Big Data and Analytics in Healthcare
Philadelphia 2013
Health Information Exchanges and Big Data:
Challenges and Opportunities
Hadi Kharrazi, MHI, MD, PhD
Johns Hopkins University
[email protected]
Big Data and HIEs
The presentation in a nutshell

Introduction

Big Data and Healthcare

HIE History, Architecture and Services

HIE Examples (Indiana HIE, CRISP)

HIE Big Data Exploration (translating CPGs into population health data)

HIE and Population Health IT Framework

JHU Center for Population Health IT (CPHIT)
© Hadi Kharrazi @ JHSPH-HPM
2
Big Data and HIEs
Introduction
•
o
o
o
Hadi Kharrazi, MHI, MD, PhD
Assistant Professor, Johns Hopkins School of Public Health
Department of Health Policy and Management
Joint Appointment, Johns Hopkins School of Medicine
Division of Health Sciences Informatics
Assistant Director
Johns Hopkins Center for Population Health IT (CPHIT)
o
E: [email protected]
L: http://www.linkedin.com/in/kharrazi/
W: http://jhsph.edu/cphit
o
Research interests: Population DSS, HIE & ACO Analytics
o
o
© Hadi Kharrazi @ JHSPH-HPM
3
Big Data and HIEs
Big Data and Healthcare

Statistics:
o Data size: 2012 = 500 petabytes  2020 = 25,000 petabytes
o Cost: 2009 = 17.6% GDP ($2.9t)  2025 = 25.0% GDP
o Big Data Saving: ~$300b/yr

Definition:
o Big Data is a collection of data sets so large and complex that it becomes
difficult to process using on-hand database management tools or traditional
data processing applications.

History in Healthcare
o Last time: We pushed the data boundary in the Human Genome Project (1990)
o This time: Massive roll out of EHRs (Meaningful-Use) and Integration of
various source of data (genomic, mobile, patient-generated, exchanges)
© Hadi Kharrazi @ JHSPH-HPM
4
Big Data and HIEs
Big Data and Healthcare (cont.)

Big Data Specs – data driven 4Vs:
o Volume  quantity (size)
o Variety  type (structure, standardized, ontology)
o Veracity  quality (meaning, completeness, accuracy)
o Velocity  time (real-time, timeliness)
a
Volume
Veracity
b
c
d
e
…
Variety


Velocity
?
© Hadi Kharrazi @ JHSPH-HPM
5
Big Data and HIEs
Big Data and Healthcare (cont.)
Biomedical informatics methods,
techniques, and theories
Basic Research
Bioinformatics
Applied Research
Imaging
Informatics
Population
HIT
Public Health
Informatics
Clinical
Informatics
Molecular Research
Consumer Health
Informatics
Health Research
Biomedical informatics as a basic science
© Hadi Kharrazi @ JHSPH-HPM
6
Big Data and HIEs
Copyright H.Kharrazi
Big Data and Healthcare (cont.)
Big Data Specs – clinical driven 5Ms:
o Measure  size, type, quality (QMs, digitization, …meaningful use)
o Mapping  integration, interoperability (information exchanges, EDWs)
o Methods  analytics (exploration, visualization, predictive, …replacing RCTs)
o Meanings  knowledge (EBM, CPG, …meaningful use)
o Matching  outcomes (triple aims, ACA)
Consumer
Public
HIEs
Expl/Visu
NwHIN
?
© Hadi Kharrazi @ JHSPH-HPM
Statistical
Data Mining
CPG
Machine
Learning
??
Application
Knowledge
CDSS
EBM
ACA
Theories
???
Outcome
HITECH
MU1,2,3
Matching
Bio Signal
ACOs
Analytics
Modeling
Simulation
Meaning
Genomics
EDWs
Measure
EHRs
Mandates
Business
Standards
Methods
Mandates
CQMs
Devices
Mapping

Triple Aim
Other
policies
?????
??
7
Big Data and HIEs
Copyright H.Kharrazi
Big Data and Healthcare (cont.)

Johns Hopkins
JH Community
Physicians
JH Hospitals
JH Health Care
JH ACO
Epic
JCHIP / JCASE (CMMI)
© Hadi Kharrazi @ JHSPH-HPM
Measure
-
Mapping
-
Methods
??
Meaning
???
Matching
??
8
Big Data and HIEs
HIE History
Copyright HiMSS
© Hadi Kharrazi @ JHSPH-HPM
9
Big Data and HIEs
HIE History (cont.) > ONC
ONC relationship with DHHS
© Hadi Kharrazi @ JHSPH-HPM
Copyright HiMSS
10
Big Data and HIEs
HIE History (cont.) > Health Information Organization (HIO)

HIE (verb): The electronic movement of health-related information among disparate
organizations according to nationally recognized standards in an authorized and
secure manner.

HIO (noun): An organization that oversees and governs the exchange activities of
health-related information among independent stakeholders and disparate
organizations according to nationally recognized standards in an authorized and
secure manner.

An HIO can be described by many acronyms, including:
o
State Level Health Information Exchange (SLHIE)
o
Regional Health Information Exchange (RHIO)
o
Regional Health Information Network (RHIN)
o
Health Information Exchange Networks (HIE[N])
o
Others: Integrated Delivery Systems (IDNs); Physician practices HIEs; Payerled HIEs; and, Disease-specific HIEs.
Copyright HiMSS
© Hadi Kharrazi @ JHSPH-HPM
11
Big Data and HIEs
HIE Architecture > Centralized
PHRs
EHRs
Claims
Registries
Data Repository
LABs
PACS
Centralized HIE
© Hadi Kharrazi @ JHSPH-HPM
Copyright Regenstrief Institute
12
Big Data and HIEs
HIE Architecture (cont.) > Federated
PHRs
EHRs
Claims
Registries
LABs
PACS
Federated Inconsistent HIE
© Hadi Kharrazi @ JHSPH-HPM
Copyright Regenstrief Institute
13
Big Data and HIEs
HIE Architecture (cont.) > Federated
PHRs
EHRs
Claims
MPI
LABs
PACS
Federated Consistent HIE
© Hadi Kharrazi @ JHSPH-HPM
Registries
Copyright Regenstrief Institute
14
Big Data and HIEs
HIE Architecture (cont.) > Hybrid
PHRs
EHRs
Data
Repository
LABs
EHRs
Claims
PACS
MPI
Registries
LABs
Hybrid HIE
© Hadi Kharrazi @ JHSPH-HPM
Claims
Copyright Regenstrief Institute
15
Big Data and HIEs
HIE Architecture (cont.) > Switch
PHRs
EHRs
Claims
Data
Standardization
Machine
LABs
PACS
Switch HIE
© Hadi Kharrazi @ JHSPH-HPM
Registries
Copyright Regenstrief Institute
16
Big Data and HIEs
HIE Architecture (cont.) > Patient Centric
Other…
EHRs
Claims
Registries
LABs
PACS
Patient Centric HIE (e.g., PHR controlled)
© Hadi Kharrazi @ JHSPH-HPM
Copyright Regenstrief Institute
17
Big Data and HIEs
HIE Architecture (cont.)
Health Information Organizations / Regional HIOs
© Hadi Kharrazi @ JHSPH-HPM
Copyright Regenstrief Institute
18
Big Data and HIEs
HIE Architecture (cont.) > NwHIN
Connecting RHIOs / NwHIN
© Hadi Kharrazi @ JHSPH-HPM
Copyright Regenstrief Institute
19
Big Data and HIEs
HIE Architecture (cont.) > NwHIN
NHII (2004)  NHIN (2010)  NwHIN (2011)
Nationwide Health Information Network (NwHIN)
© Hadi Kharrazi @ JHSPH-HPM
Copyright HiMSS
20
Big Data and HIEs
HIE Services > Core Services

Presentation Services: login, patient look-up, request patient
records, view data

Business Application Services: e-Prescribing, EMR, lab,
radiology, eligibility checking, problem list/visit history

Data Management Services: data persistence/access,
value/code sets, key manage.

Data Storage Services: message logs, XML Schemas,
Provider/User Directory

Integration Services: message translation/transport, HL7
mapping, EMR adapter

System Management Services: system config, audit/logging,
exception handling
Measure
-
Mapping
-
Methods
?
Meaning
?
Matching
?
Copyright HiMSS
© Hadi Kharrazi @ JHSPH-HPM
21
Big Data and HIEs
HIE Services > Data Services by Constituency



Hospitals:

Public Health:
o
Clinical messaging
o
Needs assessment
o
Medication reconciliation
o
Biosurveilance
o
Shared EHR
o
Reportable conditions
o
Eligibility checking
o
Results delivery
Physicians:
o
Result reporting
o
Secure document sharing
o
Shared EHR
o
Clinical decision support
o
Eligibility checking
Laboratory:
o
Clinical messaging
o
Orders

Mapping
??
Methods
??
Meaning
?
Matching
-
Personal Health Records
Researchers:
o

?
Consumers:
o

Measure
De-identified longitudinal clinical data
Payers:
o
Quality measure
o
Claims adjustment
o
Secure document transfer
Copyright HiMSS
© Hadi Kharrazi @ JHSPH-HPM
22
Big Data and HIEs
HIE Services > Emerging Services

Next Generation Analytics
o
Measure
-
Mapping
-
Methods
???
Meaning
???
Matching
??
Data warehouse, data analytics and business intelligence
o
Quality reporting support
o
Performance management
o
Fraud and abuse identification and prevention
o
Care gap identification
o
Care and disease management
o
Public health monitoring and analysis
o
Population monitoring and predictive profiling
Copyright HiMSS
© Hadi Kharrazi @ JHSPH-HPM
23
Big Data and HIEs
HIE Examples > IHIE
The Indiana HIE (IHIE) includes (as of mid-2011):
•
Federated Consistent Databases
•
22 hospital systems  ~70 hospitals
•
5 large medical groups and clinics & 5 payors
•
Several free-standing labs and imaging centers
•
State and local public health agencies
•
10.75 million unique patients
•
20 million registration events
•
3 billion coded results
•
38 million dictated reports
•
9 million radiology reports
•
12 million drug orders
•
577,000 EKG tracings
•
120 million radiology images
Copyright Regenstrief Institute
© Hadi Kharrazi @ JHSPH-HPM
24
Big Data and HIEs
HIE Examples (cont.) > CRISP (Chesapeake Regional Info. Sys. for our Patients)

Focus Areas:
Progress Metric
Result
o Query Portal Growth
Organizations Live
o Direct Secure Messaging
Hospitals (Total 48)
48
o Encounter Notification System (ENS)
Hospital Clinical Data Feeds
(Total 143 - Lab, Radiology, Clinical Docs)
86
o Encounter Reporting System (ERS)
National Labs
2
o Health Benefits Exchange integration
Radiology Centers (Non-Hospital)
5
Identities and Queries
Master Patient Index (MPI) Identities
~4M
Opt-Outs
~1500
Queries (Past 30 Days)
~3500
Data Feeds Available
Lab Results
~16M
Radiology Reports
~5M
Clinical Documents
~2M
Copyright CRISP
© Hadi Kharrazi @ JHSPH-HPM
25
Big Data and HIEs
HIE Big Data Exploration > Translating CPGs into Population Health Data
Clinical Practice Guidelines
Computerized CPG
(CARE, Arden syntax)
Manual Programming
for each CPG
One patient at the time
Relational
DB
© Hadi Kharrazi @ JHSPH-HPM
Computerized
AHRQ’s
Computerized
NGC
Popup
CPGs
(guideline.gov)
alert
(Know.Rep.Lang.)
DSS
mysql_connect(‘localhost’,
Arden
Syntax:
‘root’, ‘pass’, ‘ehr');
Patient
has diabetes
and
has
not
had an
*Blood Pressure,
Weight/Body
Mass
Index
(BMI)
- Every
adults:
Blood
eye
exam
in visit.
two For
years.
Based
onpressure
mysql_query(‘SELECT
logic:
patient_id
FROM
patient
target
goal xxx
<130/80
mm
Hg;
BMIto
(body
mass
guideline
you
may
want
consider
WHERE
if (creatinine
patient.Fname
is not number)
= ‘peter’or
AND
(calcium is
index) <25 kg/m2.
asking
for an
consultation.
not number)
patient.Lname
or=eye
‘Johnson’’);
(phosphate is not number) then
For children: Blood pressure target goal <90th
$results
conclude
= mysql_fetch();
false;
percentile adjusted for age, height, and gender;
endif;
BMI-for-age <85th percentile.
$patient_id
product :=
= $results[‘patient_id’];
calcium * phosphate;
----------------------------------------------------Foot Exam (for adults) - Thorough visual
CARE Language:
mysql_query
(‘SELECT * FROM lab WHERE
inspection every diabetes care visit; pedal
patient_id = $patient_id’ and lab_term =
pulses, neurological exam yearly.
BEGIN BLOCK CHOLESTEROL
‘HBA1c’);
Dilated Eye Exam (by trained expert) - Type 1:
DEFINE "MALE_PT"
$results
= mysql_fetch();
{VALUE} AS "SEX" IS = 'M'
Five years post diagnosis, then every year.
DEFINE "CHOLESTEROL LAST" AS LAST "CHOL
TESTS" [BEFORE
$lab_result
= $results[‘lab_term’];
"DATE"] EXIST
Depression - Probe for emotional/physical
IF "CHOLESTEROL LAST" WAS AFTER 5 YEARS
factors linked to depression yearly; treat
AGO
IF
($lab_result > 9){
aggressively with counseling, medication,
THEN
ECHOEXIT
“Patient
CHOLESTEROL
has a high HBA1c level”;
and/or referral.
ELSE CONTINUE
}
26
Big Data and HIEs
HIE Big Data Exploration (cont.)
Clinical Practice Guidelines
NQF QM eMeasures
Computerized CPG
(CARE, Arden syntax)
Manual Programming
for each CPG
One patient at the time
Relational
DB 1
© Hadi Kharrazi @ JHSPH-HPM
CDSS
SQL n
Relational
DB n
CDSS
SQL…
Relational
DB…
CDSS
SQL 2
Relational
DB 2
Automated for all CPGs
All patient at the time
CDSS SQL 1
Population
Health
Informatics
27
Big Data and HIEs
HIE Big Data Exploration (cont.)
NQF
Lexical
Integration
CARE
Arden
XML
Syntactic
Mapping
Renewable
CPG
Consortium
CPG
Semantic
Integration
CDSS
SQL n
Relational
DB n
CDSS
SQL…
Relational
DB…
CDSS
SQL 2
Relational
DB 2
CDSS
SQL 1
Relational
DB 1
XSLT trans.
Generating Renewable CDSS
© Hadi Kharrazi @ JHSPH-HPM
Pragmatic
Integration
28
Big Data and HIEs
HIE Big Data Exploration (cont.)
Language Specific
© Hadi Kharrazi @ JHSPH-HPM
Language Free
29
Big Data and HIEs
HIE Big Data Exploration (cont.)
Rules
Population
One patient*
59
0.18
CHOLESTEROL_1.sql
180
0.11
CHOLESTEROL_2.sql
1
0.12
DIABETES_1.sql
541
0.56
DIABETES_2.sql
1561
0.47
DIABETES_3.sql
182
0.40
DIABETES_4.sql
121
0.40
DIABETES_5.sql
1620
1.51
…
…
TCL_2.sql
67
0.12
TCL_3.sql
69
0.18
TCL_4.sql
61
0.17
1
0.07
WHOLE_1.sql
127
0.16
WHOLE_2.sql
129
0.13
1249
0.73
CD4_1.sql
….
TETANUS_SHOT_1.sql
Average (sec)
Runtime per second
* each row shows average of 20 individual runs
© Hadi Kharrazi @ JHSPH-HPM
30
Big Data and HIEs
HIE Big Data Exploration (cont.)
70000
Institute 1: All CDS Hits (regardless of missed proportion)
60000
50000
40000
30000
20000
10000
0
MEDICINE_1.sql
HIV_PT_1.sql
TCL_3.sql
PREVENT_STOOL_1.sql
CD4_1.sql
WHOLE_1.sql
TCL_4.sql
IDC_REM_1.sql
OBFORM_6.sql
PREVENT_MAM_1.sql
K_1.sql
FLU_SHOT_5.sql
DIABETES_5.sql
TCL_1.sql
WHOLE_2.sql
PREVENT_STOOL_4.sql
DIABETES_1.sql
K_2.sql
PREVENT_MAM_3.sql
FLU_SHOT_1.sql
TCL_2.sql
LIPID_MANAGEMENT_1.sql
TETANUS_SHOT_1.sql
PEDS_9.sql
PEDS_8.sql
PEDS_7.sql
PEDS_1.sql
PEDS_11.sql
PEDS_10.sql
FLU_SHOT_3.sql
HEPVAX_B_1.sql
FLU_SHOT_4.sql
PNVX2_1.sql
PREVENT_MAM_2.sql
31
© Hadi Kharrazi @ JHSPH-HPM
CHOLESTEROL_1.sql
CHOLESTEROL_2.sql
PREVENT_STOOL_2.sql
PREVENT_PAP_1.sql
MEDICINE_2.sql
OBFORM_5.sql
OBFORM_2.sql
OBFORM_3.sql
OBFORM_1.sql
applicable CARE rules
Applying all applicable CARE rules for Institute-1
Some rules are never hit for institute 1 (not shown in this diagram)
Big Data and HIEs
HIE Big Data Exploration (cont.)
rules
institutes
Institute versus Rules (Raw Numbers)
(zero rules / zero institutes are removed)
© Hadi Kharrazi @ JHSPH-HPM
32
Big Data and HIEs
HIE Big Data Exploration (cont.)
Institute 1: Percentage of Hit / Miss Rules
30.00%
25.00%
20.00%
10.00%
Percent
5.00%
HEPVAX_B_1.sql
PREVENT_MAM_2.sql
PREVENT_MAM_3.sql
PEDS_7.sql
PEDS_1.sql
TCL_1.sql
CD4_1.sql
K_2.sql
OBFORM_3.sql
FLU_SHOT_1.sql
PEDS_9.sql
IDC_REM_1.sql
PEDS_11.sql
LIPID_MANAGEMENT_1.sql
PREVENT_STOOL_4.sql
WHOLE_2.sql
CHOLESTEROL_1.sql
OBFORM_2.sql
OBFORM_6.sql
PNVX2_1.sql
PEDS_10.sql
OBFORM_1.sql
DIABETES_5.sql
PREVENT_PAP_1.sql
PREVENT_STOOL_1.sql
FLU_SHOT_3.sql
CHOLESTEROL_2.sql
PREVENT_STOOL_2.sql
OBFORM_5.sql
K_1.sql
FLU_SHOT_4.sql
TCL_4.sql
MEDICINE_2.sql
TETANUS_SHOT_1.sql
TCL_2.sql
FLU_SHOT_5.sql
PEDS_8.sql
DIABETES_1.sql
WHOLE_1.sql
TCL_3.sql
33
© Hadi Kharrazi @ JHSPH-HPM
~12.3%
PREVENT_MAM_1.sql
MEDICINE_1.sql
HIV_PT_1.sql
0.00%
%13.49
Avg
15.00%
applicable CARE rules
Missed applicable CARE rules for institute-1
(CDSS-generated QM fingerprint)
Manual process of deciphering the rules (hit-means-miss OR hit-means-hit-only)
The lower the better (less missed)
Big Data and HIEs
HIE Big Data Exploration (cont.)
Institute 2: Inverse Percentage of Hit / Miss Rules
30.00%
25.00%
20.00%
15.00%
10.00%
%3.47
Avg
PREVENT_MAM_2.sql
PREVENT_STOOL_2.sql
PREVENT_STOOL_4.sql
HEPVAX_B_1.sql
LIPID_MANAGEMENT_1.sql
TCL_2.sql
OBFORM_5.sql
PNVX2_1.sql
PEDS_10.sql
OBFORM_1.sql
TCL_4.sql
WHOLE_2.sql
PREVENT_PAP_1.sql
FLU_SHOT_4.sql
PEDS_1.sql
PEDS_8.sql
OBFORM_2.sql
TETANUS_SHOT_1.sql
PEDS_11.sql
OBFORM_3.sql
MEDICINE_2.sql
CHOLESTEROL_1.sql
CHOLESTEROL_2.sql
FLU_SHOT_3.sql
K_2.sql
K_1.sql
PREVENT_MAM_3.sql
~5.5%
34
© Hadi Kharrazi @ JHSPH-HPM
PEDS_9.sql
PEDS_7.sql
OBFORM_6.sql
IDC_REM_1.sql
HIV_PT_1.sql
FLU_SHOT_1.sql
CD4_1.sql
0.00%
Percent
5.00%
applicable CARE rules
Missed applicable CARE rules for institute-2
(CDSS-generated QM fingerprint)
Manual process of deciphering the rules (hit-means-miss OR hit-means-hit-only)
The lower the better (less missed)
Big Data and HIEs
HIE Big Data Exploration (cont.)
Rules
Missed %
institutes
rules
Missed rate by institute
The effect size of the missed rules depends on the population size.
© Hadi Kharrazi @ JHSPH-HPM
35
Big Data and HIEs
HIE and Population HIT > Population Health Informatics
Population Health Info.
Population Health Analytics
© Hadi Kharrazi @ JHSPH-HPM
Rx Data
Admin data
repositories
EHRs
1…n
HIE
Public Health
Informatics
B
Public Health
A
PubHI to CI
Insurance
Data
CI to PubHI
Personal
Health
Records
Clinical Informatics
CI to PubHI
PubHI to CI
Population Health Data-warehouse
C
36
Big Data and HIEs
HIE and Population HIT > CDS Continuum (cont.)
Population-based HIE
PHRs
EHRs
Claims
MPI
Registries
LABs
PACS
Population-based CDSS across databases
© Hadi Kharrazi @ JHSPH-HPM
37
Big Data and HIEs
CPHIT
The Johns Hopkins
Center for Population Health Information Technology
(CPHIT, or “see-fit”)

The mission of this innovative, multi-disciplinary R&D center is to improve
the health and well-being of populations by advancing the state-of-the-art
of Health IT across public and private health organizations and systems.

CPHIT focuses on the application of electronic health records (EHRs),
mobile health (m-health) and other e-health and HIT tools targeted at
communities and populations.

Director: Dr. J. Weiner

www.jhsph.edu/cphit
Copyright Dr. J. Weiner
© Hadi Kharrazi @ JHSPH-HPM
38
Big Data and HIEs
CPHIT (cont.)
External
PH/IDS Orgs.
JH Healthcare
Solutions,
LLC
JH Health
System
CPHIT
JHU Academic
Departments
and other R&D
centers
Business
Partners
Industry
Foundations
Government
CPHIT Organizational Linkages
Copyright Dr. J. Weiner
© Hadi Kharrazi @ JHSPH-HPM
39
Big Data and HIEs
CPHIT (cont.)

CPHIT’s R&D Priorities

Development and testing of measures created from EHRs and other HIT systems to
quantify quality, outcome and need within target populations and communities.

Use and advancement of computing methodologies – including natural language
processing (NLP) and pattern recognition tools – to improve the application of nonstructured EHR data for population-based interventions.

Initiation of effective approaches for linking provider-centric EHR systems with
consumer-centric internet and mobile-based e-health applications.

Development of EHR-based tools and decision support applications to identify and
help manage high risk populations for preventive and/or chronic care and to
coordinate care within IDS/ACO.

Strategic approaches for creating an interoperable community of EHR networks and
integrating them with the current functions of public health agencies (e.g.,
surveillance and vital records).

Creation of legal/ethical and policy frameworks to support the development of
secondary use of EHRs for public health programs and research.
Copyright Dr. J. Weiner
© Hadi Kharrazi @ JHSPH-HPM
40
Big Data and HIEs
CPHIT (cont.) > Research Pipeline

Developing and Testing "e-ACGs": Using EHR / HIT Data to Enhance
Johns Hopkins ACGs' Ability to Measure Risk and Health Status

Real-time, Cross-Provider High Readmission Risk Detection and
Notification System

Evaluate the economic impact of the Shared Patient Virtual Health Record
(SPVHR) in Inpatient Settings

Prescription Drug Monitoring Program (PDMP) Evaluation

Developing and evaluating an Inter-Provider Health Screening and Quality
Measure System (IPHSQMS)

Identifying high risk pregnancies by linking claims and clinical data using
natural language processing tools
www.jhsph.edu/cphit
© Hadi Kharrazi @ JHSPH-HPM
41
Big Data and HIEs
Thank you
Q&A
www.jhsph.edu/cphit
© Hadi Kharrazi @ JHSPH-HPM
42