How new mothers’ health data is saving women’s lives in California

Dr. David Lagrew is an OB-GYN and founding member of the CMQCC. 

How did the California Maternal Quality Care Collaborative get started? 

In 2005 the state [of California] noticed that our maternal mortality rate was increasing. So CMQCC was formed in 2006 with the intent to lower maternal mortality in the state of California, funded by the state department of health. 

It became apparent that we were going to need good data to make the kinds of changes we needed to make. And so a couple of years later they initiated the Maternal Data Center. In a nutshell, we take birth certificate data, which is public domain, and we combine that with a discharge file, a set of the discharge codes, and other information. From that we can derive all sorts of statistics. If you go on the website, there’s actually a sample that you can play with. It’s really a pretty slick application.

What has happened to the maternal mortality rate in California since this project launched? 

The maternal mortality rate fell by more than 50 percent between 2006 and 2013, even as rates continued to rise across the U.S. 

Can you tell us more about the kinds of data you collect? How do you know what interventions are working or not working?  

We can get state data on all hospitals in California. But also we’ve started what we call our active track, which today most hospitals participate in. Rather than waiting, the data center offers a way that people can send their file on a monthly basis. And that way they can get back their statistics 45 days after the close of the month. And so that from a quality improvement standpoint, obviously that’s a better way to do it because the more timely the data the more impact it has.

Our first focus was on maternal hemorrhage or bleeding problems…The big metrics we use are transfusion rates––there’s a metric we have called severe maternal morbidity. That’s a measure of, “Hey, did the mom go to the intensive care unit? Or did she get a big transfusion? Or did she have kidney failure, etc.?” That we can calculate automatically. The same with most of the C-section rates. There are certain metrics that are defined out there and certain metrics that frankly you have to kind of create. There is an organization called The National Quality Forum (NQF). It’s actually an organization that vets quality metrics. And so we actually took some metrics NQF and got them blessed. It’s a pretty rigorous process that is set up to do it. 

Occasionally there are some things, breastfeeding for example, where there’s just no discharge code. So you have to sample a small set of your population—what the data center does is give you a list of patients and say, “Hey, go check these charts and see what happened.” Then we use that data to manually track outcomes.

How does the data then inform best practices to improve maternal healthcare? How do you share out the results of your findings? 

We started putting out toolkits, which I call “collections of good ideas for change”. These always include some metrics, [informed by data we’ve collected from hospitals]. Then we would hold these collaboratives with hospitals so they could learn to implement the toolkits in their particular care settings. They started sharing ideas and actually gave us ways to improve the toolkits. It caught on by the early teens—by 2011, 2012. Usually somewhere between 20 and 40 hospitals at a time would be in these collaboratives.

Can you tell us about one particularly important innovation that’s taught in the toolkits? 

The hemorrhage cart is one of the most straightforward examples. In the toolkit, we have clear instructions that hospitals can follow, step-by-step, to immediately address something like dangerous blood loss, and it tells you what equipment hospitals need to have on-hand and ready to go.

So the maternal data is really the core of the project, and then that informs the toolkits, the toolkits inform the hospitals, and it continues in a loop. Is that right? 

If we couldn’t get [the data back from the hospitals], then it’d be pretty hard to tell how we’re implementing these toolkits. A lot of our metrics are so-called balancing measures that help you make sure you’re not hurting anybody. So for example, when we’re lowering the C-section rate, are we causing more morbidity to the mom or the baby? And the answer is no. Doctors and nurses as a group are pretty skeptical, and so you have to have pretty high-quality, meaningful data. 

Do you think about how the metrics apply to different demographics? Can you tell us more about the The California Birth Equity Collaborative?

We’ve been able to lower our maternal mortality rate here in California, which is the good news. The bad news is, although we have lowered maternal mortality for all racial groups, for African-Americans, it’s two to three times the rates of other groups. 

So now we’re going back trying to figure out what’s in the data. What can we learn to help us bridge that gap?

We’ve looked at a number of different metrics and combined them with some of the socioeconomic data that we have and other risk factors. And it boils down to,“that’s not the answer.” You have to practice equity, in other words, give women what they need to improve their health and wellbeing. We are just finishing up a toolkit to work on racial disparities. 

Can you tell me more about what these collaboratives are, that you’ve mentioned? 

We get a group of hospitals together and then we have them work on these change concepts together. When I say collaboratives, I really mean that. [The collaboratives] usually last about a year, and by the end, [they are] the experts in the room because they’re the ones with the real life experience. 

You learn very quickly in healthcare that you can’t force providers to do things. You really have to win their hearts and minds. So if you don’t design something well that’d be palatable, you can get a little change, but it doesn’t stick. That’s the real key to all this: coming up with changes that can A) be accomplished and B) stay changed for the better.

Are these toolkits being used in other areas beyond California?

Our toolkits are open source. Because it’s public domain and not proprietary, we have released them. With the hemorrhage toolkit, we put it up on our website, and literally within a few weeks we had hundreds and then soon to be thousands of downloads of the toolkit. It’s probably been translated to 10, 12 languages. 

There’s now a [multi-state] push. Here on the West Coast, Oregon and Washington have started their own collaborative to help do some of the things we’ve done. In fact there was funding through Congress to get these initiatives going in all states. 

What were some key elements for success, in making something like the CMQCC happen? 

One of the key elements is how multidisciplinary it’s been. [In addition to] health professionals like nurses and midwives, there are health administrators, and organizations like the March of Dimes. We also have a nice mixture between academic centers and researchers and actual community providers. Our executive committee is about 20 to 25 people, well represented by the different constituencies.

Do you have any other thoughts, on things we haven’t asked about? 

Hopefully what we’ve done in California will spread because I think everybody could benefit from this. Having said that, though—boy, the work is never done. It’s always a work in progress. But we’ve come a long way.