Adding Value to the EMR: A Clinical Perspective (Webinar)


Adding Value to the EMR: A Clinical Perspective (Transcript)

Welcome to an ADVANCE Informational Webinar

[Danielle]

Hello everyone and welcome to our ADVANCE Webinar sponsored by Health Catalyst, Add Value to Your EMR.

About the Speaker

And now we would like to tell you a little bit about today’s speaker, Dr. Charles Macias. He’s a recognized authority on leading and achieving large-­‐scale healthcare improvement initiatives using data and analytics to drive positive change. Today, he will share insights related to Texas Children’s Hospital’s (TCH) focus on quality. TCH was recently awarded the CHIME-­‐AHA Transformational Leadership award for its excellence in developing and deploying transformational information technology that improves the delivery care and streamlines administrative services.

Today, he will address common challenges and offer proven solutions related to capitated or value-­‐based care reimbursements. With a unique clinical perspective, he implores clinicians and administrators to foster cost/care tradeoff conversations. While this is a topic that some of you may be nervous to address, he will reveal why such conversations are necessary for quality patient care and critical from a population health perspective.

Contrary to common belief, Dr. Macias will also speak to evidence that in today’s existing infrastructures, there is a low correlation between the best care and cost. Improving quality of care with emphasis on efficiency and effectiveness will most often decrease the health system’s overall costs.

Underlying all of TCH’s success is a focus on data warehousing and analytics, along with organizational strategies that drive transformed data into the hands of clinicians and administrators. This ultimately helps them identify improvement opportunities that result in “optimum care” while eliminating waste.

Once again, I would like to remind you that Health Catalyst is our sponsor for today’s webinar.

Now, I am going to turn the screen over to Dr. Macias so he can start the presentation.  Thank you again for attending today.

Adding Value to the EMR: A Clinical Perspective

[Dr. Charles G. Macias]

Thank you, Danielle. It’s a pleasure to be able to speak to this group, specifically about an exploration of how Texas Children’s has looked at value and how we have considered the EMR within a clinical perspective. And I can’t emphasize this point enough that this has really taken the view of the clinician as we describe value.

Poll Question #1

So Danielle, if you can open the poll, I’d just like to begin with a poll for the audience to get a better understanding of who we are as a group here today. So I’d like to ask, what is your primary area of focus?  Are you a physician/clinical care provider?  Are you engaged in quality activities within your job? Are you engaged in information systems and information systems development?  Are you engaged in finance? Are you an administrative executive? Or do you fall into none of these categories?

Poll Results

[Danielle]

15% are physicians or clinical care providers, 23% are quality, 38% information systems, 5% finance, 18% admin executive.

[Dr. Charles G. Macias]

So this is a wonderful combination and illustration of specifically about why this is so important, why it touches so many lives and so much of what is embedded within a healthcare infrastructure for a group of people that are dedicated to one mission – that is improving the quality of care of the patients that we see.

Objectives

So in terms of objectives for the remaining time that I have with you, I’m really hoping to describe the power of pairing an Electronic Data Warehouse (EDW) with an EMR in order to realize care improvement and to look at subsequent waste reduction and cost savings.

We want to understand early results of Texas Children’s Hospital’s (TCH) cultural shift to focus on value and understand that link between quality and cost.

And finally we’ll discuss how Texas Children’s Hospital’s (TCH) focus on linking clinical science and payment models, as well as operation science have driven financial stewardship and a lot of our earlier success and continued successes in population health management.

The Healthcare Value Equation

So I think it’s helpful to consider this within the framework of value.  We are in a very changing environment where it has become important to us at healthcare infrastructures to really put a cost on value to understand what value is about.

I like referring to the Porter equation for value where value is equal to quality over cost. I think this is important in an environment where cost is marginally increasing, healthcare must markedly improve quality in order to improve value. So the role of EMRs then becomes absolutely critical. The adoption of the EMRs in clinical systems should help to push the quality agenda. But it’s important to recognize that EMRs alone may not be sufficient to deliver data intelligence, to really deliver data in a transformed meaningful way that will improve value. And that’s the exploration, that’s our framework for today.

Quality?

Now, it’s also challenging to drill down to the definition of quality, but I think that there are frameworks out there that help us to understand how we might define quality. The Institute of Medicine in Crossing The Quality Chasm has described six domains where quality can be seen as care that’s safe, avoiding injuries to patients from care intended to help them; that is equitable but it doesn’t vary because of personal characteristics provided by the patient towards family; that is patient-­‐centered, that we provide care that is respectful to and responsive to individual patient preferences, needs, and values; that it’s timely, it reduces waits and potentially harmful delays; that it’s efficient, avoids waste of equipment, supplies, ideas, and energy, and that it’s effective, that services are based on scientific knowledge to all who could benefit and it accomplishes what it sets out to accomplish.  The recent additions of access to care and care coordination help provide a lot more granularity to these definitions of quality.

But I really want to call your attention to this component here, that care is efficient, that it avoids waste of equipment, supplies, ideas, and energy – because when we begin to think of cost, when we begin to think of value, within the definition of quality, we already see the underpinning from where finances link on a clinical level.  And we’re very grateful that the IOM has given us these domains to consider.

Best Practices Do Exist

But I think its best practices have been well described in the literature, as well as from our personal experiences. So Best Care at Lower Cost, the IOM 2013 report, has interestingly touched upon this issue. There are best care examples that come from communities, not policymakers, and they inevitably involve patients, doctors, nurses, and other providers working together, a fundamental tenant of quality improvement, that we work in teams, that we work in collaborative and only in that way can we see improvement in quality.  And this is, I’m thankful to the words of Donald Berwick, former administrator of the CMS during the session entitled, “Controlling healthcare costs while improving quality.”  He mentioned specifically a healthcare project in Alaska, where team-­‐based care has resulted in 50% fewer hospital bed days and 53% fewer ED admissions and 65% fewer specialty visits.

By one estimate, roughly 75,000 deaths might have been averted in 2005 if every state had delivered care at the quality level of the best performing state, an astounding number when we look at healthcare across the US.

To really understand that relationship and begin to explore what the total cost is, a lower cost or a higher cost versus a lower quality or a higher quality, we can turn to some examples of hospitals in southwestern Pennsylvania who were paid an average of $18,000 to perform heart bypass surgeries, where others were paid as much as $35,000 for the same procedure.  Wide variations in charges. Similarly, patients for heart valve surgery ranged from a low of $24,000 to a high of $54,000. But what’s most important about this is that the lowest-­‐priced hospitals had lower mortality and readmission rates (i.e., better quality) than the highest-­‐priced hospitals.  Really shaking our notion that we have to be investing in huge dollars in order to obtain higher quality.

Poll Question #2

[Dr. Charles G. Macias]

So I want to stop and go up. I’ll turn this back over to you, Danielle, for poll question #2. We want to get a sense for the audience, how concerned are you about realizing return on your investment from your electronic medical record investment?  Are you A – very concerned, B – somewhat concerned, C – neutral, D – slightly concerned, or E – not concerned at all?

Poll Results

[Danielle]

44% very concerned, 31% somewhat, 18% neutral, 5% slightly, and 3% not concerned.

[Dr. Charles G. Macias]

So, you’ll notice that the vast majority of our audience today is neutral to very concerned. Very few people are not concerned about realizing the return on the investment from the EMR investment.  I think that really speaks volumes to the importance of that conversation.

ROI on EHRs Proves Difficult

So return on investment on electronic health records proves very difficult.  In second look, few savings from digital health records.  The RAND report forecasted an $81 billion annual U.S. savings. However, however, that’s 2005. “Seven years later, the empirical data on the technology’s impact on healthcare efficiency and safety are mixed,” and I’m quoting, “And the annual healthcare expenditures in the United States have grown by $800 billion.”

So what were the reasons for the disappointing performance on health IT to date? They attributed to several factors.  One, sluggish adoption of health IT systems, coupled with the choice of systems that are neither interoperable nor easy to use.  Very important to clinicians, like myself, who really depend on the systems that exist within our entire hospital to provide the data that we need to deliver care. The second, the failure of healthcare providers and institutions to reengineer care processes to reap the full benefits of health IT. Now, this takes the element, the interest, the importance to a much higher level, not about the individual provider but understanding what our systems are doing in order to improve quality of care – how have we changed our operations, our processes, our data management strategies now that we’ve instituted the EMR.

Do physicians like where that’s taking us?  Well if you look at that last point, EHRs, red tape eroding physician job satisfaction, most physicians expressed deep frustration with costly and overly complicated EHRs that have fallen far short of the promise to improve practice efficiency. In fact, in one particular survey, and I’ve  selected one here that’s really at the lower end of this satisfaction, 20% of physicians wanted to return to the paper world. So there’s a real tension for the individual provider between fighting and to improve the EMR, what do I invest in my healthcare infrastructure in the time that is my own personal time versus simply accepting what I have and spending late nights catching up on data entry. A real challenge for the average clinician.

About Texas Children’s Hospital

So let me tell you about our response. Let me explain a little bit about Texas Children’s and how we’ve related the patient experience, healthcare expenditures, the EMR, and the enterprise data warehouse.

Texas Children’s is Houston-­‐based and nationally renowned for providing top pediatric and women’s care. We provide a full continuum of services with just about every pediatric subspecialty known at several sites within our umbrellas.  We’re committed to developing clinical effectiveness guidelines to deliver the highest quality care possible and we do this in a systematic fashion.  In order to get some sense of generalized ability, how do we fair, how do we compare against your institution? We have 469 beds: our annual inpatient admissions, over 21,000; annual outpatient visits over 1.4 million; emergency room visits 82,000 a year and climbing; inpatient surgeries 8,600; outpatient surgeries 14,000. We’re a very large children’s hospital.

Pareto 80/20 Principle in Healthcare

And as we began to look at our different care processes, as we began to understand where are the opportunities to improve quality, to look at these gaps in quality either because there’s a high prevalence of disease, there’s a high morbidity and mortality, because it’s costly and there’s large variability in providing that care amongst the providers, we needed a more systematic approach to this. We know from data  management and quality improvement that Pareto charts provide an ideal opportunity.  So using our Enterprise Data Warehouse, we’re able to look at a small percentage, and you see here the number of care processes, 10 care processes, that account for 53.2% of our variable cost. Known throughout the healthcare world is the Pareto 80/20 rule, where 20% of the population may consume 80% of the cost or 80% of the opportunities are really provided in 20% of the changes.

Asthma

So let me give you an example that’s far more specific than that. As we identified through a key process analysis, an analysis that allowed us to look at numbers through our enterprise data warehouse and looked at where those gaps were, now relates that clinically to say, of the disease processes where we may have opportunities for bridging the gaps in quality, where do we stand the greatest opportunities, and asthma emerged within those analysis.

Asthma on a national level affects about 7 million children in the US, about 80,000 children in Houston. It’s the most common chronic disease of childhood. Acute asthma accounted for about 3,000 emergency department visits in 2011 here at Texas Children’s and about 800 hospital admissions, so clearly a large target. And despite the fact that there have been national asthma practice guidelines for the National Asthma Education and Prevention Program since 1991, most recently updated in 2007, hospitalizations and ED visits have not markedly decreased.

Severity Adjusted Variation

So we had our burning platforms.  We could drill down further to look at the care that we provided here at Texas Children’s and now look at severity of the illness across four levels.

What you see here in this illustration through our output is our APR-­‐DRG severity at four levels, with level 1 being in blue, level 2 being in red, level 3 being in green, and level 4 being in yellow. Each of the bubbles is representing different providers and you can see the variable direct cost for some of our outliers. Why it is that the same level of severity, so take the bubble on the highest, snap that at the lowest left side on the X axis on the yellow bubbles, same level of severity. Those providers are providing care that is much cheaper than $45,000 in same level of severity. So we can begin to understand where variability, unwanted variability, in care is not attributable to the severity of the disease but instead to practice patterns of the provider.

Correlation Between Cost and High Quality Care is Low

This really helped us to understand what that relationship was between quality and cost and what we should be demanding.  And in fact, if we look at the pediatric literature, this particular study published in the Journal of Pediatrics in this year looking at data from the pediatric health information system database.  This is data from 21 member hospitals that were included that followed two quality of care metrics for each of three disease processes.  This was gastroenteritis, asthma, and simple febrile seizures. What was noted were huge variations in practice, that increased costs were not associated with lower admission rates or with 3-­‐day ED revisit rates. The implication?  That optimal care could be delivered at a lower cost than today’s care, that we’re simply not seeing a demand to expend a higher dollar amount as an expectation to see better quality.

Consumer Care/Cost Uncertainty

So what does that mean for the consumers? Well, consider the framework of a consumer – patient – family. They trust their physicians, they hope for the best, they struggle to understand cost and care, they don’t often know what they are getting, and they don’t always get great outcomes.  But what they do expect is value.

They expect that the investment that they make in their healthcare is gonna return a level of trust and a level of outcome that’s important to them.

Challenge of Healthcare

So the challenge then for the healthcare providers, for the physicians specifically, is to understand how to bridge that gap. So consider where the physician’s direction is from.  The physicians are often driven by science and key values.  They can be overwhelmed with medical literature, anywhere from 400,000 to 600,000 journal articles published each year. How do they assimilate all that? They’re not well trained to turn that experience into high quality patient outcomes.  We teach a traditional model within our medical training systems that really doesn’t address a lot of quality. We’re really only beginning to see a change in that. So where we’re left today is helping the physicians with the key values, with an attention to science to provide them transparency of local data so that they understand how to drive the change patterns, their care patterns.

Poll Question #3

So at this point, I’ll turn this back over to Danielle for poll question #3, for non-­‐clinical attendees or non-­‐ practicing physicians in attendance, during what percentage of patient visits are your physicians talking about cost and care tradeoffs? Is it A – 80-­‐100%, B – 60-­‐79%, C – 40-­‐59%, D – 20-­‐39%, E, 00-­‐90%.

And, Danielle, when you close the poll, if you could report those results.

Poll Results

Poll Question #4

So  for  practicing  physicians  in  attendance,  during  what  percentage  of  patient  visits  are  physicians  in  your organization talking about cost and care tradeoffs?  So A -­‐  is that 80-­‐100%, B -­‐  is that 60-­‐79%, C – is that 40-­‐ 59%, D – is that 20-­‐39%, and E – is that 00-­‐19%?

Poll Question #4

And I see that our poll is open.

Poll Results

Danielle:          And Dr. Macias, 38% said 0-­‐19% of the time, 41% said 20-­‐39% of the time, nobody clicked 40-­‐ 59%, 13% are at 60-­‐79% of the time, and just 9% are 80-­‐100% of the time.

Dr. Macias:    So this is of clinical providers. Danielle:          Yes.

Dr. Macias:    Now, for poll question 3…

Danielle:       Right. I can actually go back and share that with you again if you would like.

Poll Results (Non-­‐clinicians)

[Danielle]

For non-­‐clinicians 38% are 0-­‐19%, 23% at 20-­‐39%, 15% at 40-­‐59%, 9% at 60-­‐79%, but 15% of the non-­‐clinical providers talk about it 80-­‐100% of the time versus the 9% for the other group. So there’s a slight difference there.

[Dr. Charles G. Macias]

So I think what that illustrates is exactly my point that for the clinical providers, this isn’t typically an area that we feel comfortable with. We just don’t have those tough care tradeoff conversations with our patients.  For administrators or non-­‐clinical providers, I think there’s an unreal expectation or an unreal assessment that in fact it’s happening a lot more than they think its happening.  And I think this is part of the challenge with the culture of understanding cost and value.

Physicians and Care Cost:  Clinical Decision

So as you can see in this particular slide, when we talk about physicians and care cost, when we understand where resources, where value, where costs fit into clinical decision making, clinical decision making is much more than what we would often hope is a big chunk of it, which is the evidence. It’s not simply saying what does the science tell us, but in fact that we must understand the science as a clinical care provider, we must understand the patient’s values and preferences, we must understand our colleague’s values and preferences in order to make a decision about what that diagnostic testing or management strategy must be, but we must understand it within the context of resource issues – what are the resources within our infrastructure, what are the resources available to the patient.

Physicians and Care Cost

Now, this whole concept of clinical expertise is this particular model becomes much more evident for how it should link into value and cost. That’s part of that decision.

The New Healthcare

So although it has been taboo in the past for physicians to take cost into consideration, I would make the point that without money, there is no mission, there is no expansion, there is no innovation, there is no healthcare, that we can’t move into a new era of healthcare without changing our cultures so that providers understand the improvement that can be created, they can understand the story that their data tells, and that they can understand this within the context of what it takes to administratively run their systems of care, that data linked to systems of care can drive these quality initiatives because it puts all levels equal. We understand where science merges with data transformation merges with process changes.

TCH’s Clinical Integration Strategy

So here at Texas Children’s, we took the theoretical model and integrated it into a system that we call our clinical systems integration strategy. Our chart was to build a comprehensive, integrated and evidence-­‐based quality and safety program that would result in measurable improvements in process and quality care, that we would collect and meaningfully, and this is a very important part of this, meaningfully transform and use that data to provide information about the clinical operations linked to operational processes, that we would implement an enterprises-­‐wide data management infrastructure that would leverage our current clinical systems; starting with Epic, and financial information in order to provide easy-­‐to-­‐access, meaningful and relevant data to assist in accelerating our improvements, strategies we already have in place for improving quality, simply taking them to lightning speed in both clinical and operational processes.

TCH’s EDW Architecture

How did we do this? While working with our consultants with Health Catalyst, we were able to look at all our financial sources, our administrative data sources, our EMR sources, departmental sources, patient satisfaction sources, our human resources, and link these within our Metadata bank or Enterprise Data Warehouse, taking a systematic approach to developing subject area marts within both the clinical processes and the operations processes. So the part most concerning to me on a day-­‐to-­‐day basis was, of course, the clinical realm of it.  So for us, that meant asthma, appendectomies, deliveries at the Pavilion for Women, pneumonia, diabetes, and surgery, to begin with. You can see our operations changes on the other end. But what this is really meant to illustrate is that we take different silos of data, we standardize the terminology behind them, the definitions, the cohort definition, populate them into subject area marts tailored against a balanced scorecard of metrics that are meaningful for the content expert that developed them, and then transform them, as you can see here below, into our visualization.

How TCH Defines Quality

So this allowed us then to link to quality. So how does TCH define quality?  Really in four strategies.  So the first is to look at quality within our balanced scorecards, within our approaches, using the Institute of Medicine’s quality domains. The second, by recognizing the importance of minimizing unintended, unwanted variation in the healthcare delivery system – one, the healthcare system that we were so integrally involved on.  The third, understanding a definition, one put out by Taylor earlier in a look at Medicare back in the early 1990s which described quality as the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge. Ergo, number 4, looking at systematic infusion of evidence as the best current professional knowledge that integrates operational improvement and data transformation. Critical because unless we could provide the data transformations, we could never understand whether we were truly improving desired health outcomes here with the clinical relevance of how all of that fit.

Approach to Improving Processes of Care

So our approach to improving the process of care was to organize permanent, integrated workgroup teams. They’re kinds of teams that Don Berwick described, the collaborative teams, the teams that would work from the bottoms up level to improve clinical programs or clinical services over a long term. We would integrate critical elements of evidence-­‐based practice into the delivery of care. This is the science of best known practices from managing disease processes to get rid of the unwanted variation and integrating the best known outcomes possible, and establishing baseline measures, AIM statements, measurable goals and ongoing review of results versus targets with rapid cycle process improvement.

Now, it was informed not by care process teams that had data that they had to wait for 6 to 9 months, but real-­‐time click and point data or as real time as possible, 23 hours and 59 minutes for our downloads in order to see our visualizations provided directly to clinicians.

Quality & Clinical Evidence-­‐Based Team

So the structure that we created was this metaphor that we called a flatbed truck, where each of our care processes represents a different disease process. So asthma, for example, diabetic ketoacidosis, pneumonia. All of this is lead by representation of our collaborative teams, our physicians, our clinical directors, our operations directors, and all of it is structured with the support of an evidence-­‐based flatbed, that’s clinical care guidelines systematically developed using knowledge managers to drive the team’s activities, data architects that could converse with Epic, as well as provide visualization and infrastructure and help us to integrate elements of Epic that were necessary that might not have been included in our quality improvement rapid cycle process strategies.

Data Drives Waste Reduction

So it was through this that we understood the data could drive quite a bit of our waste reduction.

Alternative Approaches to Waste Reduction

Option 1: Focus on Outliers – the prescriptive approach

So think of data waste reduction in the way that we did. And I’m going to provide you two approaches to waste reduction – one in which we see a normal distribution with excellent outcomes on the left, poor outcomes on the right; where a traditional option of focusing on the outliers, that’s the quality improvement strategy, where we look at those poor outcomes, and we take a prescriptive approach. We identify the extreme cases with the potential for high cost from bad outcomes and we eliminate that tail.

Alternative Approaches to Waste Reduction

Option 1: Focus on Outliers – the prescriptive approach

And now what we have is a normal distribution again with the tail gone. We’ve gotten rid of poor outcomes. The problem with that is if we set that level at the 1.96 standard deviation, we’ve only eliminated 2.5% of our waste.

Alternative Approaches to Waste Reduction

Option 2:  Focus on Inliers – improving quality outcomes across the majority

In contrast to that approach, we can look at a waste reduction approach that recognizes the same starting point of normal distribution but instead now, we focus our teams on the inliers.  We’re improving quality outcomes across the majority of the care processes. We’re identifying best practices either through research, that’s through evidence, through our integrated evidence-­‐based guidelines, as well as analytics, understanding the outcomes of our prior rapid cycle process improvement, understanding the outcomes of where we have been before, our baseline, and we develop guidelines and protocols to reduce inlier variation that we now tightened the curve.

Alternative Approaches to Waste Reduction

Option 2:  Focus on Inliers – improving quality outcomes across the majority

And instead of focusing on the poor outcomes, we maximized our distribution within the excellent outcomes. We shifted the cases that lie above the means toward the excellent end of the spectrum, providing a hugely more significant impact and a strategy that needs to be prescriptive at that 2.5% of bad cases.

Improving Cost Structure Through Waste Reduction

So we can think of three buckets of waste that we then focus on decreasing.

Improving Cost of Structure Through Waste Reduction: Ordering Waste

We have ordering waste. This is ordering the tests that are neither diagnostic nor contributory.

Care Redesign Methodology

So in our approach, we can look at our care design methodology, where evidence supports, evidence is equivocal for or evidence is against, where evidence there is either data from my analytics or evidence in the literature.  And so, as an example, quicker steroid delivery for asthmatics leads to better outcomes.  There is clear evidence that supports that.

Evidence Equivocal – the role of hypertonic saline and bronchodilators in select patients with bronchiolitis.  I just don’t know its efficacy. Literature has not described that.

Evidence against – chest X-­‐ray utilization in patients with known asthma or steroids in bronchiolitis. Just doesn’t improve outcomes.

Now we can understand this bucket.

Cost Per Case and Case Volumes

And when we understand that this may be one or two cases but we compile that against time and we look at the number of services that provides it through our output, we understand that adding all those average cost per case means a huge amount of waste.

Asthma: Care Process Teams Cohort, Percentage of Chest X-­‐rays Ordered (Oct. 2010 – Apr. 2013)

So we begin looking at strategies then where we use our Enterprise Data Warehouse and taking the examples of evidence against chest X-­‐ray ordering for asthmatics. We can drive rapid cycle process improvement. And what you see in this statistical process control chart is a widely variable performance with a huge level of chest X-­‐ray ordering, over 65%. But with rapid cycle process improvement, we were able to drive those numbers to 35% where this was taken now closer to 20%. So taking us from some of the highest ordering levels to some  of the lowest in the country amongst children’s hospitals.  This is the way that we could reduce that bucket of ordering waste.

Improving Cost Structure Through Waste Reduction: Workflow Waste

For workflow waste, we could look at variations in our workflow to understand how we could improve wait time.

Flow chart of a patient with acute gastroenteritis through the TCH Emergency Department: Existing Process

So here’s an example of looking at the management of acute gastroenteritis where we process map. We understand the patient goes through all these multiple points, and I bring your attention to all the elements that the patient takes on – patient waiting, patient put in ED room, patient evaluated by nurse, by medical student, by ED residents, by fellow, and care doesn’t begin until the attending of that patient.

Flow chart of a patient with acute gastroenteritis through the TCH Emergency Department

But linking evidence-­‐based practice to this, integrating those strategies within our EMR, using our data to understand rapid cycle process improvement, our map looks very different.  We initiate care at the triage point early. So by the time the attending physician sees the patient, that’s really a decision about whether it’s fine to discharge, markedly decreasing our (33:42) of time and markedly decreasing or left without being (33:46).

Improving Cost Structure Through Waste Reduction: Defect Waste

The third bucket of waste that we strived to decrease is the defect waste. This is the air waste. This is the waste that occurs from the unwanted effects of our care.

Clinical Decision Support to Minimize Errors

So where this becomes clearly important is clinical decision support to minimize errors, perhaps best described with medication errors, huge bucket for safety issues.  We can streamline our processes and improve those processes by understanding the outcomes and seeing the data that results from interventions to improve management bundles for central line bloodstream infections or to look at other errors or other safety issues that add cost in care delivery models.

Shifting Quality Improvement Culture to Effectiveness and Efficiency

So we shifted quality improvement culture to one of effectiveness and efficiency. And we’ve done this because there’s a huge stewardship responsibility for our physicians.  I’ve mentioned before that it’s often taboo to talk about the financial component, to talk about cost, but I think when we’re faced with the 13% gross domestic product being devoted to our healthcare infrastructures, it is incumbent upon all clinical care providers to understand that they need to be good stewards of the dollars. And so, here at Texas Children’s, we were faced with a burning platform. Our state was moving from a cost-­‐based reimbursement model to a more capitated model of care reimbursement through APR-­‐DRGs. So then it became clear where the cash value of waste occurred.

APR-­‐DRG Transition Calculator

We’ve actually created these APR-­‐DRG transition calculators that help us to understand where in the pre-­‐APR-­‐ DRG world, we could look at the cost of delivery and care for patient with this particular APR-­‐DRG.  And here, what you’re seeing is a level 1 severity patient with asthma and how that shifts from the pre-­‐APR-­‐DRG to the post given the interventions that we’ve undertaken with quality improvement.   Now, we can look at a system level, add a dollar value to that and we can use our Enterprise Data Warehouse to understand where costs fit into our decisions.

It is never saying that cost should be the driver for all clinical decisions but quite the contrary. When we understand best practice output, when we understand the changes that we can undergo by improving the effectiveness and the efficiency of care, if we can add a dollar value to that, we can then understand the outcomes of those financial decisions and prioritize.  If all of the things are equal and we can improve that cost at the same that we’re improving patient outcomes, there’s much more than we can do for our mission and  for our vision when we are operating with better financial stewardship.

Registry Financial Score Card

So what you see before you is the output from our visualization for only one of our tabs out of many in our metrics, and this is one that’s made now visible to our providers.  They can actually see cost of care or charges for care, they can see payments that go well for that particular disease process.  And so now they understand the healthcare system a little bit better.

Examples of Demonstrating ROI

So if I could then distill all of our discussion today and say, if I had to look at a return on investment, start at this discussion with how do we add value to our EMR, what have we done with our Enterprise Data Warehouse. Well, we’ve improved clinical care. The most important outcomes that I as a clinical care provider could describe, we have really seen changes and improvement in both process metrics and outcome metrics. It’s been able to decrease length of stay in over 10 disease processes.  We’ve decreased readmission rate, or asthma for example. I demonstrated how we decreased unnecessary chest X-­‐ray utilization. This is really amounted to over $4 million in savings across several disease processes.

We’ve reduced waste by systemizing our reporting. We’ve estimated that our Enterprise Data Warehouse reports, that through our visualizations allow us to click and point report, cost about 70% less to build with the 50 hours that is previously used to take them to get our asthma reports of an Epic report writer.

And we’ve been able to improve labor productivity by using tools that allow global views for increased operational efficiency.

Population Management

This has taken us one step closer to where we view our real goal and that’s driving value across a system that results in a healthier population, a population management strategy, taking all elements of our systems of care here at Texas Children’s, our health plans, our pediatric hospital and subspecialty clinic, our pediatric practices, and our Pavilion for Women and understanding how the enterprise-­‐wide data management infrastructure can support, promote and drive better outcomes in all of those areas.

The Healthcare Value Equation

So if I take you back now to the healthcare value equation, I want to remind you what that looked like, where value equals quality over cost. Here at Texas Children’s, we recognized the over $100 million investment we’ve put into our EMR but we’ve also recognized the opportunities for linkages to clinical decision support that allowed using the Enterprise Data Warehouse to link to science, the operations, and the data management strategies now linked to drive but to drive at lightning speed our rapid cycle process improvement that used to take us months to do previously.  We do this understanding and driving the importance of financial stewardship with our providers, helping them to understand where better mission and better vision to take us and we’re driving value through a higher quality of care delivery.

Questions and Answers

And with that, I will leave you my contact information and open this up to questions.

Questions

Danielle:         Thank you, Dr. Macias. We do have a couple questions from the attendees.  So I’m just going to share them with you and the audience now.

First question, where can I focus my strength to contribute more in quality area? Currently, I am working as a quality and medical staff coordinator.

Dr. Macias:     So if I understand the question correctly, the question comes from a person who is working in a quality department, working also with medical staff. And the question really is how do I, and if I can over read the question, that’s really how do I change the outcome, how do I look at the opportunities I can provide to improve the outcomes that my providers may get. And please correct me at any time if I am misleading the question.  But I think most of that strength is  going to come from a framework that helps us to understand how we link data, how we link science, and how we link operations, or what we depend most on from our colleagues in the quality department is what we can do to either drive rapid cycle process improvement, lean improvement, whatever strategies are going to improve quality of outcomes but helping to link the clinicians with the science of it, helping them to see the data that can be made transparent from it, and then understand what next iterative improvement can be done. It’s really all about team building and communication and being transparent about the data that will improve those outcomes.

Danielle:         Okay.  And next question, what was the timeframe from the deployment of EDW to benefit realization?

Dr. Macias:     Excellent question.  So what is the timeframe from deployment of the EDW? So I’m going to start by qualifying the question and that is it’s not simply a matter of flipping a switch. Much of what we realize we had in existence of what – I’m gonna go back to that triad that we considered to be important in the value concept – that we must merge science, we must merge data transformation, and we must merge process improvement in a single strategy that helps drive the improvement in outcomes.

So at Texas Children’s we had already engaged in the development of evidence-­‐based guidelines that were systematically developed.  I will tell you that we have now since 2007 developed over 38 systematically derived comprehensive guidelines of care with over 150 evidence-­‐based order sets that are now integrated into our EMR. We were beginning to implement, I would have to say, at a snail’s pace within our evidence-­‐based outcomes center but at a much quicker pace through our quality and outcomes management group simply not linked to a data source that could provide the data more rapidly. So we were dependent on driving much of that care process improvement with data that we would either extract from national databases that might be 6 to 9 months delayed or that we would do by chart pulls, which was labor-­‐intensive, or from Epic which required quite a bit of time and effort from an Epic report writer and need I also say very costly to pull.

So as we implemented our Enterprise Data Warehouse about slightly over two years ago, we began to construct the governance, the relationships, defining the cohorts for the first two care processes, which were asthma and appendicitis, and within 6 months or so, we were already beginning to see some of the data and some of the output from the teams. We’re now at a point where at any given point and time, our care process team can manage anywhere from 3 to 6 rapid cycle process improvement strategies that may be at different stages, but getting those teams up and running does take a considerable amount of time and effort. But you have to consider you’re also looking at it within the confines of what your culture allows, what is  your deployment strategy, am I teaching advanced quality improvement methodologies, is their quality improvement there coming from our residents where vast knowledge foundation already exists or am I having to build that solely into the infrastructure?

So the answer to that question is it’s completely dependent on where you are in your own timelines for deploying a culture of quality and safety, but our early realization of output given where we were was as early as 6 months.

Danielle:         Thank you. Our next question, how should we assist physicians in having those cost of care versus value of care conversations with patients?  And that asker is in admin.  She’s not a physician herself.

Dr. Macias:     So really the question is what is the tension between the cost of care and value of care and how do you approach a culture of discussions with physicians, understanding the difference  between the two.

So I’m going to start by giving my perception of not just physicians but all clinicians, and in my world, clinicians especially who work in the world of pediatrics.  I am very lucky.  I work with a very dedicated specialty subspecialty that is really focused on improving the outcomes of care for children.  Not a single one of us on a regular basis goes to work in the morning, thinking how can I just get by, how can I just get to work and do my job by the skin of my teeth.  And the reality is we do what we do because we believe in the outcomes of healthcare.  And I would say that’s true with the vast majority of healthcare providers.  Where the gap of cost of care to value of care understanding is we fail to provide data on a meaningful level that has some relevance to the delivery of care.  So if I, circa 2000, when I had none of these tools available to me, was working in the emergency department, accumulating the list of all the patients I saw with asthma, I couldn’t aggregate that data.  I didn’t have any way of understanding today’s 30-­‐ minute difference in my strategy to go to triage and write orders in the emergency department as soon as the patient had triage for steroids, impact the outcome of care because of systems for data collection simply didn’t exist.   That would mean me getting special permission to derive data from chart pulls, compile it, try to understand difference in those populations, and then  tell myself, yes, that made a difference.  Well, working in a 60-­‐hour workweek at that point in time, it just wasn’t feasible to deliver that kind of understanding of the value of care.

In a world in which data transparency is provided, we can accelerate that discussion really playing upon the strength and the desire of the clinical care providers, that they want to  achieve good outcomes for their patients. So that if they understand that doing these  strategies rather than these strategies results in a better outcome because our local data demonstrates that, that’s really what’s gonna motivate them. That’s really what’s gonna help them understand what the value of care is all about. Assigning a dollar value to that, understanding how much that charge was or how much that cost was only adds that much more to their appreciation and their driver for improving the outcomes of healthcare.  When you can systematize it the way that we’ve done, you’re now able to provide dashboards that help people to understand not only where their outcomes for their patients were as it relates to their clinical care but what the financial outcomes look like and how these cost savings could add to more expanded programs, more community health programs, so a better population health management.

Danielle:         Thank you. And someone else asks, does your data warehouse help you to realize a better return on investment on your EMRs?  And if yes, how exactly do you measure that ROI?

Dr. Macias:     Wow, that is a million dollar, quite more than a – probably a trillion dollar question.  There had been much written in the quality literature on return on investment and how one begins to quantify it, and I think part of the biggest challenge for that is understanding all the factors that go into cost. We often focus improvements in cost by looking at decreases in expenditures. So we can look at the cost of implementing the quality program but it’s the displaced cost that we often don’t measure. So what are the untoward effects, what are the balance metrics that may have changed that we actually do need to measure, what are the improvements in patient satisfaction that lead to better referrals? I’m gonna have to cheat on the answer to this – is to say that I can’t exactly quantify the return on investment because there’s so many hidden costs within the quality strategies that the literature doesn’t well inform us on but we know what some of the internal costs are and we built some of those structures financially, some of those models, to understand what some of those, at least the return on investments are. But I will say at this stage in the game, we do a very poor job, as does most of our healthcare system across the US in quantifying the hidden cost of quality, in quantifying the hidden benefits of quality, what are all the external costs, and I think that’s what makes the ROI difficult to calculate.

So personally, I focus much more on the value equation and understanding these are the cost reductions that we’ve seen and we know that this is valued better with better patient outcomes.  And if we can parallel better efficiency and effectiveness with a lower cost and a better satisfied population, then I think we’re headed in the right direction.  It is the right return on investment.  I just can’t give you a number.

Danielle:         Okay. Well thank you for attempting to answer that. And our next attendee asks, where were vertical and horizontal processes integrated and streamlined prior to building the EDW?

Dr. Macias:     So question regarding vertical processes and horizontal processes, and is the person asking the question asking about vertical processes of clinical care or vertical operational processes?

Danielle:         They didn’t specify. That was the exact question and I don’t know what fields they’re coming from.

Dr. Macias:     Okay.  So I’m going to interpret this to really look at vertical and horizontal processes of care delivery.  And so, when we look at asthma as a vertical process to say I can look at the disease process and look at all that it encompasses, prior to the Enterprise Data Warehouse, all of that was really being driven by whatever collaborative efforts and guidelines of care we may have gathered to volunteerism. There wasn’t a systematic approach to that, there wasn’t a governance that would either define what that should look like, or more importantly, what metrics should be used along the vertical clinical care processes.

For the horizontal, I would say that that was much more challenging because we tended to think of silos of care, so that if the emergency department was looking at the management of septic shock with the asthmatic and the NICU was looking at management of septic shock with the neonate with fever in the newborn and the hemog unit was looking at septic shock in their bone marrow patients with fever and neutropenia, there wasn’t a lot of opportunity for horizontal connection. This has provided data across silos. It’s also put people into the same room. It’s also built teams bottom up with not just creating that science piece that I talked about in extracting that but it’s also linking that to data transformation across silos of care, that we have a better view of the horizontal processes, as well as at the transformation. So linking again data transformation, science, and process improvement but not doing it in a manually extractive fashion.

Danielle:         Ok. Great. Thank you. And did you have a hard time getting leadership buy-­‐in regarding implementation process?  And what strategies did you use?

Dr. Macias:     So I think we’re very fortunate here at Texas Children’s Hospital. We’ve had a very proactive information services group. Our chief information officer has on more than one occasion been very transparent about her decision-­‐making, and I have to attribute a lot of that support to my colleagues at quality and outcomes management, to our chief quality officers and patient safety officer, and to our administrators, especially our chief information officer.  And I’m going to hone in on something that resonates as something that our CIO had said to me in a conversation about evidence-­‐based guidelines and what I’m trying to do in the clinical realm and what our teams are trying to do – And her words to me were, “I don’t want to be the clinical decision-­‐maker. It is not my strength. It is not my expertise, but instead I want to be able to collaborate. I want to be able to build and share the vision that we all share but provide the right components of it.” And I think that probably was the most telling piece of where engaging it really was leveraged.  It was working with the chief information officer and our IS group to understand that our information technology strategy should not be driving the workflow and should not be driving the processes of care but instead it should be iterative, it should be side-­‐by-­‐side discussions, it should be collaborations with the governance that recognizes the importance of that side-­‐by-­‐side decision-­‐making.

Danielle:         Ok.  Thank you. Thank you for that. Someone asks, have you implemented the integration to diagnostic imaging with your EMR?

Dr. Macias:     So we have several elements that we have integrated in our Enterprise Data Warehouse. We’re looking at some clinical efficiency issues currently. I work at pediatric hospitals and ionizing radiation is a very important issue. Some of the literature out there had suggested that there are about 500 cancer deaths attributable to 600,000 CTs. And so, when we begin to look at those numbers about harm, it becomes much more salient that we look at the science of this linked to the efficiency piece of it. So we’ve taken strategies that are crosscutting to say where are the opportunities to decrease ordering waste and to decrease errors if we consider what could be a calculated defect from ionizing radiation and look at risk-­‐benefit in the evidence arena, in the science arena to understand how that should inform a better standardized process of care. The example of that being how do we decrease chest X-­‐rays for asthma. We can look at that across closed head injuries and how we minimize CT utilization.  Currently, our appendicitis team is evolving to the arena of looking at CTs versus ultrasound use versus clinical diagnosis.  A lot of opportunities there. I see those as the more horizontal approaches.

Danielle:         Ok. Thank you. And one final question for this afternoon, how do you group episodes of care to identify outcomes?

Dr. Macias:     So, how do we group episodes of care? I could interpret that in a couple of different ways. I would say that we start at the most basic clinical definition. So we look at episodes of care as they relate to what is scientifically reasonable, that they be added to a particular cohort. So as an example, if I look at only administrative databases, most of the health services literature would tell me that if I try to identify a population of children with this process, I may be about 30% accurate. That’s exactly what the case was when we looked at our populations with asthma. So when we looked at children with asthma and then we, now using an Enterprise Data Warehouse, went to the medical definitions and science definitions, and say let’s not just look at ICD-­‐9 codes, let’s couple this with the administration of the beta agonist, use of steroids with prior history of MDI prescription.  So we can begin to assemble a cohort that takes all of those episodes of care and say this is much more likely that this disease process was common to all of those. It’s a much better crunch in population.  And when we in fact measured that against a goal standard, a grant-­‐funded asthma surveillance project that was looking at all respiratory diseases in every case, we were only one off out of those 800 admissions. So that really tells us that using these data strategies can help identify these episodes, cluster them in a clinically meaningful manner.

Now, if you take this on from an administrative realm, I will tell you that strategies that only look at the APR-­‐DRG world grant this proprietary and we don’t know what elements completely go into building those APR-­‐DRG clusters but it’s often what informs clinical decision making on an administrative level, could be one of, especially in the world of pediatrics.  If I look at clusters as they’ve been applied to the world of adults, I think fractions don’t matter as much to our group. A fever in the neonate is hugely important.  So our build through our Enterprise Data Warehouse was really trying to shift away from an administrative definition into episodes of care that were congruent with what we expect from the science and the administrative coding for those and was able to cluster them differently.  Now, we could take a clinical approach that’s clinically meaningful. We’re going to hit a population that isn’t gonna be the 30% of but instead we’re doing targeted rapid cycle process improvements and measuring that in the clusters that are meaningful.

Danielle:         Well thank you, Dr. Macias, for answering that question and all of our other questions today.

Thank you everyone for attending this informative event today. I would like to remind you once again that our sponsor for this webinar was Health Catalyst. Thank you to Dr. Macias and thank you to everyone for attending.

[END OF TRANSCRIPT]