Is Big Data a Big Deal...or Not?

© 2015 Health Catalyst
www.healthcatalyst.com
Creative Commons Copyright
cIs Big Data a Big Deal… or Not?
Dale Sanders
EVP, Product Development
Health Catalyst
February 2016

Creative Commons Copyright 2
Agenda
Destination
Where are
we now?
What should we
do?
• Where should we be
headed with healthcare IT,
in general, and how does
Big Data fit into that?
• What are the guiding
concepts, influences
and vision?
• Does healthcare data
require Big Data now?
• If not, when?
• So, what should we do
and when?
• What will it look like?

www.healthcatalyst.comCreative Commons Copyright
Where are we headed?

What’s the real question we should ask?
• Is Big Data a big deal or not?
Or maybe…
• Is the Cloud a big deal or
not?
• The Cloud is making Big
Data accessible, affordable,
and transparent for everyone
4

• Commoditization is going up, complexity is going down
• At some point, we should stop caring about the technology below the
Data Content layer and just concern ourselves with the services above it.
In the Cloud

c
6
For better and worse,
software has overtaken the
impact of heroic leadership
as the greatest agent of
change in human behavior.
We have to build software
that deliberately borrows
lessons from the software
that has changed human
behavior.

c
7
The Facebook and Amazon EMR
From a blog I wrote in 2010

c
8

c
9

90% of the screen space is driven dynamically,
by context, through analytics and algorithms
in the background that are nudging your
decisions through suggestive analytics
& collective intelligence
10

At Health Catalyst, I’m asking the same thing you
are: To what degree do we go ‘Big Data’ and when?
To be or not to be…

Relational vs.
Hadoop Analytic Technology
"Technology Life Cycle". Licensed under CC BY-SA 3.0 via Wikimedia Commons –
https://commons.wikimedia.org/wiki/File:Tecnology_Life_Cycle.png
0 Time
BusinessGain
Vital Life
R&D
The Technology Lifecycle Path
A D
L
M Maturity
Hadoop
Relational
databases for
analytics

What is Big Data, anyway?
15

What does it take to reach the Big Data threshold?
What are the numbers?

Let’s look at Health Catalyst’s numbers
What are the numbers in
healthcare?

We Are Not “Big Data” in Healthcare Yet
19

Just Beginning: Digitization of Health
20
“EMR data represents ~8% of the data we need for population health and precision
medicine.” -- Alberta Secondary Use Data Project
The Growing Ecosystem of Human
Health Data
Healthcare
Encounter
Data
7x24
Biometric
Data
Consumer
Data
Genomic
&
Familial
Data
Social
Data
Outcomes
Data

Volume, Velocity, and Variety aren’t the only reasons to move
Dear Data…

The Three Loops of Clinical Decision Support

Physicians are 15x more likely to change
their ordering and treatment protocols if
presented with substantiating data at the
point of care vs. presented with the same
data in a clinical process improvement
meeting.
Kawamoto, et al, BMJ, 2005
24

We are evolving from an offline, data
aggregator and analysis company, to a
real-time data production and decision
support company, integrating the
knowledge derived from analytics into the
workflow of our clients and their patients,
wherever that decision workflow occurs.
25

There’s not a cultural change problem
among physicians in healthcare. There’s a
software and data problem. We send
physicians out to drive without a
speedometer– while CMS, insurance
companies, and administrators have a
radar gun-- then we penalize them when
they drive too fast or too slow.
26

The Right Data
To The Right Person
At The Right Time
In The Right Modality
27

Where are we now?

Oracle, Microsoft, IBM might as well start giving their databases
away for free
You get a lot from the community

Pure open source Big Data platform
The Hortonworks Platform

Gartner Survey 2015
32
Hadoop Investment Plans

Obstacles to Hadoop adoption

• Health Catalyst Analytics Platform
• Operations & Performance Management
• Financial Decision Support
• Clinical Analytics & Decision Support
• Research Informatics
• Precision Medicine
• Population Health & Accountable Care
• Care Management & Patient Relations
• Comparative Analytics (CAFÉ)
• Collective Intelligence
34
Our New Product Lines

c
35
We are going to build
software that supports
this aspirational
statement from a
physician to a patient:
“I can make a health
optimization recommendation
for you, informed not only by
the latest clinical trials, but also
by our local and regional data
about patients like you; the
real-world health outcomes over
time of every patient like you
who has had your illness; and
the level of your interest and
ability to engage in your own
care -- and in turn I can tell you
within a specified range of
confidence, which treatment
has the greatest chance of
success for a patient
specifically like you and how
much that treatment will cost.”
Training Data
Machine
Learning
Algorithm
HypothesisTest Data Performance
Acknowledgements to the Learning Health System

Patient Flight Path Mapped to vision statement…
11
“I can make a health optimization recommendation for you, informed not only by the latest clinical trials, but also by local and regional
data about patients like you; the real-world health outcomes over time of every patient like you who has had your illness; and the
level of your interest and ability to engage in your own care – and in turn, I can tell you within a specified range of confidence, which
treatment has the greatest chance of success for a patient specifically like you and how much that treatment will likely cost.”

The House of Health Catalyst

The Health Catalyst Big Data strategy
39

Selection Criteria
1. Total Cost of Ownership
2. Flexibility
3. Scalability
4. Security
5. Reliability
40

The Microsoft Evaluation
Factors Health Catalyst Score Microsoft Score
Affordability  TCO $1.75m (with HC labor)
 Cost Spikes when new
hardware is needed or
hardware needs to be
replaced
9  TCO: $XXX (with HC labor)
 Competitive price pressure
drives costs down over time.
No need to overprovision
capacity.
10
Scalability  Cyclical h/w purchase model
 No regional deployment
5  Virtually unlimited total capacity
available on demand.
 29 Regions WW (lower latency,
get the data closer to the
customer)
10
Flexibility  Finer control of allocations of
RAM/CPU to a given
customer within existing
hardware deployment.
10  VMs at prescribed sizes, with
new sizes getting added all the
time.
8
Security  Security strong but clamped
down, fixed, and manual
6  Trust Center
 Compliance
 DDoS Detection
 Security Center
 SDLC
10
Reliability  ~99% SLA 8  99.9%+ SLA
 Financially backed SLA
10
• 10 point
must
system

Rea life architectural example
45

Q & A

Questions submitted prior to webinar
Debbie Sweet:
I’m hoping that you are planning to (or can) discuss the
appropriateness of a big data or Hadoop data store for different types
of healthcare data. One thing I hear from IT people at health care
organizations is that they feel like the business/clinical folks hear “big
data” or Hadoop and think that is the panacea and everything in their
data warehouse should move to Hadoop. My understanding and
experience with this is that there are times when this data structure is
helpful and times when there won’t really be much net gain from
moving from RDBMS to Hadoop.
47

Questions submitted prior to webinar
Eric Cole:
Where do you see big data having the greatest positive and negative impacts for
health care organizations?
Do you anticipate individual clinics and hospitals will be motivated by the potential
benefits of combining big data with machine learning types of applications?
What challenges do health care organizations face when considering adoption of
techniques like machine learning and big data?
I can imagine how the Internet of Things can be a factor in the future. Not only will we
have the regular data sources to contend with but also many intelligent devices will be
collecting a wide variety of types of data. Potentially, useful data. Do you think as the
volume of data being collected continues to grow, potentially dramatically, is going to
motivate adoption of big data technologies in health care organizations?
48

Thank You
49

Is Big Data a Big Deal...or Not?

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to Is Big Data a Big Deal...or Not? (20)

More from Health Catalyst (20)

Recently uploaded (20)

Is Big Data a Big Deal...or Not?