Joshua Rosenthal, PhD and Elizabeth Young
Government Data and You:
Using Public Use Files to Guide Value-Based Care
Decisions and Deliver High-Value Care
How Medicare Public Use Files, along with a variety of
additional Public Use Files, can be used to provide strategic
insights for health plans, physician groups, and hospital
systems looking to provide high-value care.
Background on the currently available datasets and their
strengths and limitations with the goal of inspiring new
avenues of workforce research.
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Context: Public Data, Public-Good Goals
Overview: Uses and Examples
Trying Your Hand: Sources and Samples
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Context: Public Data, Public-Good Goals
Overview: Uses and Examples
Trying Your Hand: Sources and Samples
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Government is releasing lots of data*
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Need market to adopt – use to create value
Cf. Weather & Geo-Location data
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Context
Exit
Exit w/ Good Multiple
Idea
Prototype
Funded
1 %*
1 %*
1 %*
1 %*
*Health Care Start Ups fail at astounding,
disproportionate rates
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Most ‘Successes’, Aren’t
Measuring Value Creation
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Measuring Value Creation
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
User Guide (ALT)
Consultants /Professional Services,
Providers
SaaS-baseddata / analytic
platforms
Measuring Value Creation
Joshua Rosenthal, PhD and Elizabeth Young
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Health care has not done so well
“Just wait ‘till next year Financial Services & Energy!”
In this race vs. other verticals
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
*Direct-to-Consumer Note: People don’t like to pay out of pocket for something they don’t like to do or don’t want to know about
Challenge: Direct to Consumer Apps
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Speaking at a major health care conference near you
Challenge: Fluff in nice suits
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
I have better engineering / architecture
Hmm, ‘fixing the pipes’was not the answer
Challenge: Noise from Tech
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
I have better design & experience
Hmm, the pretty colors on my social app didn’t stop me
Challenge: Bubbles (design)
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
My data is bigger than yours
Challenge: Buzz (cf. big data)
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
* Source: Dartmouth Atlas for Unwarranted Variation
Challenge: Perverse Incentives
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Perverse Incentives: Fee for Service
Payers aggregate –but some have not historically been “health care”
“Actually, I make more money off of bad drivers.”*
(Read w/ accent)
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Some Hospitals/Providers may historically generated revenueby filling rooms* Keep patients away?!?
I was trying to book you for an extra night!
Perverse Incentives: Fee for Service
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Perverse Incentives: Disrupting Status Quos
David Wennberg, MDRowdMap Advisory Board
Open Health Data Outperforms Claims and Electronic / Personal Health Records, & Devices for Risk & Cost
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Challenge: Interpretation of Data
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Causal?
Then again… Maybe there’s something to it?
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Review
Health Care Start Ups fail at astounding,
disproportionate rates
FAIL
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Solution: NEW INCENTIVESGovernment Announces Sun-setting of Fee for Services
Government Announces New Pay-for-Value Models
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
New Government Released Referral Data(Patient flows between PCPS, specialists, hospitals and post acute centers)
Dartmouth Atlas for Unwarranted Variation(Decades of research and data on unwarranted variation by condition and geography to keep things apples-to-apples for comparisons, hence “Unwarranted” in the name)
New Government Released Performance Data (Individual providers, groups, hospitals and post acute centers including the new part B&D)
Provider Pattern Intensity Profiles and Risk Readiness for every provider, hospital, post acute center in the US. All preloaded with no IT.
Solution: NEW DATA
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Answering the Open Challenge:Connecting Open Data to an Meaningful Business Model
DIY it!
Guide: http://tinyurl.com/l2nxvw9
Sources: http://tinyurl.com/ko7jnfa
Game: http://tinyurl.com/kspxr4e
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Context: Public Data, Public-Good Goals
Overview: Uses and Examples
Trying Your Hand: Sources and Samples
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
CMS: 50% of FFS will be gone by 2018
The business context has changed. Health plans, government payers, providers, and hospital systems need
to develop Risk-Readiness strategies to excel as they transition from fee-for-service to pay-for value.
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
PUF Files Featured NationallyUS CTO on RowdMap: “Visionary
Genius”
What you can do [without a PhD]With mashups of gov’t data (CMS HHS, Gov, CDC)
Chronic prevalence & physician supplyPopulation Health ReportPopulation Report Card
Match practice patterns to the right risk arrangements – PFV ReadinessGroup Risk-Readiness SM ReportPhysician Risk-Readiness SM ReportHospital Risk-Readiness SM ReportPost Acute Center Risk-Readiness SM ReportRisk-Readiness SM Arrangement Match-Maker
Manage clinical care and costs –Remove No Value CareGroup Unnecessary Cost ReportPhysician Unnecessary Cost ReportHospital Unnecessary Cost ReportPost Acute Center Unnecessary Cost ReportUnnecessary Cost Referral and Value Chain Report
What you can do [without a PhD]With mashups of gov’t data (CMS HHS, Gov, CDC)
Chronic Prevalence &Physician Supply
Match Practice Patterns to the right Risk Arrangements – PFV Readiness
Manage Clinical Care and Costs –Remove No Value Care
Diabetes Prevalence -Westchester
Use this data to allocate providers and care management resources around condition-specific population needs by zip.
Locate clinics, health fairs, etc. based on chronic needs.
Income
Obesity
Depression
Health Opportunity Index
Demand and Supply
What type of populations?
Medicare FFS Geo. Variation: http://go.cms.gov/1D8j7LECDC Behavioral Risk Factor Surveillance: http://1.usa.gov/1PzcisTMedicare FFS Part B: http://go.cms.gov/OCmyoyMedicare FFS Part D: http://bit.ly/1mGyBxk
Lots of diabetics but few PCPs
Lots of diabetics and lots of PCPs
PCP Density –Westchester
32
Demand and Supply
County ProfilesLargest Counties In Ohio
People use this data to calibrate expectations for profitability by incorporating population health and provider performance into
product strategy. Use excess to subsidize operations in counties with fewer high-performing resources
Risk ScoresTotal Cost
PMPM Reimbursement Overall Star
Chronic Star
Health Rank
MA Profit Opportunity -
MA
Profit Opportunity -
Exchange
MA Eligibles
MA Enrolled
Exchange Subsidized
ExchangeEnrolled
Compare to National and Regional Benchmarks
Medicare FFS Geo. Variation: http://go.cms.gov/1D8j7LECDC Behavioral Risk Factor Surveillance: http://1.usa.gov/1PzcisTMedicare FFS Part B: http://go.cms.gov/OCmyoyMedicare FFS Part D: http://bit.ly/1mGyBxk
What you can do [without a PhD]With mashups of gov’t data (CMS HHS, Gov, CDC)
Chronic Prevalence &Physician Supply
Match Practice Patterns to the right Risk Arrangements – PFV Readiness
Manage Clinical Care and Costs –Remove No Value Care
At the core of Risk-Readiness SM is
Unwarranted Variation:
RowdMap applies the Dartmouth Atlas for Unwarranted Variation methodologies to data on Medicare Parts B & D. This research has been repeatedly validated over the last 30 years and we now have a national data set to apply the methodologies at a large scale.
The estimated 30% of medical expense that goes to unnecessary care. This unnecessary spend drives billing in a fee-for-serve economic model, but success in pay-for-value comes from managing and mitigating these pockets of variation.
Every provider has a unique practice pattern that informs Risk-Readiness SM
Pay for Value Readiness
Los Angeles, CA
Compare to National or Regional Benchmarks
Pay for Value Readiness
Provider Profiles
Identify highly efficient, Risk-Ready practices and physicians to profitably grow into. Improve profitability of lower performing practices with large panel sizes through modified arrangements
or performance improvement plans.
Medicare FFS Part B: http://go.cms.gov/OCmyoyMedicare FFS Part D: http://bit.ly/1mGyBxkReferrals: http://1.usa.gov/1FzoEOV
Identify high and low performing hospitals and post-acute facilities— are there post acute facilities that hospitals with poor chronic
readmits are routing members to?
Pay for Value Readiness
Dartmouth Atlas: http://bit.ly/1GXvlJpCMS Hospital Compare: https://goo.gl/p8MtoICMS Hospital Readmissions: http://goo.gl/02KnQdCMS Nursing Home Compare: https://goo.gl/3DpT8m
EOL Hosp Days: Which hospitals fewer end-of-life days than their peers?
Chronic Admits: Which hospitals see their most chronic population repeatedly/ with the most frequency?
Cardiac Imaging: Which hospitals are more likely to over-utilize cardiac imaging compared to their peers?
Pay for Value Readiness
Great profile for aggressive risk
Tread carefully for some risk
Match appropriate risk arrangements based on provider practice patterns and Population characteristics within a geography.
What you can do [without a PhD]With mashups of gov’t data (CMS HHS, Gov, CDC)
Chronic Prevalence &Physician Supply
Match Practice Patterns to the right Risk Arrangements – PFV Readiness
Manage Clinical Care and Costs –Remove No Value Care
Low-Value Care looks at a different category of spending
Shift focus from clinical edits, audits, and recovery efforts to identifying care that is clinically appropriate, but unnecessary. Historical efforts have shown returns, but they only look at a fraction of total spending. Unnecessary care can account for up to 30% of total spending and provides significantly larger opportunities for cost containment and quality improvement.
Clinically Appropriate, but Unnecessary Care
(30% of spend)
Claims Spend for a Health Plan
Necessary Utilization(70%)
“It’s generally agreed that about 30 percent of what we spend on health care is unnecessary. If we
eliminate the unneeded care, there are more than enough resources in
our system to cover everybody.”
-Dr. Elliott Fisher,Dartmouth Institute for Health Policy
Remove Low-Value CareManage Unnecessary Spend
Remove Low-Value CareManage Unnecessary Spend
RowdMap tackles the 30% of the U.S. health care spend that goes to clinically appropriate, but unnecessary care
Over $9B in Orange County, CA
How much unnecessary spend is in your market?
Over $66B in Florida
$850 Billion Unnecessary Spend* in 2014
Least Unnecessary Spend
Most Unnecessary Spend
RowdMap tackles the 30% of U.S. health care spend that goes to clinically appropriate, but unnecessary care. RowdMap’s models identify the cost-savings opportunities in a geography based on the collective intensity of care delivered by doctors in that area.
* Unnecessary Spend =
(Dartmouth Avg cost) * (Population) * (RowdMap Network Opportunity Index)
Unnecessary Spend in Florida
In Broward Co. alone, there is over $7.6B in unnecessary spend.
Let’s look at which hospitals, groups and physicians account for this and for what conditions
Remove Low-Value CareManage Unnecessary Spend
Physician Marketshare by Major Clinical Categories
Match appropriate risk arrangements based on provider practice patterns and Population characteristics within a geography.
Hospital Marketshareby Major Clinical Categories
Provider Group Marketshareby Major Clinical Categories
Unnecessary Spend in BrowardBy condition across hospitals, groups and physicians
This Physician.Let’s start here
This GroupThis Hospital
Circulatory
Muscular-skeletal
Respiratory
Remove Low-Value CareManage Unnecessary Spend
Physicians Driving Unnecessary Care in Broward
Musculoskeletal care is major contributor to unnecessary spend in Broward. Let’s take a physician who is not an outlier but in the middle of the pack such as Dr. Spend*. Let’s walk through what his clinically acceptable, but medically unnecessary, practice pattern creates
in unnecessary spend.
Remove Low-Value CareManage Unnecessary Spend
Referral Patterns and Physician Value Chains
Identify high performing providers and downstream referral patterns. Encourage referrals to
high-performing specialists.
Remove Low-Value CareManage Unnecessary Spend
Least Unnecessary Spend
Most Unnecessary Spend
Option 2: Reinforce highest-performing referral
and care pathways. Increase the number of patient interactions
with green dot doctors.
Option 1: Change provider behavior. Requires lots of provider education. Requires
payer to make up a significant portion of a provider’s revenue. Increase the number of
green dot doctors.
Zoom to zip
Remove Low-Value CareManage Unnecessary Spend
If had same ratio as :
• His decompression rate would drop from 6.01 to 0.436 per patient.
• Which translates to 2,608 fewer decompressions per year.
• At an average cost of $332 per decompression, this represents potential savings of over $850K
If decompression to fusion rate were average for orthopedic surgeons:
• He would have 1629 fewer decompressions for a potential savings of $540K.
*Actual physician names have been changed.
For every 10 back fusions, does 103 decompressions
For every 10 back fusions, does 2 decompressions.
Dr. Save*
Dr. Spend’s
Dr. Spend*
Dr. Save*
That’s one physician, with one procedure, in one clinical condition. This savings would not be picked up in unit cost or utilization analysis,
but cumulatively dwarfs fraud, waste and abuse outliers. Intense practice patterns like this power FFS arrangements
but success in Pay for Value comes from identifying Risk-Ready providers.
Dr. Spend*
Remove Low-Value CareManage Unnecessary Spend
Government Data and You:
Using Public Use Files to Guide
Value-Based Care Decisions
and Deliver High-Value Care
Joshua Rosenthal, PhD and Elizabeth Young
Context: Public Data, Public-Good Goals
Overview: Uses and Examples
Trying Your Hand: Sources and Samples
TRY YOUR HAND AT GOV DATA
Before data products, tools, widgets, benchmarks, APIs, systems…
Before access and security; before interoperability, standards…
Before automation and efficiency…
Taxonomy
… basis for moving to new paradigm from fee-for-service to pay for value
… bridges business & technology/data… bridges internal & external users
... across CMS… identifiable beneficiary databases and non-beneficiary files
Helps to get started:
TRY YOUR HAND AT GOV DATA
http://www.medicare.gov/download/downloaddb.asp
Example Taxonomy
TRY YOUR HAND AT GOV DATA
Sample record from the downloaded file
What is this? How can I use it to answer business questions?
Example Taxonomy
TRY YOUR HAND AT GOV DATA
Example Taxonomy
In order to get insight, I need the data in a meaningful business structure!
TRY YOUR HAND AT GOV DATA
Taxonomy defines business entities and the relationships among them
TRY YOUR HAND AT GOV DATA
Taxonomy defines attributes for business entities
TRY YOUR HAND AT GOV DATA
Taxonomy at the center of any system
If you want to do something meaningful with the data, go through taxonomy
TRY YOUR HAND AT GOV DATA
[Referrals: http://1.usa.gov/1FzoEOV ]
[Variation: http://go.cms.gov/1D8j7LE ]
[Shared Savings: http://go.cms.gov/1Hh8vx0 ]
[Medicare FFS Part B: http://go.cms.gov/OCmyoy ]
[Medicare FFS Part D: http://bit.ly/1mGyBxk ]
[Medicaid: http://go.cms.gov/1z7b5ic ]
[Dartmouth: http://bit.ly/1GXvlJp ]
[Behaviors: http://1.usa.gov/1PzcisT ]
[Health Data All Stars: http://bit.ly/1GAsVC3 ]
Places to Start: