The Deployment System: Creating the Organizational Infrastructure to Support Sustainable Change (Webinar)


The Deployment System:
An Organizational Structure to Support Sustainable Change
John L. Haughom, MD
July 2014

[John L. Haughom, MD]

Thank you, Tyler.  Again, it is my honor and privilege to be with you.  Let’s start with the initial poll question, which is the usual one, trying to figure out our audience.

Poll Question #1

What is your primary functional area of expertise?

What is your primary functional area of expertise?  Clinician, executive, finance, IT, or other?  So I’ll give you a few seconds here to respond.  And Tyler, take it away.

[Tyler Morgan]

Alright.  We’ll go ahead and close that poll.  And let’s take a look.  It looks like we’ve got 12% a clinician, 20% executive, 3% finance, 28% IT, and 37% other.

[John L. Haughom, MD]

Implementing an Effective System of Production in Healthcare

Okay.  As we have discussed in the last couple webinars, I believe all healthcare organizations would benefit greatly from developing a systemic approach to improvement.  In essence, this is a strategy for managing complexity and change with the goal of achieving scalable and sustainable improvements and outcomes over time.  You know, in my experience, such an effective framework is illustrated on this slide.  It divides the strategy of the three critically important components that in combination can ignite sustainable, meaningful, and scalable change.  In the last two webinars, we discussed the first of the three, the analytic system.  Today, we will discuss the deployment system.

Deployment System Components

Chapter 5 of the book explores the importance of having an effective deployment system in place.  The goal of chapter 5 is to enable the reader to identify the essential teams and their key interactions, explain the benefits of using an era or agile approach for improvement, and understand how to leverage the organization’s analytic system, accelerate lean process improvement.

Now, just as a heads up, with today’s presentation, I decided to take a different approach than I have previously taken.  In prior webinars, I essentially presented an overview of the chapter being covered in that webinar.  However, based on questions and comments people have shared with me over the past few weeks, it’s becoming apparent that many of you who participate in these webinars have downloaded the book and are in the process of reading it.  Of course, I’m appreciative and gratified that you are taking the time to read it.  Out of respect for those of you who are reading the book and have read chapter 5, I am going to use today’s presentation to do two things.  First, provide demonstrations of modern analytics that clearly show how it is essential to operational and clinical improvement and how it enables a new way, a better way of delivering care; and secondly, to emphasize and elaborate some key points in the chapter rather than simply reviewing the entire chapter.  If you have not read the chapter, do not worry.  I believe this presentation will still be very interesting and useful to you and you can read the chapter later for an in-depth discussion of the key components of the deployment system.

A Demonstration…

At the outset, let’s turn our attention again to a demonstration of modern analytics tools.  These demonstrations will set the stage for a discussion of key elements of a deployment system.  This slide lists most of Health Catalyst applications, though not all of them.  Once again, they are categorized into three categories – foundational, discovery, and advanced applications.  Collectively, these applications provide tremendous flexibility and power to analyze and improve clinical care and operational processes.  I put a yellow star next to the Key Process Analysis or KPA discovery application that I demonstrated in the webinar last month.  This tool allows organizations to identify the most significant areas of opportunity to improve care while also eliminating substantial cost.  Today, I will be demonstrating the additional applications including the Population Explorer and one of the advanced population suites highlighted by the red stars on the side.

So let’s begin with the demonstration of the Population Explorer and how it can be useful to organizations.  Just give me a second here while I switch off.

[Tyler Morgan]

Dr. Haughom, while you are pulling that up, we’ve had several questions and comments asking for a link to your ebook and we’d like to let everybody know that we will provide that link to the ebook in both the chat section of this webinar.  We will also include that link in the email that we send post webinar.

LIVE DEMONSTRATION

[05:27]

Population Definitions

[John L. Haughom, MD]

Sure.  Thank you, Tyler.  Okay.  I’ll start out in populations (05:33) on the population tab here.  The Population Explorer is often used early on by organizations to doing specific patient populations and to understand some basic risk profiles associated with those populations.  The application is based on the Catalyst clinical integration hierarchy that we discussed last month.  And out of the box, applications are largely founded on administrative codes which allows organizations to identify populations early in the process.  Today, what I’m going to do is continue with the heart failure example that I used last month.  So let me type in heart failure.

Heart Failure

Heart Failure.  I can select the heart failure population simply by typing that and select it.

What you see at the top here is the list of rules that we’ve used in heart failure population of patients.  Some of them are based on CMS criteria, some of them are based on additional criteria for diastolic heart failure cases with systolic hypertension.

[Tyler Morgan]

We apologize for the technical difficulties.   We’ll have Dr. Haughom back out and then jump back in here to make sure that you guys can see exactly what he is talking about.

[John L. Haughom, MD]

Okay.  I apologize.  So we’ve selected the heart failure population and these are the basic criteria or rules that define that population and mostly it’s the administrative codes at this point but we’ll talk later about how to add additional criteria.  In the table below, we’re looking at the parameters that have been used to define these rules.  Initially, we’re looking at administrative codes because that is what the starter set includes.  But as I indicated, these parameters can easily be augmented with other information, including clinical criteria like ejection fraction, medications, lab results, and other important information.  Ultimately, it is important to add these additional criteria because improvement teams can variably want to refine the clinical population under study.  During the second demonstration this morning, looking at the heart failure advanced application, we will see that additional clinical rules in fact have been applied.

Population Summary

Clicking on the summary tab, we see some basic information about this population.  Because we’re looking at a heart failure set of patients, we see the 30-day readmission rates trended over time in this top graphic.  Here, we’re looking at cost but we can look at a variety of financial metrics, including charges, payments, and variable costs.  This data varies depending on how our Catalyst clients are doing costing but we can usually pull in some level of costing data.  This illustrates an average inpatient length of stay.  It’s trended over time and we’re also showing the average across all the time right here.  It’s important to understand that we have risk profiles underneath each population.

Demographics

And this demographics tab shows us some of that information.  We’ve looked at the various age buckets.

And we can easily look at the subset of the populations just by drawing a box around this upper chart for the older patients, which we probably would want to pay attention to.  Now, we’re looking at older patients.  We can look at them by gender, and we can also look at their risk profile.  This is the Charleson Index.  It’s a risk stratifier that’s based on diagnosis of comorbid conditions, as well as the age of the patient.  The higher the number, the more risks there is associated with these patients.  This comes with one of our products called the comorbidity analyzer and it’s used as a basic risk stratifier.  Again, that’s part of our foundational roll-up shortly after our platform is installed.

Demographics

It’s important to note that these tools allow us to look at other population metrics across the continuum of care.  Now, in the first tab, we were looking at primarily inpatient metrics around readmissions and length of stay.  Under the demographics tab that we just looked at, we were looking at primarily demographic information.

Primary Care Visits

Under this primary care visit tab, we start to look at some metrics really to primary care visits.  Obviously, it’s important for us to make sure that patients who have chronic conditions such as heart failure are making regular visits to primary care.  What this graph shows is the number of days since their last primary care visit.  The bars indicate how many patients with heart failure in each category.  Now, I should point out that this is actually real data that’s been de-identified but there are about 40,000 or 50,000 charts in here that you’re looking at and it is real.  It’s actual data.

On the right-hand side of this chart, we can see there are a large number of heart failure patients who haven’t been in for over a year.  This is something that we would probably want to investigate and resolve.  So let me select those patients.

Primary Care Visits

We can isolate these patients who haven’t been in for quite some time in that primary care clinic.  In the table below here, we have an actual listing of the patients with their name, medical record name, times his last seen, and phone number.  And it’s important to note that this list is sorted by the highest risk patients that you see in the red here.  This ability to sort by risk is very powerful.  It helps you focus and prioritize your efforts to the clinic.  You can imagine that care managers and medical assistants would spend a lot of valuable time using this information.

Cross Cohort Comparison

Another very valuable piece of functionality of this tool is being able to compare across the organization.  We can do this under the cohort tab here.  We’re looking at right now in a heart failure cohort.  So this is a comparative and it allows us to look at two separate populations.  One of the populations is highlighted in blue and the other in grey.  In this case, we’re looking at heart failure in blue, as illustrated here.  And because nothing is selected down below here, this is the health system’s entire rest of the population that’s illustrated in grey.  So we’re seeing all of the rest of the patients compared with heart failure patients.  Of course, we see a much higher readmission rate with heart failure than a general readmission rate, which is only 7.1% in this particular example with this data.

We can also do things like compare different facilities in an organization.  So if I want to look at heart failure patients across two different populations, I can click on this button to select heart failure.

Cross Cohort Comparison

Now, I’ve got heart failure represented by both blue and grey, so they’re identical lines.

Cross Cohort Comparison

But I can go up here and I can select, let’s say Granite Hospital for this one.  And down here I’ll select Millrock, which of course is fictitious.  And now I have a comparative view of these two different institutions which are both part of my health system that are caring for heart failure patients.  When I do this, I see a big difference in their readmission rates.

I also see a big difference in their length of stay and in their financial metrics.  Using this tool, we could ask things like are we seeing sicker patients compared to the other facility or are there things we can learn about the cost and quality of care in one facility care to another.  I’m sure that it’s obvious to you that this kind of functionality can be very helpful if you work in a health system with multiple hospitals and clinics because it will help you more precisely manage populations of patients in variations in care or cost.  I don’t have time to go through all of the tabs of these tools but I think I’ve given you a good idea of what you could do in our Population Explorer to better understand different patient populations and the risk profiles, the details about those populations and about different parameters about those populations.

END OF LIVE DEMONSTRATION [16:31]

So let me go now back to the slides…

Improvement Methodology

Okay.  When I discussed the improvement methodology in the book, I’m borrowing from the Institute for Healthcare Improvement’s methodology.  The IHI methodology, as you first defined a goal for a specific care process, such as improving heart failure readmissions, the goal can be thought of as an umbrella.  It is the desired endpoint for a specific clinical problem like heart failure readmissions.  Improvement teams then work at identifying AIM statements that help you achieve the particular goal of reducing heart failure readmissions.  As reviewed in chapter 5 of the book, AIM statements are written, measurable, time-sensitive objectives that move the team toward the goal of reducing heart failure readmissions.  Improvement teams can often identify a handful of AIM statements designed to achieve a specific goal.

Let’s now review 5 AIM statements that could help a heart failure readmission improvement team achieve the goals of reducing their readmission rate.

CV Heart Failure AIM #1

The first one is almost routinely the first AIM statement that clinicians select because they want to validate the data and establish a baseline.  While they’re excited to get the data, clinicians still want to make sure that the data is accurate.  This is reasonable and is definitely appropriate.  That’s’ the first AIM statement that often revolves around this type of validation and baseline setting.

CV Heart Failure AIM #2

The next thing improvement teams often want to do is to identify high risk patients in the population under study.  In this case, we’re looking of course at heart failure patients.  There are a variety of risk stratification models available to help predict the likelihood of the return of a heart failure patient within 30 days for readmission.  We all know that we have limited resources in healthcare and they are getting more and more limited as time goes on.  And so, organizations often want an actual (18:45) to spotlight those critical patients, those high risk patients, and aim their intervention efforts towards those high risk patients as a way to maximally reduce readmission rates, improve care and save costs.

CV Heart Failure AIM #3

The next three AIMs that I’m going to review, identify interventions that are known to reduce heart failure readmissions – there’s no magic bullet to keep patients out of the hospital and a single intervention may or may not show significant results on longer readmissions.  So we often use a bundle of interventions that are known to collectively make a difference.  Published research has identified several factors that in combination seem to help reduce heart failure readmission rates.  These next three AIM statements are three of those recommendations.  The first is making sure that high risk heart failure patient has a follow-up appointment in the clinic within 48 to 72 hours after discharge.

CV Heart Failure Aim #4

The next AIM statement revolves around medication reconciliations and tracking compliance with reconciliation recommendations.  As you know, medication reconciliation helps to determine that the patient is on the right medications, that they understand what the medication does, and that they know how to take the medication correctly.  If we can get a high compliance with medication reconciliation and tracking compliance, it helps improve the readmission rate.  It’s crucial to making and keeping patients out of the hospital.

CV Heart Failure AIM #5

Lastly, it’s been shown to be very helpful to have a nurse do a follow-up phone call post discharge to check in on a patient.  Make sure that they obtained their medications from the pharmacy, that they’re going to be able to make their follow-up appointment and to answer questions.  In combination, these three AIM statements can significantly help decrease 30-day readmission rates.

Now, let’s return to our demonstration to illustrate how an advanced application can help an improvement team measure and improve these particular AIM statements over time.

[Tyler Morgan]

While Dr. Haughom is switching, we’d like to remind everyone that you can type in your questions in the questions portion of your control panel.  Also, we have sent out via the chat the link to Dr. Haughom’s ebook.  So you can click on that and obtain the ebook that way as well.

LIVE DEMONSTRATION

[21:40]

Heart Failure Readmissions

[21:40]

[John L. Haughom, MD]

Okay.  Now, you’re looking at the heart failure advanced application.  Let me orient you  to this view of the data.  Across the top are the data related to heart failure readmissions.  You can assume this data has been validated by the physician lead, the nurse expert and the clinical operations that’s involved in caring for heart failure patients at this organization.

On the right here, we have data regarding 90-day ER utilization and 90-day observation stays.  We refer to these as balance metrics.  These balance metrics are important because if we reduce readmissions, if we then have a patient showing up in the ER or in an observation rate, we’re not really managing their care optimally.  We’re not optimizing quality or maximally reducing cost.

Below that, we have the interventions that we just talked about in our AIM statements.  You can see the different interventions we’ve identified in the AIM statement.  So track with percentages.  They can be very easy actually to identify what areas need to be focused on for additional improvement.

Then at the bottom we have a graph illustrating the trending of the readmissions over time.  So really what you see on this first page is the baseline of what we described in the AIM statements.  You’ve got the readmission baseline up here, then the specific interventions that are described in the AIM statements here, and then we have tabs congruent with the various interventions that we defined along here.  By the time we’ve developed an advanced application, like this heart failure example, we start to include some more clinical criteria or definition based on work done with the improvement teams at our client organizations.  We’re not just using the administrative codes anymore.

[24:17]

So if I click here, you’ll see the criteria that are now defining this patient population.  This heart failure definition was developed by clinical leaderships sitting down with the members of their team and identifying the criteria that’s necessary for heart failure patients.  With the clinically validated population (24:34) team has validated these factors of population, the cohort they want involved.  In addition to heart failure, we have core measure definition that’s based on the CMS definition, then we have a series of ICD9 codes.  We have mappings to ICD10 but we don’t have ICD10 in the demonstration department yet.  We see a lot of ICD9 here.  And then there are other lab and medication criteria that have defined this population.  As I mentioned, this criteria can be modified, expanded, or contracted based on the requirements established by the improvement team.

[25:28]

Now, I can drill down into this even further.  So for example, if I just want to look at patients that are defined by the medication criteria, I can select that.  We now have updated data.  We’re looking at that medication specific heart failure cohort.  It’s important to know that the Catalyst platform, the data warehouse, has a necessary flexibility and infrastructure for defining these populations and to modify these rules and the visualization to create individualized stratified versions of different populations.  It’s very important to be able to provide clinicians this level of flexibility and adaptability so they can adjust the views of data, improve care, and highlight dynamic care environment, depending on where they feel they need to go.

In addition to our population stratification, we can stratify by risk.  So that’s illustrated here with a series of risk filters.  Some organizations have a simple flag and (26:22) where a physician can indicate whether their patient is high or low risk is illustrated there.  We can identify high risk patients using the Charleson Index that we just discussed as well.  That’s illustrated here.  We can identify patients based on their comorbid conditions, in this case, 1, 2 or greater than 2 comorbid conditions.  And we have a fourth opportunity here to identify by risk, this is the Catalyst heart failure risk index.  It’s another level of stratification that’s a lot more advanced.  This one is specifically targeted for heart failure patients and this particularly gives us a number between 1 and 100 that indicates the relative likelihood that one of these patients will be readmitted for this heart failure condition.  It’s (27:17) by a 90% accuracy rate predicted risk.  So it’s that most accurate risk stratifier developed on a training set specifically for individual clients and is a highly accurate way to identify these high risk patients in this population.

Heart Failure Medication Reconciliation

[27:44]

If I want to drill down on a data regarding medication reconciliation, I click on the medication reconciliation tab.  Let’s say I just want to look at the year 2013.  So I select that year.

[27:50]

What we see at the top is our medication reconciliation compliance rates over time.  The blue dots indicate whether the medication reconciliation was done within 48 hours.  The yellow is whether our pharmacist has actually reviewed it.

Down below, we have a block chart illustrating performance by different care units.  We can see the units that are doing better at medication reconciliation.  Most of them are mostly green with just a little bit of red.  But we do have one outlier here, Northwest, who’s doing a particularly poor job at medication reconciliation.  So this is how we can identify a problem here, is our problem units, in order to identify the root cause, understand where some of the problems may be coming from, and improve care over time.  We can also figure out who’s doing a good job and duplicate their processes in another unit.  You can imagine how important this information would be to a nurse manager or someone in the quality department.  Obviously, there needs to be some serious education and med training of the leadership or staff on the Northwest unit.  Without this dashboard information though, you really would never have had this information available to you.

Let’s now look at the drilldown on the phone call metric.

Heart Failure Follow-Up Phone Call
[29:23]

This metric is tracking the AIM statement that said every patient should have a follow-up phone call and it’s laid out similar to the medication reconciliation tab.  Off at the top we have the trending over time of the compliance rate with the metric.  So the number of patients who were called, the blue line; the number of patients who were called within a certain amount of time, the yellow line; and the number of patients who were reached, the orange line.  We know that in this table we actually have a list of the patients who haven’t been reached.  So if we need to intervene and started calling these patients, we have the information about them, their medical record number, their name, their age, and again it’s stratified by risk.

So if I were a very busy nurse and I only had 20 minutes in my day and I need to make some follow-up phone calls, I would know exactly where to focus.  This is an exciting tool that helps clinicians greatly improve care.

Heart Failure Follow-Up Phone Call
[30:30]

I can just go to the number of days since discharge here and select it out even more.  So let’s say I wanted to just look at the patients in the last 14 days.

[30:50]

And now you see that data.  That concludes the heart failure portion of the demonstration.

Now, just for a moment, consider the power that you’re seeing here.  Putting the analytic tools like this in the hands of clinicians can greatly improve their ability to deliver the best possible care at the lowest possible cost.  This represents capabilities that clinicians have never had before.  This is in fact healthcare delivered in a better way.  Some of you may now know where the title of the book came from.  This will indeed be entirely new era of opportunity.  We will have tools and data that will vastly increase our ability to improve the quality and cost of care while eliminating waste.  This is way beyond our father’s and mother’s era at healthcare.

So now let’s return to the slides and talk about the deployment system.

Two Important Questions…

So let’s assume that we have the analytic capability that I just demonstrated in place in our organization and improvement teams are eagerly using it to improve care and they’re making progress.  As they make progress, the organization is going to want to spread the improvements and sustain them over time.  As I indicated at the outset of this presentation, I’m not going to discuss all of the elements of a successful deployment system.  I believe that that’s pretty well covered in chapter 5 of the book.  However, in the remainder of this presentation, I do want to discuss two important questions.  If you have analytics in place, why is a deployment system necessary?  And secondly, what is the most important ingredient that makes a deployment system successful.  I want to emphasize that by focusing my remaining comments on these two comments, I am in no way suggesting that these are one or two things that are required to implement an effective deployment system.  There are many other elements of an effective deployment system that are well covered in chapter 5 but these are two important questions that must be considered at the outset.  So let’s look at the first question.

Why is a deployment system necessary?

With respect to the first question, is a deployment system really necessary?  The evidence would suggest the answer is profoundly yes.  You absolutely need an effective deployment system to spread best practice throughout an organization, to sustain improvements over time, and to support continuous learning and a continuous improvement over time.

For different results, something has to change…

These quotes by Albert Einstein and Paul Batalden, the physician at Dartmouth, effectively argue why we need to pay attention to how a system is organized and how it operates if you really want to change its outcomes.  “Insanity is doing the same thing over and over and expecting different results,” said by Einstein.  And “Every system is perfectly designed to get the result that it gets.”  So the fact of the matter is that if you do nothing to change the way a system is organized and how it operates, any system, including the health system, you cannot expect to get different results over time.  And in fact, you will not get different results over time.  As we’ve discussed in prior webinars, the current health system has accomplished some really great things but it’s also being overwhelmed by the complexity and producing some serious quality, safety, waste, and cost issues.  If we do nothing to change how we organize and operate, we would be foolish to expect different results.

Outcome of a Traditional Deployment System

Even if you managed to temporarily improved a process under our present system, it is very difficult, probably nearly impossible, to spread and sustain against.  There are simply too many processes and too much complexity in our current health system to ensure that this happens, unless you have an effective deployment system to ensure that the improvements are spread and sustained.  Organizational teams that can drive scalable improvements are one component of a deployment system.  To improve their deployment system, an organization needs to start by establishing permanent teams that take ownership of the quality, cost, and patient satisfaction associated with care delivery.  An organization also needs to organize team structures, provide training and roles, allow teams to design their own solutions and to ensure improvements are implemented consistently over time and throughout the organization.  Organizations often deploy teams when they need to make a change but few do it in the manner that supports scalable and sustainable gains.  As a result, they often enjoy temporary success followed by return to baseline performance, as illustrated in this slide.  Common characteristics of such teams include being temporarily assigned, receiving little or no support from members of the organization’s technical and operations team, approaching the work like a project with defined beginning and end and having no access to an analytic system.  Without having an effective deployment system, an organization cannot effectively deploy and sustain gains.  Without having an effective analytic and deployment system, organizations cannot continuously improve.  I believe the demonstrations that I’ve showed you just a moment ago can help you understand why this analytic system is so important.

Deployment System Team Structure

Team Interactions

When an organization begins to develop their deployment system, we recommend that they assemble a few essential teams – guidance teams, clinical implementation teams, and work groups.  All of these teams have several things in common – they are permanent, they support care process families, and they integrate clinical, operational, and technical experts.  The make-up and role of this team and the team interactions are illustrated in this slide.  These teams and their roles are described in detail in the chapter 5 of the book and I will not review them now.  I just want to emphasize that in order to achieve, spread, and sustain improvements, an organization needs a structure that will support back.  Otherwise, they will leave their success to chance.  What is illustrated here and what’s discussed in detail in chapter 5 is an example of the type of team structure that we know works, we know it works, basically that has been successful on many health systems, and we definitely recommend it.

So the answer to the first question, do you really need to design and implement an effective deployment system, is absolutely you do.  Without an effective deployment system, an organization will not be able to achieve, spread, and scale and sustain improvements over time.

Two Important Questions…

Now, let’s consider the second question.  What is the most important ingredient to make a deployment system successful?  Once again, in answering this question, I want to emphasize, this is not the only important ingredient for a successful deployment system.  I only want to highlight that in my view, it is at or very near the top of the list of important ingredients for success.  Other important ingredients were discussed in chapter 5.

Healthcare: The Way It Should Be

Recently, I’ve had a number of people familiar with the book asked me why I took the task of writing them.  It was needed task.  I put a few months into researching the book followed by roughly 10 months of writing and editing and working with the team.  At the outset of undertaking the project, my goal was actually to compile the best knowledge available regarding how an organization can successfully participate in or perhaps more accurately survive healthcare transformation.

About two weeks ago, my good friend and colleague, Paul Horstmeier, pointed out that he had concluded that the book was really a handbook.  And I agree.  Indeed it is a handbook.  It’s a compilation of some of the best thinking and experience regarding the present and true state of healthcare.  I believe this really is a good handbook that can help organizations over time.  As of two weeks ago, I would have said it’s a handbook for transformation.  It’s probably that.  However, on further reflection, as I considered this presentation, I decided it was more specifically a handbook for engagement.  The reason I say that is that engagement of those in healthcare in the transformation process, especially healthcare’s frontline workers, clinicians, is absolutely essential if healthcare is going to change.  Without broad engagement, especially of clinicians, healthcare will not change.  The day we achieve broad engagement is the day healthcare will profoundly change.  Thus, the book really is about explaining why healthcare has to change, what you need to know and do to successfully participate in the change, and what success will look like when we get there, which we will cover in part 3 of the book and in the future webinar.

Levels of Engagement

Now, if you look at any given group of people, their level of engagement covers the spectrum between engaged and disengaged.  Actually, I decided to add the word “relatively” because there really are very few people in healthcare that are entirely disengaged.  As I have said in the past, healthcare is a system loaded with people who are bright, well educated, and committed to serving patients well.  We are fortunate to have lots of people who want to believe they are A students.  Still, we all know that some are more engaged than others.  We all recognize the engaged group.  They tend to be what I call “opinion leaders”.  They’re hungry for best practice and they are deeply committed to continuous learning.  We all know who these folks are.  They are the physicians that physicians go to if they need help or who they send their family and friends to if they need help.  They are the nurses that everybody wants to take care of them in the hospital or clinic.  They are the operations leaders that routinely outperform goals and win awards.

The more disengaged tend to be less hungry and are more comfortable doing things the way they’ve always done them.

Up until about 5 to 7 years ago, most of the engaged folks were comfortable with the traditional craft of medicine.  Things that worked for them in the past for many many years and they saw really little need to change.  However, over the past 5 to 7 years, I’ve witnessed the profound change in this group and I see that all across the country.  They are becoming aware of healthcare’s challenges and they’re beginning to see the need for changes.  They frequently go through a period of frustration and skepticism but that’s okay, because eventually they get to the point where they’re increasingly looking for solutions.

The more disengaged would remain in a state of needing more convincing.

Engaging the “Smart Cogs”

In changing the culture of an organization, I believe that it’s important to understand the diffusion of innovation model that was built by Everett Rogers, as depicted in this slide.  Every medical staff, every nursing staff, every operational group has the opinion leaders.  These are the clinicians and operational leaders who want to be the best.  They are the engaged and they are the innovators and early adopters of healthcare.  In my experience, they typically represent about 15% to 20% of every group.  It’s important to recognize this because they can make it easier to move an organization forward.  If you win these engaged opinion leaders over to a new reality, they will move the rest of the curve for you.  So focus your efforts there.  I know this works because I’ve seen it work.  I used it in my 20 years as a senior executive.  This whole approach is discussed in more detail in the book, so I’ll not elaborate on it more here.  But it is important to understand.

Linking the three systems

Clinical Integration Hierarchy

This slide is my attempt to pull together everything that I have been talking about in one digital or one graphic.  We are in the midst of discussing the three systems, which are essentially the three components of a sound strategy that organizations can use to ensure their ability to support meaningful and sustainable improvements and outcomes.  We have now covered two of the three of these components – the analytic system and today, the deployment system.  In the next webinar, we’ll begin the discussion of the third component of the strategy, the content system.  We’ve also introduced the concept of a clinical integration hierarchy over the last two webinars.  This is essentially a framework to help us organize our thinking about the complex realm of healthcare delivery.  At the top of the heart, we have the major clinical programs that comprise the majority of healthcare delivery.  In turn, each clinical program is comprised of a series of care process families.  They constitute the majority of care in each clinical program.

Finally, there are a series of care processes that constitute the majority of care in each care process family.  In this case, I’ve illustrated the heart failure example, which is made up of care processes including pulmonary heart disease, cardiomyopathy, congestive heart failure, and valve disorders.  The heart failure care process families comprise of four components that represent the majority of care in the heart failure care process family.  Of course, there are other components of the heart failure care process family but these four represent the majority of care in the heart failure care process family and they therefore represent the majority of the opportunity to improve care and the opportunity to lower cost.  And so you want to focus your attention there.

Today, we also discussed the elements of an effective deployment system, including a permanent team structure that has been shown to be effective in realizing, spreading, and sustaining improvements over time.  The guidance team manages an overall clinical program, like cardiovascular that we were discussing today.  The clinical implementation team manages care process families, each of their respective clinical programs.  In cardiovascular, that is rhythm disorders, vascular disorders, ischemic heart disease, and heart failure.  The various work groups work on improving the various care processes within each care process family.  Four care processes within the heart failure care process family are illustrated here.  I’m hopeful this slide will help pull the various concepts that we’ve been talking about over the last three webinars together for you.

Two Additional Points…

Before I finish, I’d like to make two final important points.  As I go out of the country and I interact with clinicians, I frequently hear “I am too young and busy.  I do not have time to work on this.  I am building my practice.”  Or I hear, “I’m far too far along in my career and I cannot see why I should get involved.”  This Is for the younger clinicians.  Well, Tiger Woods was 3 years old when he shot 48 over nine holes at his hometown golf course in Cypress, California.  Mozart was 8 years when he wrote his first symphony.  Paul McCartney was 15 years old when John Lennon invited him to join a band.  Thomas Jefferson was 33 years old when he wrote the Declaration of Independence.  Think about that.  33 years old.  Mother Teresa was 40 years old when she founded the Missionaries of Charity.  Pablo Picasso was 55 when he painted his masterpiece Guernica.  Winston Churchill was 65 years old when he became Brit’s prime minister and successfully led Brits through Word War II.  Nelson Mandela was 71 years old when he was released from a South African prison.  Four years later, he was elected president of South Africa.  Michael Angelo was 72 years old when he designed the dome of St. Peter’s Basilica in Rome.  Benjamin Franklin was 79 years old when he invented bifocal eyeglasses which I’m thrilled (47:55) personally because I use them all the time.  And Frank Lloyd Wright was 91 years old when he completed his work on the Guggenheim Museum.

The point here obviously is we’re never too old or too young to make a difference.  One can argue that all of these examples I just gave are people that have special gifts and they are.  It may be true that they are.  But we all have special gifts.  This is especially true in healthcare if we had collectively.  This leads me to my second point that I definitely want to emphasize.  While these people that I’ve just described may have special gifts, and so do we, we definitely do so collectively.  Working together, we can achieve an entirely new exciting and powerful way of delivering care that I illustrated when I showed you those demonstrations.  In fact, we have to work together, and when we do, it’s powerful and it’s fun.  That’s what a deployment system is all about.  I cannot tell you how many times I’ve been in rooms where I see the light bulbs switch on for people in healthcare, when they see new possibilities and they get engaged when I see a better way of delivering care.  It’s electric.  The energy in the room is absolutely palpable.  You could feel it.  To address healthcare’s challenges and benefit patients, we simply need to replicate that experience that I’ve experienced in those rooms across the country.

Thank you for your time and attention.  Before we go to questions, let me post one additional poll question.

Poll Question #2

Does your organization have the necessary organizational structure that not only supports the spread of innovation and improvement, but also allows improvements to be sustained over time?

Does your organization have the necessary organizational structure that not only supports the spread of innovation and improvement, but also allows improvements to be sustained over time? A) 5 – definitely, down to e) 1 – not at all.

Tyler, take it away.

[Tyler Morgan]

Alright.  We have that poll up.  We’d like to remind you, as you are filling out that poll, that if you have questions, please be sure to enter them into the questions pane in your control panel.  We have had a lot of questions about the slides.  I’d like to remind everyone that we will be providing the link to the recording of this webinar, as well as link to the slides after the webinar.

I’m going to go ahead and close the poll and let’s share the results.

Dr. Haughom, we’ve got 15% that say they definitely have that structure in place, 17% at a 4, 31% at 3, 27% at 2, and 11% at 1 or not at all.

[John L. Haughom, MD]

Yeah, I think that the results accurately reflect where healthcare is.  You know, I can see the movement in the healthcare organizations across the country and as they implement advanced data systems and as they go through the transformation, they’re slowly changing their organizational structures and the way they operate.  But it’s a slow difficult process and as I discussed in chapter 5, it requires a pretty complex change management.  So, those results pretty much reflect what I would have expected, that we’re starting to move but we’re not there yet.  So thank you all for replying to that.

Alright.  Before I go to the questions, let me just show you a couple more slides here.

Thank you

Upcoming Educational Opportunities

First, we got some upcoming educational opportunities.  We have a webinar on August 14th by Steve Barlow, one of the co-founders of Health Catalyst, Six Reasons Why Healthcare Data Warehouses Fail.  And then very exciting, we have the Healthcare Analytics Summit that’s coming up on September 24th to 25th and we’ll be giving some tickets away for those.

Transforming Healthcare Through Analytics

Objective

And on this slide, you can see a list of the very exciting array of keynote speakers and the objectives of the summit.

So thank you.  I hope you all attend this.  It’s very exciting.

Tyler, I will take questions.

[Tyler Morgan]

Alright.  Well, before we jump in to the questions, we do have two passes to give away to the Healthcare Analytic Summit that I mentioned earlier.  Now, the first is a pass for a single registration.  The second is a pass for a team of three.  And before doing this, I would like to mention that because of limited space, we have asked the winners to register by August 1st to confirm that they can come.  Because August 1st is so close, we will extend that deadline request to August 15th.  So please enter the drawing if you feel you could travel to Salt Lake and could register by August 15th.  So folks, it’s very simple.  Please answer the poll question if you’d like to be entered in the giveaway and are confident that you can attend the summit on those dates.  This poll is for a single registration pass.  I’ll leave that open for just a few moments to give everyone the opportunity to enter if they would like to.  Okay.  We’ll go ahead and close that poll now.

And the next is the similar process but this is for the team ticket and this is a registration pass for a team of three.  I apologize.  Let me re-launch that poll.  I’m afraid that it closed too soon.  Hold on just a moment.

I apologize for the wait here.  Just a moment while I get that new poll up…  Alright.  Well, that is processing.  Let’s go ahead and get to some of the questions.

QUESTIONS AND ANSWERS

Dr. Haughom, we do have quite a few questions here.

QUESTIONS ANSWERS
Population health management across the continuum of care is very important.  But how does a hospital obtain that info?  Can you import comprehensive CCDA from ambulatory or from home health? Well, you know, it is very difficult unless they have alliances with the outpatient world, with physician groups and things like that.  In chapter 8 of the book, I’ll talk about accountable care organizations and we’re steadily seeing that concept of accountable care organizations take hold across the country and more and more hospital systems are aligning themselves with physician groups in one way or the other, and that’s really what you have to do and that’s what’s driving that whole trend – is the need to have integrated data across the continuum.  And I think we’ll see it continue.
If Health Catalyst is acting on multiple databases, how is data mart or reference data base kept current and secure and that it’s integrating multiple HIPAA and other privacy sensitive data sets? Well we discussed that when I discussed the analytic system a little bit and I would refer the question to chapter 4 of the book.  But my idea is that we use a Late-Binding ™ approach and the data is pretty much (55:55) and noble from each of the systems into the data warehouse in its current state.  There’s very little early binding.  Most of the binding is late.  And so, the enterprise data warehouse is constantly updated as new data is captured by the transaction system.  So it’s a fairly transparent process once it’s set up and you have the interfaces in place.

[Tyler Morgan]

I do have the new poll question fixed, so I’m going to launch that now.  This is again for the team ticket of three for those who are interested in attending the Healthcare Analytic Summit in Salt Lake City as a team.  So I’ll leave this up for just a few moments to give everyone a chance to respond.  Alright.  We’ll go ahead and close that poll now.  Thank you all for your responses.

And let’s go to our next question.

QUESTIONS AND ANSWERS

QUESTIONS ANSWERS
Yes, yes AIM statement.  How are we keeping track of this and assuring ourselves that we are being accountable to our own AIM statements?  Is it chartered, is it captured by a specific system?  (57:11) an example, are we locked to our own resources to keep track of this? No.  I mean I apologize, obviously I didn’t do a very good job of talking about the examples section here, but that’s what those dashboards are all about.  Once the improvement team has come in place, let’s say, heart failure readmissions, and they’ve done a really great job of implementing the heart failure readmissions at one of the hospitals at the health system.  And then it’s spread across the health system.  You compare it, like I showed you, with the tools that I demonstrated.  Then it’s the responsibility of those permanent teams that I talked about, the guidance team, the implementation teams, and the workgroups, to monitor it over time.  And that’s why analytic system is so important to sustain the gains because if they start slipping, you can see where they’re slipping and you can determine why they’re slipping and you can do something about it.  Without an analytic system and a good deployment system, you simply just can’t do that.  You can’t monitor it over time.  I mean you can imagine those dashboards in the hands of somebody running an apartment or running a clinic.  Every day they can call it up once or twice and they can determine how well they’re doing.
Is this block diagram based on risk analysis in process within this application? Yes.  Yeah.  What I showed you there all had at least the Charleson Risk Index applied to it.  So, it was adjusted for risk.  And that’s a question we always get from clinicians who have restarted down this road with new clients – is the risk adjusted.  Clinically it has to be and if it’s not risk adjusted, the data is much less meaningful.  So yes, it’s all risk adjusted.  Whether i showed you different hospitals or different units or different populations, these risks are adjusted.
How is the reconciliation tracked? Well it kind of depends on your organization, what the process is, but in most organization, the nurse who does the reconsideration record set is a part of the chart.  And so what’s in your electronic health record, you know that it was done and you know when it was done and you know who get it.  So you can track that over time.  So you’ve got to record the data (60:05) and most organizations, they do that at the chart as part of the discharge process.
What technology is used to build these dashboards? We have a couple of different presentation tools that we use and that particular one is – it’s a well-known application that is good for putting that top of the data warehouse.  I really apologize.  If you email me, I will reply back to you.  I’m just blocking on it right now.  But it’s a standard presentation tool.[Tyler Morgan]I believe it’s QlikView that we use for that particular one.

[John Haughom, MD]

Yes.  Yes.  That’s’ what it is.  I’m sorry.

Can you enter updates in this tool?  How do I know that I called someone, for example, so I don’t call them again.  Or another care manager calls the same person.  Most work use will be over in Epic Care or another EHR. Yeah, there’s two ways you can enter it.  Most of the time, and what we recommend, is that when a patient is called, it’s recorded in the electronic health record, like Epic or Cerner.  It’s absolutely the care.  That’s the tool supporting the care, the electronic health record, and you want to record it there.Alternatively, there are ways that you could enter directly into the data warehouse but that’s not the first solution.  The first solution is to use the tool that’s used to other care.  That’s the electronic health record.

[Tyler Morgan]

Dr. Haughom, we are at the top of the hour.  We would like to thank everyone who has joined us so far.  Are you able to stay on to continue to answer questions?

[John L. Haughom, MD]

Yes, I am.  I would be happy to.

[Tyler Morgan]

Okay.  So we will stay on to continue to answer questions past the top of the hour.  We would like to thank everyone who has joined us for up to this time period.  So let’s continue on with our questions that we received.  We still have quite a few to go through.

QUESTIONS AND ANSWERS

QUESTIONS ANSWERS
Can you discuss a stage assembly of this deployment system so we don’t get shocked by costs to assemble all these people? You know, you don’t go from A to B in a day.  You don’t put up a goal from your present state of operating to a new state with a full deployment system in place in a day.  And having done this myself as an executive 20 years in an organization, I can tell you that you start out by defining the need.  You define the need to sustain gains and to spread them throughout the organization.  And you start with roles that you have and employ people that are currently working in their jobs to do that.  But then you always put in, you start designing the system that you need and start creating those roles and spreading it throughout the organization.  And it does take some time to do that.It also varies of course on the size of the organization.  In my case, we have 11 hospitals and about 2,000 people in a clinic.  So it was a big enough organization, so we could put people into these roles pretty much full time to manage over time and it definitely paid for itself because we’re able to save enough money to do that.In a smaller organization, however, you can’t quite do it that way.  You still have to have a function.  You still have to have a role but it may have to be part of somebody’s job rather than their entire job.
These changes require not only infrastructure but also capital resources, how do you fund such efforts? Well, you fund it the same way you fund anything else.  What’s going on across the country is that there’s a lot less capital being invested in bricks and mortar and more and more being invested in IT.  And if you go to HIMSS or to Gartner, they show you those statistics – off the top of my head, I don’t have that most recent ones – but there’s been a steady rise in the amount of money invested in IT over time.  And the reason is what you’re seeing here, the fact of the matter is that future organizations simply are not going to be able to operate.  They’re not going to be able to survive unless they invest in things like this.  And so, we’re going to see more and more dollars invested in IT and fewer dollars invested in things that we’ve invested in in the past, specifically bricks and mortar.
How can we achieve patient engagement if we have so much trouble getting physician engagement? I think it’s the same way.  First of all, let me address trouble engaging physicians.  I talked about this (66:01) in the book.  The majority of doctors and nurses, really I’d say 90%, maybe 95%, get up there and wanting to believe they are the best they could be.  They really do want to be “A” students.  They just don’t see the need for change, like I talked about today, and they don’t have data to know that they’re not “A” students they think they are.  And in my experience, if you go out there and you give them a reasonable argument, which I tried to articulate in the book, that things have to change and that the way they’re going to change is pretty exciting and empowering for them, it really is going to be empowering for clinicians and you give them the data that they start shifting.  Now, it’s a bell-shaped curve, like that Roger’s Diffusion of Innovation that I showed you.  You obviously have the opinion leaders, the 20% that shift first, and then you have the early adopters that shift second.  But they change.  I started this in 1993.  It was very difficult then.  Nobody wanted to talk about change.  Now, it’s changing and this is the thing that I run into is what I showed in the last slide.  Some people say “I’m too young or too old”.  But the reality is that the clinicians have already changed.In terms of patients, I think it’s very similar.  We had a major effort in my prior role trying to make patients engage.  And in fact, we put a lot of patient representatives on improvement teams that we had.  And boy, I tell you, the patients had no trouble getting excited about this stuff.  I mean I had no trouble becoming Evangelist about it when they started going out.  In chapter 8 of the book, and I’ll cover this in the future webinar, I’ll talk about a lot of technologies that are coming down the pipe that are designed to inform and support patient decisions and the studies on those technologies are mind-blowing.  The patient satisfaction is huge within the patients.  Definitely the patients and their family is definitely going to engage with them.So it takes an effort but most people care about their health and if you give them the tools that they need to be engaged, they’ll get engaged the majority of the cases.
Where does change management and training fit in the deployment system? Oh it’s huge, it’s huge.  You know, I skipped over that and I had a slide here and I didn’t have enough time to cover it, so I took it out.  But there are so many important parts of doing this successfully and I reviewed it in the book, you know, the analytic system, the technology and all that.  But the bottom line is managing the complex changes in this organization is the biggest challenge and it takes very strong leadership, it takes an acceptance that they’d have to change, it takes a strong vision about where things are going and that’s what I tried to articulate in the book.  It takes commitment.  It takes good data.  And so, change management is huge and in the chapter 5 of the book, I talked about doing a readiness assessment and I illustrate 3 different examples of readiness assessments for organizations to prepare themselves for the change and determine how ready they are.  It is a very big deal and it takes strong leadership and it takes real commitment.  No question about that.

[Tyler Morgan]

Alright.  Well that brings us to the end of our questions and answers time.  Before we close the webinar, I do have one last poll question.  See, our webinars are meant to be educational about various aspects affecting our industry, particularly from an analytics perspective.  We had many requests, however, for more information about what Health Catalyst does and what our products are.  So if you are interested in a Health Catalyst introductory demo, please take the time to respond to this last poll question.

While you’re doing now, I would just like to remind everyone that shortly after this webinar, you will receive an email with links to the recording of this webinar, the presentation slides, the poll question results, and the names of the winners of the Summit pass giveaways.  Also, please look forward to the transcript notification we will send you once the transcript is ready.

I’ll go ahead and close this poll now.  Thank you for your participation.  On behalf of Dr. Haughom, as well as the folks at Health Catalyst, thank you for joining us today.

This webinar is now concluded.

[END OF TRANSCRIPT]