shortages of providers, burnout, and this is across all of U.S. healthcare. They’re strapped for time so they can’t provide whole-health care,” Sparks says. While there’s a lot of focus on maternity deserts where lack of care is stark, these same issues strike in urban environments where hospitals are nearby: Patients are bounced between silos, providers are overworked, and the situation creates a systemic challenge.

The second bucket is the patient. “We have a churn-and-burn provider situation, and patients may not have access to care. There are screenings they may not have,” says Sparks. “Or they have no childcare for their other children and so they may struggle to make it to medical appointments that would address blood pressure issues. We see this play out in maternal complications.”

The third bucket is the payment model itself. “We’ve evolved this company to address the systemic challenges in the first and second buckets, but the third is how do we pay for it?” says Sparks. “We can’t address buckets one and two unless we address the payment model so that rewards transformative care. That’s how we see core problems resolved.”

How to inspire transformative care

There’s been movement toward enabling value-based care in recent years, and CMS has gotten out in front of value-based as a model, but it needs to cascade into other areas, Sparks notes.

“In addition to the care solutions we provide, we’ve become a value-based care enabler for our clients,” says Sparks. Some clients might still be 90% fee-for-service and asking how to cross that chasm. There needs to be a way to make the change easy to administer and remove barriers to entry, says Sparks.

“We want to make it easy for providers: What does maternity cost, and where are the greatest opportunities to improve care for mom and baby across an entire population? Are there high NICU rates or non-delivery rates? Are factors not being cared for at the hospital?” says Sparks. “What are the percentages of diabetes, hypertension, depression before the patients become pregnant, and how does that impact pregnancy?”

Armed with that information, step two is providing the tools to move toward transformative care. “We can’t go to a provider and say, ‘Reduce newborn OB costs by 10%,’ ” says Sparks. “We have to help show them the way. ‘Here’s how we’re going to do it, how we’ll help diabetic or hypertension patients.’ ”

How to scale change

How do you scale care transformation, particularly in regions where there aren’t enough OB practices for the patients, let alone practices willing to take a risk by changing to a value-based care model?

How the care is delivered comes into play here. While you can’t deliver a baby over telehealth, Sparks says, a vast number of virtual services can monitor the mother’s health throughout the pregnancy.

“You can aggregate care in these rural areas under a virtual clinic model,” she says. “You can do a lot of prenatal care and risk assessment virtually during the 40 weeks leading up to the birth.”

These virtual models can perform risk assessments, issue blood pressure cuffs that can be monitored remotely to track the mother’s health, and enable physicians to interface with other care providers on the ground to coordinate with the labor and delivery team. “There’s a lot that can be done to extend care, but it’s a model that needs to be tested to get the evidence-based assessment,” Sparks says.

This isn’t just applicable to rural areas, though, she notes. “Everything goes wrong in the month between appointments,” says Sparks. “That’s the stuff you don’t know about. We see the impact of escalations between visits.”

A key area of Wildflower’s model is using health advocates and coaches in the home. This enables watching for indicators like blood pressure spikes and assessing for depression that may need immediate intervention.

Patient education also plays an important role in improving maternal care outcomes so patients take advantage of the care and screenings offered. “With Medicaid, it’s often about education: ensuring they know they’re not going to have copays, not going to receive surprise bills, educating women about their own benefits,” says Sparks.

Since the CDC report was published, additional studies have come out on the topic of maternal death rates and patient populations. Sparks notes that the information wasn’t new to those working in the trenches of this issue, but it did shine a brighter light on existing problems.

“The racial inequalities were not highlighted previously the way they have been the past few years. The tragedy of high maternal death rates based on ethnicity and race has been an area that is not a surprise, but it saddens and angers you,” she says. “It’s tied up not just in the clinical, but in society and how much bias is placed in that, and it’s something that’s not as easy to drive as care compliance.”

Wildflower takes a bottom-up approach. “When we look at patients with diabetes, we know we have poorer outcomes across specific ethnicities and racial groups—and also look at the data by LBTQA as well. We can then look at the data and deliver more services based on risk profile, whether clinical factors or socioeconomic factors,” says Sparks. “For groups we know have a higher risk of poor outcomes as a combination of these factors, we’re personalizing these patient journeys to move the needle.”

Because race, ethnicity, and orientation are often not included in claims data, looking at the whole-person picture sometimes poses a challenge. “The first step is making sure payers and providers work together to go beyond claims and really measure how we’re doing with all these sub-populations,” says Sparks.

Change is hard in healthcare, and often slow because of the risks involved when changing processes. However, Wildflower is confident it can succeed based on what it’s already helped customers and patients achieve, such as:

  • Threefold increase in identification of high-risk patients
  • 38% fewer missed appointments
  • $43,000 in savings for high-risk patients and $2,000 in savings for low-risk patients

“We need, just like anything else, to have the move to value-based care to inspire people emotionally and pragmatically, and need to line those up economically and financially,” says Sparks. “I hope we’ll be able to move in a way that’s not just about women’s health in terms of having a baby but a whole-women’s-health model. If we have a broader model instead of being siloed and have the whole care team align to a payment model that supports it, that’s nirvana to me. That’s where I hope to go and that we can be a part of it.”

Matt Phillion is a freelance writer covering healthcare, cybersecurity, and more. He can be reached at