Healthy Data Podcast_ Kendra Folh (Memorial Hermann) & Jordan Cooper (InterSystems)-20250520_103030-Meeting Recording
May 20, 2025, 2:30PM
18m 31s
Jordan Cooper 0:03
Women and children service line for the system. Kendra, thank you so much for joining us today.
Kendra Folh 0:07
I’m very excited to be here.
Thank you for having me.
Jordan Cooper 0:09
Yeah. So we have a wide set of issues that we’re going to be discussing today, all pertaining to data surprises, healthy data, podcast.
I’d like to ask you to start us off with a conversation or description of how data the current state of data at Memorial Hermann.
What kind of issues are arising from clean or lack of clean data at Memorial Hermann?
Kendra Folh 0:35
Well, I think I could definitely speak to what is the state of Theta for all maternal health and you know, not just within my organization, but I also do a lot of work at the state of Texas level and so significant opportunity to really understand what’s happening in.
Population. We know maternal health is population health, so how do we get data that’s relevant, actionable and timely, so we can really steer the ship. We know we have opportunities to improve our maternal mortality and morbidity, but where is the data that really helps us other than we?
Have a problem.
How do we fix this problem?
And so how do we find the data?
How do we clean that data?
There’s multiple different sources of this to really create.
A story so we know where we can go and prove.
Jordan Cooper 1:17
So what are some of I like to?
I think it’s most helpful for our listeners if you walk us through a particular use case, I think many listeners may say yes, you know this resonates.
We do have an issue with mortality morbidity relating to maternal health, but we don’t really know what to do.
What are the different data sources you’re trying to reconcile and what are some of the challenges in trying to have actionable real time insights to improve?
Maternal health.
Off.
Kendra Folh 1:44
Well, I think what the current state is and we think of what are our data sources and I think you know, Data’s A1280 kg, like we say, a clinical side, I’m an L&D nurse by background.
So data is a newer place for us. As clinicians, we know we have our EMR which in Memorial Hermann we currently converted to EPIC.
So we have EMR data, we have administrative coding data which is really where a lot of our reporting is coming from.
So CMS is now really moving towards.
Our severe maternal morbidity.
Diagnostic codes.
That’s all claims data. There was just a recent article that came out by Popeue that said that really it’s only about 65% accurate.
Jordan Cooper 2:15
Mm hmm.
Kendra Folh 2:22
So how did we take the ICD 10 code, which is really where all the claims data is coming from?
The billing data.
The reportable data is coming from, but then how do we use our epic to give us reports?
That’s gonna give us some good information and mix that together within the state of Texas and is coming to theaters near a lot of us across the nation as levels of care. And that’s where we are gonna be designated as.
Organization at what level of care and what is the acuity level of moms and neonatal babies?
Even the NICU that you can take care of that requires a significant amount of case review.
So that’s another, you know, data source that is based on real obstruction, which I think as clinicians that’s our preference.
We really want to dig into that case.
We want to understand all the variables.
What happened?
So how do we take the administrative data, the EMR data?
And the abstracted data and create one source of truth that we can really guide.
Our teams to move improvement.
So that’s the work that we’re doing right now.
It’s really trying to pull that together to create one source.
Jordan Cooper 3:31
Yeah. And I understand that it’s been difficult to get clinicians to buy into that administrative coding data and to get them to trust the claims data because there may be a lack of trust in the validity of that payer data.
How are you overcoming that kind of skepticism on the part of clinicians?
Kendra Folh 3:50
We have to know our data and I think that’s, you know, whether you’re on the clinical side or you’re on the on the tech side is we’ve really got to review it.
We’ve got to get dirty with it.
You got to get your hands in the data.
So what we have found, we are seeing the same thing within our administrative data that’s also being reported.
Our validity is a little bit better as far as reliability, but I mean you’re really looking at about 80% of that because the administrative data.
Has a tendency to overcode history of, so that’s what we’re seeing is when we’re getting triggered from mom that had a very severe complication.
Based on a coding we’re looking back and realizing she has a history of some cardiomyopathy or a history of stroke and then that pops up as a trigger and that’s what we’re publicly reporting and that’s where we lose the trust of our clinicians and our program managers and.
Our nursing abstractors at the site saying, hey, our data’s not right. So we’ve got to take those triggers. We’ve got to abstract and really understand it.
And then we have to visualize it, make it digestible and bring it back to the teams into a product that they can.
Actually mute.
So very labor intensive.
Jordan Cooper 4:56
So.
Kendra Folh 4:57
But to have the move forward to using our EMR to do some of that cleaning abstraction for us.
Jordan Cooper 5:01
So as all of our listeners will know, there are two main buckets of analytic of reporting.
There’s analytical and operational. The analytical often people associate with.
Historical chart reviews and claims reviews for something that happened potentially months ago and operational something that happens real time is actionable at the point of care.
Have you any experience working with fire or is that on your road map at all?
Kendra Folh 5:29
Well, we definitely have.
And so I think we we use fire more from an operational efficiency perspective where we’re thinking, OK, hey, where do we have barriers for discharge?
So how do we build that into our Emrs to flag a a mom that has already shown me some signs or changes in vital signs?
Or maybe there’s some delays that are already occurring so that way we can mitigate that quickly. Just really to ensure efficiency decrease our length of stay.
But we’re really wanting to do is how do we actually use fire and look at safe postpartum hemorrhage risk assessment.
How do we know a mom truly is at risk for hemorrhage and utilizing that technology to alert these teams to increase the awareness to really monitor that mom closely?
Jordan Cooper 6:13
Got it.
So I think that there’s a if there’s a real time risk of hemorrhage, you’re looking at using these reports to prevent bad outcomes and the target.
You can’t monitor every single woman who delivers for being at high risk for a postpartum hemorrhage, but you’re kind of identifying the ideal target population where the care gap may be.
Have you had any success in reducing the rates of postpartum hemorrhage because of these risk assessments and the analytics?
Provided by fire in real time.
Kendra Folh 6:46
Not yet, but we are looking at is really the maternal early warning and so that’s the analysis that we’re doing right now or what are the signs, the vital signs that a mom has that really has subtle changes that are increasing her risk for morbidity, what we’re looking.
At right now is. What is that impact on ICU? So how soon can we identify a change in a vital sign, whether it’s a mom’s heart rate, her mean arterial blood pressure, or even just some of the hypertension disease?
What are those warnings that are going to prevent her from going to the ICU for early?
OK.
Jordan Cooper 7:17
Hi Kendra I think you mentioned that you are A6 Sigma black belt and there have been different reports that have offer conflicting insights and you’ve been working to reconcile those.
Kendra Folh 7:21
Yeah.
Jordan Cooper 7:28
Would you mind elaborating upon your efforts there?
Kendra Folh 7:32
You know, we’ll just stay with epic.
I mean, there’s some really great product there and there’s some good reporting. I think what happens is if we have so much data that’s coming out and the definitions conflict, So what populations are we actually looking at?
How are these reports written?
But how are we gonna use them? That is our biggest struggle is not only just having the consistency and the cleanness of the data, but having to use, for example.
With severe maternal morbidity, which is a very important measure and definition for those severe complications.
It’s her mom’s.
There are three definitions that we have to report and have to educate our teams on.
So you have three different reports coming out with three different numerators, 3 different denominators, different definitions.
What do you do with that?
What do you act on that?
So it’s taking a step back, aligning that up strategically to create one system of truth that our teams can act on.
Jordan Cooper 8:25
Got it.
So it sounds like there’s an issue that we covered at the beginning of this discussion about whether data is clean or dirty, but it’s also possible to have fully accurate data in three different reports that conflict with each other, because even though the data is accurate, the defin.
‘S conflicting aren’t standardized.
And so that presents a challenge for healthcare providers who are looking to have one golden record and source of truth.
Kendra Folh 8:53
Correct. And then if it was April 30 and then if you’re having reporting issues or workflow entry issues or the documentation’s not done in the right place, it’s not gonna feed your report.
Jordan Cooper 8:53
Have you been?
Kendra Folh 9:04
So how are we’re documenting the workload is so essential in the success and the clean data process.
Jordan Cooper 9:12
Is there?
Can you walk us through what the what kind of interventions you’ve been working on to improve documentation workflow at Memorial Hermann?
Kendra Folh 9:25
So as a six figure black belt, one of the projects we had as we did our epic conversion is we’ve really had to analyze our data all across the spectrum.
What are the reports doing process mapping?
Understanding the data going, we call go to Gimba which is going and working with those clinicians at the frontline to look at the workflow.
How are they documenting?
Where are they doing it?
What makes sense to workflow and how does that documentation?
Feed into the reporting.
The revenue.
The charge drops and all of the output that are required.
What we’re finding is a significant amount of variation there.
So we’ve had to really do some work to standardize documentation.
Standardized template.
Standardized order sets and workflows and then kind of reconfigure how we’re working in the application, then going back to the reporting and analytics to to redesign and reprogram what those workflows are.
Jordan Cooper 10:23
Many of our listeners may be thinking, you know, we’ve tried to implement standardized checklists and standardized procedures, but there’s been a lot of pushback from different kinds of clinicians.
The one physician says, well, this is the way we did it in residency and the other one does it differently.
And then this nurse was trained on this for at that institution and then left and came to this institution and wants to do it that way because it worked over there. How do you handle?
Cultural change as you’re trying to make sure that you’re following standard processes to improve the quality of data.
And therefore improve patient health with actionable real time insights.
Kendra Folh 10:59
We would be very intentional with this as an organization with almost 32,000 deliveries, we have 10 campuses and I always think of us as a star track enterprise starships.
Everybody’s got a different, different kind of ship.
They’re running where a plurist model.
So we’ve got a global 4 academic facility.
We’ve got quite a physician.
We have employee positions, so we really had to create an organizational structure and governance. That way we were collaboratively working together. So we have what we call a perinatal quality collaborative within the organization we have.
We meet every month.
Where we bring all of this together, we have work groups and so stakeholder analysis is key.
We’ve got to have the right folks at the table and leading conversation and collaborating, so we’re not wanting to be authoritative.
We as the hospital organization, I’m not managing physician practice, but we’ve got to come to some consensus. And so it’s working to consensus in the best evidence based practice.
And taking that to the right governance structures for approval.
Once those are approved, then we put those in place for standardizing them and then doing that education, Pete as well as auditing. We’ve got to audit.
We’ve got to bring that back to our quality and assurance performance improvement committees or the quality committees to make sure that we are managing that standard.
Jordan Cooper 12:12
As you mentioned Kendra, at the start of this discussion, you are the Chair of the data committee at the state level in Texas. And I think you have had some experience with reluctancy of different health systems to share data.
At the, either with the state or with each other, would you care to elaborate on?
Obviously, some patients are getting care well system and moving to another health system.
Different. How do you how do you get kind of an accurate record of 1 mother coming in to deliver?
And how?
The different data sharing issues playing into that.
Kendra Folh 12:45
One, I think we’re just not. I think we’re not getting a good story of moms working across the state.
I think we’re also struggling.
I think Epic truly with the my chart have been the best interoperability that we’ve been able to find so far. I think even just within Houston and some of the work we’re doing, to your point, women go to different organizations all the time.
I’m going to go to what’s ever easier.
I’m going to go to the Ed.
That’s fastest.
Not necessarily the same.
It’s changing that with, you know whether it’s HIPAA.
Rules, whether it’s organizational, reluctancy and protectivity of client data as well as nobody wants to to to all the things nobody wants to have their information used against them.
So even at the state level with the state is reluctant to have their data publicized.
I know there was an article that just recently came out with Pro Publica that there’s just this reluctance to be being seen in a bad light.
So everybody’s very protective.
Of the information and again adding in PIPA, well it just adds a little bit more difficulty in exchange of information.
Jordan Cooper 13:52
So there’s political barriers, economic barriers.
To share to getting interoperability across health systems.
And clinicians are just trying to provide care and it’s difficult to exchange data in a way that is in the best interest of the patient is what I’m hearing.
Kendra Folh 14:10
Correct. Absolutely.
Jordan Cooper 14:12
Now you mentioned that automating data collection reporting would potentially add value and potentially if you could automate data collection, maybe you could overcome some of those interoperable.
Ability challenges. Have you had any experience working on the automation of data collection and reporting?
Kendra Folh 14:30
Well, I think we’re trying to I think just take them within my own organization.
That’s exactly what we’re trying to do when we’re looking at how do we automate data collection and abstraction just for all of our state recording is that’s where we’re really trying to do this at a pilot level internally.
How do we utilize the EMR with the claims data to audit?
We’re building something we’re calling the motherboard.
So how do we take this motherboard feeding automatically and doing some minimal abstraction to complete the story?
And then we have a final.
Product. But we’re also looking at how do we do that at the state level?
Now California has done that well in this in this population where you like in the Emrs are automatically reporting. There’s some automation particularly with epic, with the Vermont Oxford Network and the neonatal population.
So I think there’s pockets of it. The biggest opportunity in my perspective is how do we take what the piloting that has been done and these piloting pilot in pockets.
And how do we learn from those?
To see how can we fill that up.
Jordan Cooper 15:32
Got it.
Yeah. The biggest opportunity is to scale little instances where it worked and see if you spread it across the organization and that’s when you might use your position at the chair of the data committee at the state level or the difference governance committees at each individual healthcare deli.
System.
We’ve covered a lot of different topics, kind of approaching the end of this podcast episode.
We’ve been moving quickly. We covered.
Challenges with data being clean.
Challenges with providers trusting the data challenges of trying to get the data in real time so it provides actionable insights and value to clinicians and patients when they’re interacting with each other in an encounter.
Challenges with try to reconcile different reports and then challenges and interoperability and automation for our listeners who are also experiencing some of these same challenges. What?
Is one word of advice. If there you would potentially even give to yourself.
A year ago or two years ago.
What do you wish that you knew then that you know now that may help some of our listeners as they’re beginning to go down the journey that you have been on.
Kendra Folh 16:43
The most important thing I have learned so far, whether it’s in my own organization and at the state level, if alignment is true.
So we’ve got to line up with Cnf and the external reporting measures and we’re financially impact us. We have to start from the top.
That is how we get the resources and the time to do this work and to get this data clean.
It’s very important. And So what we.
Have found the most valuable in these conversations is where we gonna make impact.
Start very simple.
What are we trying to do?
What’s our objective?
Let’s focus on mortality.
Let’s focus on morbidity decrease like the state.
How do we get the data to get there?
What does the story we need and what is the financial impact and that was the biggest piece is when we lined that all the way up with our supplemental funding, our CMS, our IQR measures.
Well, when you paint that picture, there’s a lot of monetary impact here.
Not as well as just the impact of.
The the lives of these moms and these babies.
And that was how we were able to get the resources and the time to do this at a meaningful level.
Jordan Cooper 17:52
Got this. So if you present a financial argument of return on investment to C-Suite executives, then there may be an opportunity to allocate resources in time in order to invest in these interventions to improve the quality of data where it’s needed when it’s needed, how it’s needed.
Kendra Folh 18:10
And not just the speechwrit executive.
Also to our legislature.
Jordan Cooper 18:13
And legislative as well.
All right. Well, for our listeners, this has been Kendra folh, the program director of Women’s and children’s servicen for Memorial Hermann Kendra. I’d like to thank you so much for joining us today.
Kendra Folh 18:26
Thank you so much for having me.
Jordan Cooper stopped transcription
