SlideShare una empresa de Scribd logo
1 de 23
Big Data in Health Care
What Marketers Need to Know
Tim Gilchrist, May 2014
@timgilchrist
Session Goals
• Cover
– Big Data
– Artificial Intelligence / Machine Learning
– How to be an Informed Consumer
– Applications for Marketers
– Questions
2
Big Data
The term for a collection of data sets so large and
complex that they become difficult to process
• Many data sources with different formats
• Data with missing values
• Text / Social Media
• Things that don’t fit in Excel
3
Artificial Intelligence
4
“Pay no attention to the man behind the curtain”
Machine Learning
The construction and study of systems that can
learn from data
5
Bayes
Thomas Bayes (1701 – 7 April 1761) was an English
mathematician and Presbyterian minister, known for
formulating the theorem that bears his name: Bayes'
theorem
6
Bayes theorem uses prior probabilities, combined with new observations to
calculate the probability of a hypothesis being true or false
Bayes is a natural fit to health care due to the presence of hypothesis
(diagnosis) and events (tests / observations)
Example
You are all doctors who have administered a
critical test to 50 patients
You know the test is:
– 75% Accurate
– 10% False Positives
7
– 10 of your patients tested positive
– How many are actually sick?
Bayes’ Theorem – Mammogram Example
We can present the data as a decision tree representing the
probabilities confronting doctors and patients
8
If we were to take population-level down to the individual level, much more accurate
probabilities would be possible
#1 Mammogram
#2Biopsy#3Time
Observation
Probability
What Does this Mean To Marketers?
• Big data is about discovering relationships
9
“Can’t beat a man with some insurance.
I need that health plan baaaby!”
• Then using data-driven insights to inform
strategy
Big Data Health Landscape
HIE
Member PCP Specialist
Sees PCP Gets X-Ray Sees Specialist Ambulatory
Outpatient
Analysis / Transformation
EMR
Admission
Discharge
PrescriptionClaims
Plan Data Portal
CarePathOutputsDataTypes
Direct
ConnectionWearable
Telemetry
What is Happening What Will Happen
Social PurchasedLocation
Cell
Example
Text Mining for Sales
11
Listening & Collecting
12
Signal
Training / Processing
13
Tweets extracted from
the Twitter Fire Hose with
key words “Health” and
“Plan”
1MM per day
1
@CapoeiraBatuque
“What's the best plan thru
affordable health care. Blue
cross? Blue shield? Health
net? #confused #healthcare”
@bluecalgal
Obama says don't listen to Fox,
why? Obama lied about keep
your dr, health plan, cheaper
than cell phone (bs) keep your dr
Create Training databases
for any classification desired.
+ and – outcomes used here
2
Weka turns Tweets into
numerical code that can be
analyzed by computer
“String to Word Vector”.
uses Naïve Bayes classifier
3
“What's the best plan thru affordable health care. Blue cross?
Blue shield? Health net? #confused #healthcare”
.225.357
.155
.999
ACA
AccessHealth
HealthPlanNow
Pressure
Confused
Dumb
@ Handle
# Followers
# Tweets
Profile
Retweets
Location
Date/Time
Text Analysis
Stemming/Tokenization
Demographic
Analysis
Once classifications are
established, rules can be
applied to new Tweets with
high accuracy ~90+%
4
Result
14
Other Uses
15
• Model who will be your most
valuable:
• Customer
• Facebook follower
• (Really) Determining Sentiment
• Marketing Mix Simulations
• Consumer Facing Predictive
Technology
• Prod development (HIX)
Questions?
16
Appendix
17
Discovering Relationships Between Data
We can use machine learning to form
relationships between sets of data that
are seemingly unrelated (Causal
Relationships):
• Making your bed in the morning and job
satisfaction
• Artificial Christmas trees and family “brag
letters”
• What you buy and why
18
A Causal Diagram Based on Established Relationships for Estimating
the Incidence of Coronary Heart Disease (CHD).
(Source: Comparative quantification of health risks: Conceptual framework and methodological issues)
19© Tim Gilchrist 2013
Bayes’ Theorem – Mammogram Example
Problem: Estimates of breast cancer over diagnosis range from 25%–
52%. Physicians often misinterpret their own lab results
20
Prior Probability
Chance that a woman will develop breast cancer in her 40s X 1.4%
New Event: Mammography
Ability of mammogram to detect cancer when present Y 75%
False positives Z 10%
Posterior Probability
Revised Probability, given new event (positive mammogram)
xy+z(1-x)
9.6%
A positive mammogram still leaves a 90.4% chance that the test showed something other than
cancer. When biopsies are performed on this age group, 75% are negative. Physicians rarely
consider the other 90.4%
What Does this Mean To Marketers?
21
Other Reading
Why Most Marketers Will Fail In The Era Of Big
Data
8 Marketers Doing Big Data Right
The big-data revolution in US health care:
Accelerating value and innovation
HITLAB Speaks with Tim Gilchrist, Director of
eBusiness Strategy for WellPoint
22
The State of Big Data
23

Más contenido relacionado

La actualidad más candente

La actualidad más candente (20)

444 jumping the fence of information sharing with patients
444 jumping the fence of information sharing with patients444 jumping the fence of information sharing with patients
444 jumping the fence of information sharing with patients
 
Health 2.0 WinterTech: Will Artificial Intelligence change healthcare? by Eug...
Health 2.0 WinterTech: Will Artificial Intelligence change healthcare? by Eug...Health 2.0 WinterTech: Will Artificial Intelligence change healthcare? by Eug...
Health 2.0 WinterTech: Will Artificial Intelligence change healthcare? by Eug...
 
Big Data Analytics for Smart Health Care
Big Data Analytics for Smart Health CareBig Data Analytics for Smart Health Care
Big Data Analytics for Smart Health Care
 
Utilising Smartphones to Aid in the Treatment and Management of Inflammatory...
Utilising Smartphones to Aid in the Treatment  and Management of Inflammatory...Utilising Smartphones to Aid in the Treatment  and Management of Inflammatory...
Utilising Smartphones to Aid in the Treatment and Management of Inflammatory...
 
Big data revolution in healthcare
Big data revolution in healthcare Big data revolution in healthcare
Big data revolution in healthcare
 
The big data revolution in healthcare
The big data revolution in healthcareThe big data revolution in healthcare
The big data revolution in healthcare
 
1 10 everyday reasons why statistics are important
1   10 everyday reasons why statistics are important1   10 everyday reasons why statistics are important
1 10 everyday reasons why statistics are important
 
Data drift and machine learning
Data drift and machine learningData drift and machine learning
Data drift and machine learning
 
Driving Healthcare Operations with Data Science
Driving Healthcare Operations with Data ScienceDriving Healthcare Operations with Data Science
Driving Healthcare Operations with Data Science
 
Building a Data Warehouse at Clover
Building a Data Warehouse at CloverBuilding a Data Warehouse at Clover
Building a Data Warehouse at Clover
 
Big data in healthcare
Big data in healthcareBig data in healthcare
Big data in healthcare
 
Data drift and machine learning
Data drift and machine learningData drift and machine learning
Data drift and machine learning
 
Introduction to machine_learning_us
Introduction to machine_learning_usIntroduction to machine_learning_us
Introduction to machine_learning_us
 
DSPA Insights Conference 2019
DSPA Insights Conference 2019DSPA Insights Conference 2019
DSPA Insights Conference 2019
 
Will Artificial Intelligence Change Healthcare?
Will Artificial Intelligence Change Healthcare?Will Artificial Intelligence Change Healthcare?
Will Artificial Intelligence Change Healthcare?
 
Longevity drug development is just getting started, and data is ready to lend...
Longevity drug development is just getting started, and data is ready to lend...Longevity drug development is just getting started, and data is ready to lend...
Longevity drug development is just getting started, and data is ready to lend...
 
How artificial intelligence ai assist in medicine, an example of diffrent dev...
How artificial intelligence ai assist in medicine, an example of diffrent dev...How artificial intelligence ai assist in medicine, an example of diffrent dev...
How artificial intelligence ai assist in medicine, an example of diffrent dev...
 
Big data in Healthcare & Life Sciences
Big data in Healthcare & Life SciencesBig data in Healthcare & Life Sciences
Big data in Healthcare & Life Sciences
 
Supporting Innovation and Evidence-based Communication based on Best Practice...
Supporting Innovation and Evidence-based Communication based on Best Practice...Supporting Innovation and Evidence-based Communication based on Best Practice...
Supporting Innovation and Evidence-based Communication based on Best Practice...
 
Building a Data Warehouse at Clover (PDF)
Building a Data Warehouse at Clover (PDF)Building a Data Warehouse at Clover (PDF)
Building a Data Warehouse at Clover (PDF)
 

Similar a Big data chicago v2 5 14 14

Similar a Big data chicago v2 5 14 14 (20)

Machine learning applied in health
Machine learning applied in healthMachine learning applied in health
Machine learning applied in health
 
Big Data Provides Opportunities, Challenges and a Better Future in Health and...
Big Data Provides Opportunities, Challenges and a Better Future in Health and...Big Data Provides Opportunities, Challenges and a Better Future in Health and...
Big Data Provides Opportunities, Challenges and a Better Future in Health and...
 
The Philosophy, Psychology, and Technology of Data in Healthcare
The Philosophy, Psychology, and  Technology of Data in HealthcareThe Philosophy, Psychology, and  Technology of Data in Healthcare
The Philosophy, Psychology, and Technology of Data in Healthcare
 
Health IT Summit Austin 2013 - Presentation "The Impact of All Data on Health...
Health IT Summit Austin 2013 - Presentation "The Impact of All Data on Health...Health IT Summit Austin 2013 - Presentation "The Impact of All Data on Health...
Health IT Summit Austin 2013 - Presentation "The Impact of All Data on Health...
 
Digital healthcare show - How will Artificial Intelligence in healthcare will...
Digital healthcare show - How will Artificial Intelligence in healthcare will...Digital healthcare show - How will Artificial Intelligence in healthcare will...
Digital healthcare show - How will Artificial Intelligence in healthcare will...
 
Sun==big data analytics for health care
Sun==big data analytics for health careSun==big data analytics for health care
Sun==big data analytics for health care
 
Data analytics: Are U.S. hospitals late to the party?
Data analytics: Are U.S. hospitals late to the party?Data analytics: Are U.S. hospitals late to the party?
Data analytics: Are U.S. hospitals late to the party?
 
Promise and peril: How artificial intelligence is transforming health care
Promise and peril: How artificial intelligence is transforming health carePromise and peril: How artificial intelligence is transforming health care
Promise and peril: How artificial intelligence is transforming health care
 
Data in Healthcare Community Response.docx
Data in Healthcare Community Response.docxData in Healthcare Community Response.docx
Data in Healthcare Community Response.docx
 
Big Data in Healthcare and Medical Devices
Big Data in Healthcare and Medical DevicesBig Data in Healthcare and Medical Devices
Big Data in Healthcare and Medical Devices
 
Health Data Innovation (Wolfram Data Summit)
Health Data Innovation (Wolfram Data Summit)Health Data Innovation (Wolfram Data Summit)
Health Data Innovation (Wolfram Data Summit)
 
Rock Report: Big Data by @Rock_Health
Rock Report: Big Data by @Rock_HealthRock Report: Big Data by @Rock_Health
Rock Report: Big Data by @Rock_Health
 
의료의 미래, 디지털 헬스케어: 신약개발을 중심으로
의료의 미래, 디지털 헬스케어: 신약개발을 중심으로의료의 미래, 디지털 헬스케어: 신약개발을 중심으로
의료의 미래, 디지털 헬스케어: 신약개발을 중심으로
 
From personal health data to a personalized advice
From personal health data to a personalized adviceFrom personal health data to a personalized advice
From personal health data to a personalized advice
 
Improving health care outcomes with responsible data science
Improving health care outcomes with responsible data scienceImproving health care outcomes with responsible data science
Improving health care outcomes with responsible data science
 
Big Data and its Impact on Industry (Example of the Pharmaceutical Industry)
Big Data and its Impact on Industry (Example of the Pharmaceutical Industry)Big Data and its Impact on Industry (Example of the Pharmaceutical Industry)
Big Data and its Impact on Industry (Example of the Pharmaceutical Industry)
 
2016 Healthcare Trends
2016 Healthcare Trends2016 Healthcare Trends
2016 Healthcare Trends
 
Big Data in Medicine
Big Data in MedicineBig Data in Medicine
Big Data in Medicine
 
Risk & Opportunities: Healthcare Information
Risk & Opportunities: Healthcare InformationRisk & Opportunities: Healthcare Information
Risk & Opportunities: Healthcare Information
 
AMDIS CHIME Fall Symposium
AMDIS CHIME Fall SymposiumAMDIS CHIME Fall Symposium
AMDIS CHIME Fall Symposium
 

Big data chicago v2 5 14 14

  • 1. Big Data in Health Care What Marketers Need to Know Tim Gilchrist, May 2014 @timgilchrist
  • 2. Session Goals • Cover – Big Data – Artificial Intelligence / Machine Learning – How to be an Informed Consumer – Applications for Marketers – Questions 2
  • 3. Big Data The term for a collection of data sets so large and complex that they become difficult to process • Many data sources with different formats • Data with missing values • Text / Social Media • Things that don’t fit in Excel 3
  • 4. Artificial Intelligence 4 “Pay no attention to the man behind the curtain”
  • 5. Machine Learning The construction and study of systems that can learn from data 5
  • 6. Bayes Thomas Bayes (1701 – 7 April 1761) was an English mathematician and Presbyterian minister, known for formulating the theorem that bears his name: Bayes' theorem 6 Bayes theorem uses prior probabilities, combined with new observations to calculate the probability of a hypothesis being true or false Bayes is a natural fit to health care due to the presence of hypothesis (diagnosis) and events (tests / observations)
  • 7. Example You are all doctors who have administered a critical test to 50 patients You know the test is: – 75% Accurate – 10% False Positives 7 – 10 of your patients tested positive – How many are actually sick?
  • 8. Bayes’ Theorem – Mammogram Example We can present the data as a decision tree representing the probabilities confronting doctors and patients 8 If we were to take population-level down to the individual level, much more accurate probabilities would be possible #1 Mammogram #2Biopsy#3Time Observation Probability
  • 9. What Does this Mean To Marketers? • Big data is about discovering relationships 9 “Can’t beat a man with some insurance. I need that health plan baaaby!” • Then using data-driven insights to inform strategy
  • 10. Big Data Health Landscape HIE Member PCP Specialist Sees PCP Gets X-Ray Sees Specialist Ambulatory Outpatient Analysis / Transformation EMR Admission Discharge PrescriptionClaims Plan Data Portal CarePathOutputsDataTypes Direct ConnectionWearable Telemetry What is Happening What Will Happen Social PurchasedLocation Cell
  • 13. Training / Processing 13 Tweets extracted from the Twitter Fire Hose with key words “Health” and “Plan” 1MM per day 1 @CapoeiraBatuque “What's the best plan thru affordable health care. Blue cross? Blue shield? Health net? #confused #healthcare” @bluecalgal Obama says don't listen to Fox, why? Obama lied about keep your dr, health plan, cheaper than cell phone (bs) keep your dr Create Training databases for any classification desired. + and – outcomes used here 2 Weka turns Tweets into numerical code that can be analyzed by computer “String to Word Vector”. uses Naïve Bayes classifier 3 “What's the best plan thru affordable health care. Blue cross? Blue shield? Health net? #confused #healthcare” .225.357 .155 .999 ACA AccessHealth HealthPlanNow Pressure Confused Dumb @ Handle # Followers # Tweets Profile Retweets Location Date/Time Text Analysis Stemming/Tokenization Demographic Analysis Once classifications are established, rules can be applied to new Tweets with high accuracy ~90+% 4
  • 15. Other Uses 15 • Model who will be your most valuable: • Customer • Facebook follower • (Really) Determining Sentiment • Marketing Mix Simulations • Consumer Facing Predictive Technology • Prod development (HIX)
  • 18. Discovering Relationships Between Data We can use machine learning to form relationships between sets of data that are seemingly unrelated (Causal Relationships): • Making your bed in the morning and job satisfaction • Artificial Christmas trees and family “brag letters” • What you buy and why 18
  • 19. A Causal Diagram Based on Established Relationships for Estimating the Incidence of Coronary Heart Disease (CHD). (Source: Comparative quantification of health risks: Conceptual framework and methodological issues) 19© Tim Gilchrist 2013
  • 20. Bayes’ Theorem – Mammogram Example Problem: Estimates of breast cancer over diagnosis range from 25%– 52%. Physicians often misinterpret their own lab results 20 Prior Probability Chance that a woman will develop breast cancer in her 40s X 1.4% New Event: Mammography Ability of mammogram to detect cancer when present Y 75% False positives Z 10% Posterior Probability Revised Probability, given new event (positive mammogram) xy+z(1-x) 9.6% A positive mammogram still leaves a 90.4% chance that the test showed something other than cancer. When biopsies are performed on this age group, 75% are negative. Physicians rarely consider the other 90.4%
  • 21. What Does this Mean To Marketers? 21
  • 22. Other Reading Why Most Marketers Will Fail In The Era Of Big Data 8 Marketers Doing Big Data Right The big-data revolution in US health care: Accelerating value and innovation HITLAB Speaks with Tim Gilchrist, Director of eBusiness Strategy for WellPoint 22
  • 23. The State of Big Data 23

Notas del editor

  1. Example: Hand out 50 cards 10 cards will say positive 40 will say negative Ask the group What percentage of positive group are actually sick (%9.6) Real Doctors claim 70% Actual chances you have cancer after a positive mammogram are 9.6% This is so important, but very few doctors take advantage of these facts Prior probability was only 1%
  2. Imagine the above decision tree with many many branches including patient demographics, family history, radiographic study info, environmental info, etc. The accuracy would climb tremendously It’s all about understanding the data and using it to make informed decisions. Millions of women endure unnecessary biopsies because physicians do not understand the data %80 of biopsies are negative / unnecessary Bayes at the bedside But what about marketing?
  3. Just as there are relationships between a positive mammogram and actually being sick, there are relationships between consumer habits and purchasing patterns What’s the relationship here?
  4. Here is a generalized look at our Big Data landscape If we could understand the relationships between all these data what would be possible: When people need to see a doctor before the hospital – way before the claim When people switch jobs and need new insurance / HIX When they need OTC meds etc. There is no other category that how the types and abundance of leading indicators as health! Most people do not know they ill buy a car 24hours before they do
  5. Over 1MM Tweets per day regarding health. Most of them garbage Completely unstructured, no labels, etc Surprisingly, there are many looking for health plans? Separating the signal from the noise is what machine learning can be very effective at Socially viable companies can take advantage of the signal and close sales / No voice in social, don’t bother
  6. If we take 50k tweets on health care remove 70 of those that are sales prospects and 300 that are not prospects We have a training set we can use to teach and machine learning algorithm that is 94% accurate 50k becomes a few hundred qualified leads This is a decision tree that represents a fraction of the math in this problem
  7. These are well-known in consumer marketing but not so much in health care Public health researchers have made may discoveries in this area (see appendix) Again, commercial applications lag in machine learning
  8. These results are at the population level. Given