SubHero Banner
Video Component

(bright upbeat music)

 

Welcome, my name is Lou Brooks.

 

I'm the Vice President

 

of Commercial Analytics here at Optum

 

and today I'm going to talk to you

 

about creating a clearer picture of the patient experience

 

through the integration of claims

 

and electronic health record data.

 

We're all well aware that the health ecosystem continues

 

to evolve as we collectively try to manage

 

the cost of healthcare,

 

develop new compounds to treat diseases

 

and get consumers to take a promo more proactive role

 

in managing their health.

 

We continue to drive innovation in the areas

 

of drug development, health technology, surgery,

 

care models, reimbursement and analytics,

 

but much of this innovation is based upon

 

existing data strata utilizing either claims

 

or electronic health record data in isolation.

 

We have the opportunity today

 

to further accelerate our understanding of healthcare

 

and drive noble changes

 

to many of our fundamental constructs

 

by altering the underlying data foundation

 

to many of our processes.

 

It starts quite simply by integrating claims

 

and electronic health record data at scale

 

to provide a unique data foundation

 

that enables us to see a more comprehensive view

 

of patient care, outcomes and costs,

 

become more tailored in addressing healthcare questions

 

in a more cost effective manner

 

with a single source of data,

 

and improve on our existing models and methods

 

to shift our focus to joint clinical and cost outcomes,

 

to identify those treatments and care pathways

 

that lead to the best clinical outcomes

 

for our patients at the lowest possible cost.

 

How does it all begin?

 

When we think about

 

the data itself,

 

from that standpoint,

 

on the next slide,

 

you'll see that we have a wide range

 

of data potentially available to us.

 

As we continue to work to transform healthcare,

 

a perfect example of that transformation

 

is machine learning models

 

and identifying patients that may be at risk

 

for a certain condition

 

and proactively intervening with those patients

 

to positively alter their health trajectory.

 

However, even the best of

 

those models miss a fairly large portion

 

of the appropriate patients and falsely identify others,

 

leading to some level of healthcare waste

 

and missed opportunities

 

to positively engage with patients on their health.

 

Why is that?

 

Many of the models today are based solely on claims

 

or solely on clinical data.

 

It's like using unleaded gasoline

 

when premium is needed in a sports car.

 

It runs, gets the job done,

 

but it isn't firing on all cylinders optimally,

 

so you aren't quite getting peak performance.

 

Integrating gait as the premium gasoline

 

for that sports car,

 

and it doesn't just stop at integrating claims

 

and electronic health record data.

 

As this graphic shows,

 

there's a wide range of additional data available

 

to bring into healthcare research.

 

The ultimate goal for healthcare researchers

 

is to obtain that comprehensive view of the patient,

 

including healthcare data, attitudinal information,

 

consumer profiling and purchase data

 

and leveraging all of those elements at once

 

to truly understand health outcomes and costs.

 

We aren't there yet today,

 

but the integration of claims

 

and EHR data is a big step in that direction.

 

Imagine for a moment how

 

integrated claims electronic health record data

 

could change the way we engage

 

in existing analysis across

 

the entire spectrum of healthcare.

 

Let's start by imagining being able

 

to fully assess a clinical trial protocol

 

for inclusion and exclusion criteria,

 

and then be able to drive

 

that directly back to site selection.

 

How many times have you used claims data

 

to engage in that process

 

only to be unable to identify the very specific criteria

 

that's essential to your protocol,

 

or be really targeted

 

at your protocol assessment using clinical data,

 

but be unable to identify the best sites

 

for potential trials.

 

The integration of data, claims

 

and electronic healthcare record data,

 

has the potential to support the entire work stream

 

with one set of data.

 

Let's move forward a bit and talk

 

about comparative-effectiveness studies.

 

Most of that work today is done with claims,

 

data and various methods analyzing either total cost of care

 

or the cost of the disease of interest.

 

How much better could those models be

 

if we were able to integrate clinical information

 

in case matching methods

 

and stratify the analysis by disease severity

 

or other clinical metrics of interest

 

that could provide additional illumination

 

and insights on product performance.

 

Integrated data also has the potential

 

to change the way we evaluate product performance over time.

 

Imagine for a moment,

 

being able to stratify patients being put on your drug

 

by clinical severity, measure with lab results,

 

and then tracking the clinical performance

 

and cost of those patients over time.

 

Imagine that information being available at your fingertips

 

when you sit down with a payer

 

to discuss the total cost of care

 

of the patients on your drug compared to competitors.

 

How could that change that discussion?

 

Both clinical and claims data

 

have their own individual

 

strengths and weaknesses.

 

Claims data is missing critical clinical data elements.

 

Things like lab results, observations,

 

notes information where a doctor

 

has recorded specific pieces of information on care

 

that have occurred during that particular instance.

 

Those clinical insights could provide a better understanding

 

of what's going on from a care management standpoint.

 

On the electronic health record front,

 

we're missing claims,

 

data information such as costs, eligibility,

 

and filled prescriptions.

 

As this graphic illustrates in that gray box in the middle,

 

we've got a great deal of data in common to both,

 

but the integration of the data allows us

 

to gain access to all the information.

 

As a result of that,

 

we fill in missing pieces of information

 

from either of the individual unique sources,

 

resulting in a new data foundation for healthcare analytics.

 

Let's look at an example.

 

Very simple, very straightforward,

 

but it gives you some insight as to what happens

 

when you can integrate claims

 

and electronic health record data together

 

from an analytical perspective.

 

Let's imagine that we have a patient

 

who has visited the emergency room.

 

With claims data,

 

we know that they had a claim for an outpatient visit,

 

they went to the emergency room,

 

we saw that it was billed related to diabetes

 

and the cost was about $7,000.

 

With EHR data,

 

we can easily confirm the visit and the diagnosis

 

because those data elements are in common

 

to both sources of data.

 

In EHR data however,

 

provides a wider range of additional data on the patient

 

that would be unseen

 

in just a straight-claims-based-analysis.

 

Including basic observational data such as height,

 

weight, blood pressure,

 

as well as extremely high A1C level,

 

and what medications were prescribed to the patient.

 

This additional information only scratches the surface

 

of what's available

 

from an electronic health record perspective,

 

including things like symptomology,

 

patient reported medications

 

and other health data

 

that will provide context

 

around the specific interaction

 

and the specific treatment decisions.

 

Let's go back to claims for a moment.

 

With that integration of data,

 

we're now able to connect that fill information

 

from the pharmacy,

 

with the electronic health record information

 

and see that the patient actually followed through

 

with the three written prescriptions

 

from that emergency room visit

 

and filled those three prescriptions.

 

And this journey can continue from there,

 

and you can demonstrate a wide range of impacts

 

from a treatment decision standpoint with providers.

 

We'll talk a little bit more about that in a moment.

 

The integration of data does come with its challenges

 

and its limits.

 

Perhaps the biggest three challenges are first, sample size.

 

The more disparate data sources we integrate together,

 

the smaller the sample of data that you have to work with.

 

So it's essential to work with the largest sets

 

of individual claims and electronic health record data,

 

to give you that largest intersection you can

 

from an analysis standpoint.

 

Completeness can also be difficult.

 

On the claim side,

 

many claims sources are eligibility controlled,

 

so you know what you're missing.

 

But as you start to integrate various sources

 

of electronic health record data,

 

you may or may not be missing particular pieces of data.

 

The same can hold true with claims sources,

 

depending on whether or not they're open or closed sources.

 

The final challenge is really around resource competency.

 

You need to rethink the way that you model

 

and you analyze data once you start integrating it.

 

The old claims based algorithms don't hold true perfectly

 

and that research that you've been doing

 

from a clinical perspective, isn't perfect either.

 

You have to rethink the way

 

that you are working with the data

 

and your researchers need to be retrained.

 

We must also remember

 

that privacy is a fundamental component

 

and a responsibility of all of us

 

as we're working with these integrated data source.

 

While it doesn't stop the integration of data,

 

it does complicate it

 

and it complicates what you can do with it.

 

You need to make sure

 

that we're working to look at that data

 

and make sure that we are hyper compliant as we work

 

through all of this integration standpoint.

 

Because, while it limits what we can do,

 

it doesn't prevent outright analytics

 

with that integrated data source.

 

So let's change gears now.

 

We've talking theoretical for the last 10 or 15 minutes,

 

let's move into some actual use cases

 

to show you what we can do with integrated data.

 

I've got a few examples of how

 

that integrated data can be utilized,

 

translated into analytic value

 

and change how we generate insights.

 

The first example is perhaps the easiest.

 

It's the "low- hanging fruit".

 

And we've touched briefly upon it

 

with the emergency room example previously.

 

The integrated data provides a one stop shop

 

for truly understanding the journey

 

of a patient to get treatment from that interaction

 

and written prescription in the office,

 

through the subsequent fill and refills

 

of those prescriptions over time.

 

In this example,

 

we get to see a piece of information

 

that isn't normally available in a claims analysis.

 

That information is

 

the actual written prescription or order.

 

The physician's intended treatment for that patient.

 

Once we have that information,

 

we can move to claims.

 

Claims now allows us to cycle

 

through the administrative process

 

of filling the prescription

 

from the point of presentation to the pharmacy,

 

you through the utilization management programs

 

that might exist

 

and the fill and subsequent pickup by the patient.

 

The claims data also allows us

 

to track subsequent prescriptions and adherence,

 

and if treatment changes do occur,

 

the electronic health record data gives us

 

those reasons for change

 

such as the example in the lower right hand of your screen.

 

So as we can demonstrate with the three product examples,

 

we can see

 

not only how many prescriptions were originally written,

 

how many of those were presented to the pharmacy

 

and ultimately how many of those got into

 

the hands of the individual patient.

 

Having insight along this pathway offers many opportunities

 

for all of us providers, payers,

 

and life savings companies alike,

 

to develop and target programs to maximize

 

the number of patients

 

that they're get their prescriptions

 

and stay on their medications.

 

Imagine for a moment,

 

just one simple example.

 

A provider has access to the integrated data.

 

They can follow a patient who doesn't present

 

that written prescription to the pharmacy

 

and contact them to discuss treatment again,

 

and depending on the reason why

 

that patient decided not to present the prescription,

 

work to overcome that barrier to filling

 

and getting that patient treated.

 

Now let's look at a integration of the data,

 

utilizing both clinical metrics and cost.

 

We're going to look at the correlation between cost

 

and clinical outcomes

 

for type 2 diabetes patients.

 

We took a very simple approach to this,

 

identifying type 2 diabetics in 2016

 

and calculating their baseline,

 

A1c and BMI levels at the end of 2016.

 

We then had a year's worth of data for them throughout 2017,

 

and we tracked all of their healthcare interactions

 

and expenditures over the course of 2017,

 

and we deciled them.

 

Decile one being the most expensive or highest cost segment,

 

and decile 10 being the lowest.

 

The 80-20 rule does hold true,

 

so you'll know that the smaller deciles

 

are in the higher costs of one, two, three, four, and five,

 

and the larger segment memberships are

 

in deciles eight, nine and 10.

 

We then looked at their clinical metrics at the end of 2017,

 

to evaluate what the change was in those metrics

 

and how it related to cost.

 

We show that there's a positive correlation

 

between higher spend

 

and better clinical outcomes moving down

 

and to the left on the graphic.

 

But it is far from perfect correlation.

 

Just opens up a wide range of questions

 

and potential future analytic opportunities.

 

Imagine for a moment doing a comparative study

 

and being able to control for a wider range of confounders

 

and develop metrics on the most cost

 

and health outcome-effective therapies.

 

Imagine for a moment being able

 

to show a payer

 

your more cost-effective

 

and have a greater impact on clinical outcomes.

 

Imagine then being able to take that information

 

and build better pathways

 

to get a more impactful management of population health,

 

and a lowering of costs simultaneously.

 

We're going to get a little heavier

 

into the clinical metrics associated

 

with integrated data in this next example.

 

And our goal is to really evaluate

 

how integrated data can gain better insight

 

into the impacts of interventions.

 

Repeated health measures are a integral part

 

of the integrated data.

 

And they very succinctly allow us

 

to evaluate interventions based on

 

the actual elements of interest.

 

Let's take the case of bariatric surgery.

 

In the example in the lower left hand graphic

 

on your screen,

 

we have segmented bariatric surgery patients

 

over a one year period after surgery,

 

based on their BMI pre

 

and through out the entire year.

 

We know that the average cost

 

of a bariatric surgery runs about $27,000,but in this analysis we found that 8% of those surgeries

 

and 8% of the patients that received them,

 

saw no appreciable weight loss in the course of a year.

 

Imagine for a moment that we could build that

 

into some type of monitoring

 

to help alter that performance trajectory

 

and perhaps even change the performance metrics

 

to only reimburse physicians when we saw a positive gains

 

in healthcare outcomes related to the surgical intervention.

 

Let's take it a step further.

 

We subset these patients to just type 2 diabetics.

 

And we examined how they were able to manage

 

their type 2 diabetes during

 

that one year period post-bariatric surgery.

 

In the example on the right hand part of our screen,

 

we found that 64% of the patients

 

that were type 2 diabetic

 

and had bariatric surgery,

 

also had uncontrolled diabetes

 

during that one year period.

 

They also cost on average three times more than

 

the patients in the lower right hand part of that graphic,

 

that had the best performance in both weight loss,

 

and were better able to control their diabetes.

 

We've found that there were many differences using

 

the EHR data in healthcare engagement

 

between those groups as well.

 

Differences in engagement with providers,

 

diet, and exercise,

 

and the integration of the data allowed us

 

to see all of that.

 

And allows us now, to set to identify those issues,

 

evaluate the performance of these interventions,

 

set up new interventions, policies and protocols

 

to improve patient care overall.

 

A fourth used case is more targeted

 

towards life sciences companies,

 

and one of the greatest challenges that they face

 

in evaluating and understanding the investments

 

and the return that they they've gotten

 

from those investments in terms of product performance.

 

These companies spend millions,

 

hundreds of millions of dollars developing drugs,

 

and ultimately more marketing them

 

and bringing them into a space where we can utilize them

 

to improve the quality of life for our patients.

 

The problem is that monitoring many

 

of these markets is difficult because

 

the data itself resides in different silos.

 

Let's take the case of "The Immunology Space".

 

Truly understanding market share on a weekly

 

or even a monthly basis is difficult because

 

while many of the products go through the pharmacy channel,

 

still many others are administered in a physician's office

 

and paid through through the medical benefit.

 

Those two claims streams are independent

 

and they have different time lags associated with them.

 

Integrating claims and electronic health record data,

 

allows organizations to gain clarity in their markets

 

by combining all of those products together

 

in a single view in near real-time.

 

As I mentioned,

 

those medical claims tend to lag in some cases,

 

days, weeks, and even months,

 

which means that you can't evaluate properly market share

 

or product performance until all of

 

that information comes into play.

 

EMRs, and in this case EHRs,

 

are updated on a more routine basis and as a result,

 

we can integrate all of the data together

 

to give us all of the medically administered products

 

on the same case as the pharmacy products

 

and demonstrate market share in real-time.

 

Is to also opens up on the next slide,

 

the ability to track other things with the data as well,

 

looking at indications,

 

and why products are being prescribed from that perspective.

 

This type of market clarity on a weekly basis

 

can enable better reaction to competitive threats

 

and maximizing new opportunities for Life Science Companies.

 

Our last example

 

is all about tactics.

 

I spoke earlier about the resources that you need.

 

You need to be able to change them

 

as you work with integrated data.

 

And it's likely the case today

 

that your organization may not yet be at a point where

 

you can tactically implement programs, pathways,

 

or other strategies using integrated data.

 

It may be that you only have claims data

 

to leverage for your operations,

 

or perhaps only clinical data.

 

If you recall in my earlier in my presentation,

 

I showed you a graphic of the different types

 

of data found in EHR and claims,

 

and I noted a group of variables that are in common to both.

 

One of the advantages of integrated data is

 

you can build models using that common subset

 

to move from the strategic to the tactical,

 

even if you don't yet have the infrastructure in place

 

to leverage integrated data at scale for operations.

 

As we round this out,

 

I searched for a comment on healthcare data and change,

 

but was unsuccessful.

 

So I defaulted to one of my favorite quotes

 

about Charles Darwin.

 

And it really is,

 

is ultimately focused on change

 

and our ability to survive by being able to adapt to change.

 

As we are well aware,

 

as we are developing new models

 

and new methods to solve problems in healthcare,

 

integrated data will help fuel that change

 

and enable us to realize greater efficiency

 

and insight generation.

 

Imagine the ability to address value based

 

on clinical values and costs simultaneously.

 

Think of how we could develop new value based contracts,

 

reimbursement models and care pathways

 

all using one integrated source of data.

 

It is the future of healthcare analytics.

 

I hope you've found today's session informative,

 

and I thank you for your time.

 

(bright upbeat music)

Text

Integrated data is the driver of change

Integrated data can help us realize greater efficiency, new reimbursement models and more proactive care pathways for better outcomes. Join Lou Brooks, vice president of Commercial Analytics at Optum, as he discusses how integrated data is ushering in the future of health care analytics.

Icon Picker V2
""