Download Kharrazi slides

DHSI/CPHIT Grand Round
Oct.19.2012
Automating the Contextualization
of Population-Based CDSS in
Tethered EHR/PHRs
Hadi Kharrazi, MHI, MD, PhD
Johns Hopkins University
[email protected]
Automating Population-Based CDSS …
Disclosures



Hadi Kharrazi reports having a financial relationship with Indiana
University from Aug 2008 to Aug 2012.
Hadi Kharrazi reports having a financial relationship with Johns Hopkins
University since Aug 2012.
Hadi Kharrazi reports that he has no financial relationship with any
commercial interest.
Speaker: Hadi Kharrazi, MHI, MD, PhD
Health Sciences Informatics Grand Rounds
© Hadi Kharrazi @ JHSPH-HPM
Oct 19.2012
2
Automating Population-Based CDSS …
Title: Automating the Contextualization of
Population-Based CDSS in Tethered EHR/PHRs
Lecture Objectives





Describe the types of HIT solutions (e.g., EHR, PHR) and how they
complement Population Health IT (PopHI)
Explain the relationship among CPGs, CDS languages, and CDS systems
Demonstrate how CDSS can be automated and applied in the context of
PopHI
Explain how population-based CDSS can bridge the gap between clinical
informatics and PubHI
Discuss the opportunities and challenges of population-based CDSS
Speaker: Hadi Kharrazi, MHI, MD, PhD
Health Sciences Informatics Grand Rounds
© Hadi Kharrazi @ JHSPH-HPM
Oct 19.2012
3
Automating Population-Based CDSS …
The presentation in a nutshell

Introduction

Background
o
Clinical Informatics (CI)
o
Public Health Informatics (PubHI)
o
Health Information Exchange (HIE) role in PubHI
o
Population Health Informatics (PopHI) - bridging the gap between CI and PubHI
o
Clinical Decision Support (CDS) continuum and PopHI

Regenstrief Institute / INPC / Indiana HIE

Previous Work
o
Developing renewable CDS systems
o
Application and results of CDS automation in INPC RMRS
o
Contextualizing CDSS in CHI Platforms (PHR)
o
Population-based CDSS

Center for Population Health IT (CPHIT) / Future Work

Q&A
© Hadi Kharrazi @ JHSPH-HPM
4
Automating Population-Based CDSS …
Introduction
© Hadi Kharrazi @ JHSPH-HPM
5
Automating Population-Based CDSS …
Introduction
•
Hadi Kharrazi, MHI, MD, PhD
“You are a doctor who is lost in cyberspace!”
Current affiliation:
(my wife)
JHSPH-HPM (Aug 2012)
Associate Director CPHIT
Director: Dr. J. Weiner
Previous affiliation:
Indiana University
Dalhousie University
Research interests:
Population Health IT


© Hadi Kharrazi @ JHSPH-HPM
CDSS/QM in HIE environments
Personal Health Records (PHR)
6
Automating Population-Based CDSS …
Introduction
Basic Research
Biomedical informatics methods,
techniques, and theories
PopHI
Bioinformatics
Applied Research
Imaging
Informatics
Public Health
Informatics
Clinical
Informatics
Consumer Health
Informatics
Molecular Research
Health Research
Biomedical informatics as a basic science
© Hadi Kharrazi @ JHSPH-HPM
7
Automating Population-Based CDSS …
Background
© Hadi Kharrazi @ JHSPH-HPM
8
Automating Population-Based CDSS …
Background > Clinical Informatics (CI)
Clinical Information Systems
If patient has x, y then do z
CDSS
Administrative
Accounting
Clinical Inpatient
EHR
EMR
Clinical Informatics
Clinical Outpatient
JHH EPR
Laboratory
Imaging
QM,
KD…
Non-clinical
© Hadi Kharrazi @ JHSPH-HPM
9
Automating Population-Based CDSS …
Background > Public Health Informatics (PubHI)
Registries
Commun.
Diseases
Reports
EHR
Claims
1
EHR
Public Health
Informatics
Public Health
Nonclinical
?
2
Syndromic
Surveillance
?
Quality
Measures
Periodic public reports and alerts
© Hadi Kharrazi @ JHSPH-HPM
10
Automating Population-Based CDSS …
Background > Health Information Exchange (HIE)
Other
sources
Registries
Commun.
Diseases
EHR
n
EHR
Public Health
Informatics
…
EHR
Public
Health
HIE
1
Syndromic
Surveillance
Quality
Measures
© Hadi Kharrazi @ JHSPH-HPM
11
Automating Population-Based CDSS …
Background > Population Health Informatics (PopHI)
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
12
Automating Population-Based CDSS …
Background > Clinical Decision Support (CDS) Continuum
If patient has x, y, has gene w, lives in Baltimore, is high risk, is Hispanic, has insurance q, and more
population health rules…, and you don’t have an MRI then do z”
© Hadi Kharrazi @ JHSPH-HPM
If patient has x, y and
you don’t have an MRI
then do z’
Organizational Info.
Clinical Informatics
Bioinformatics
If patient has x,
y, and gene w
then do z’
If patient has x, y with
insurance q then do z’
If patient has x, y and
is in high risk
population then do z’
If patient has x, y and
is Hispanic then do z’
Population Health Informatics
If patient has x, y and
lives in Baltimore
then do z’
Public Health Info
If patient has x, y
then do z
13
Automating Population-Based CDSS …
Regenstrief Institute / INPC / Indiana HIE
© Hadi Kharrazi @ JHSPH-HPM
14
Automating Population-Based CDSS …
Regenstrief Institute > INPC
•
Established in 1969 by philanthropist Sam Regenstrief
•
Regenstrief receives $3 million per year in core support from the
Regenstrief Foundation
•
Annual operating budget of approximately $23 million derived from grants
and contracts
•
Logical Observation Identifiers Names and Codes (LOINC) system, a
standard nomenclature that enables the electronic transmission of clinical
data from laboratories
•
Regenstrief Medical Records System (RMRS) was developed 35 yrs ago.
RMRS has a database of 6 million patients, with 900 million on-line coded
results, 20 million full reports including diagnostic studies, procedure
results, operative notes and discharge summaries, and 65 million radiology
images.
•
Indiana Network for Patient Care (INPC) was created in 1996. It is a citywide clinical informatics network of 11 different hospital facilities and more
than 100 geographically distributed clinics and day surgery facilities
Copyright Regenstrief Institute
© Hadi Kharrazi @ JHSPH-HPM
15
Automating Population-Based CDSS …
Regenstrief Institute > Indiana HIE (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
16
Automating Population-Based CDSS …
Regenstrief Institute > PubHI > ELR Data Completeness
Overhage JM, Grannis S, McDonald CJ. Am J Public Health. 2008 Feb
© Hadi Kharrazi @ JHSPH-HPM
Copyright Regenstrief Institute
17
Automating Population-Based CDSS …
Regenstrief Institute > PubHI > Timeliness of Data (CC vs ICD9)
Copyright Regenstrief Institute
© Hadi Kharrazi @ JHSPH-HPM
18
Automating Population-Based CDSS …
Regenstrief Institute > PubHI > Notifiable Condition Detector (NCD)
Copyright Regenstrief Institute
© Hadi Kharrazi @ JHSPH-HPM
19
Automating Population-Based CDSS …
Regenstrief Institute > PubHI > Syndromic Surveillance Service (SSS)
Copyright Regenstrief Institute
© Hadi Kharrazi @ JHSPH-HPM
20
Automating Population-Based CDSS …
Previous Work
© Hadi Kharrazi @ JHSPH-HPM
21
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS
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
}
22
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS

Issues with current CDSS

Often successful CDS RCTs are not replicable (renewable) in
other clinical settings.

CDS knowledge interchange standards have existed for as long
as clinical data interchange standards (about 2 decades), yet
sharing of computable, actionable knowledge in practice is
extremely limited

Current computerized knowledge languages representing CPGs
are not standardized: Arden syntax, CARE, GELLO, GLIF

The majority of the current CDSS systems are hard coded, as
experienced in a simpler format in CHI studies, which make
them very limited due to maintenance issues.
© Hadi Kharrazi @ JHSPH-HPM
23
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS
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
24
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS

Phases to Develop Renewable CDSS
1.
Lexical Integration (mapping of idiosyncratic codes to
standard terms)
2.
Syntactic Mapping of Different Concrete Representation
Languages

Parse original content as XML (e.g., Chaperon)

XSLT Transformation (e.g., SAXON)
3.
Semantic Integration of Procedural Knowledge into
Declarative Rules (SQL commands)
4.
Pragmatic Integration (query enhancements, implementation
and evaluation)
© Hadi Kharrazi @ JHSPH-HPM
25
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS
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
26
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS
Language Specific
© Hadi Kharrazi @ JHSPH-HPM
Language Free
27
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS
BEGIN BLOCK CHOLESTEROL
DEFINE "MALE_PT" {VALUE} AS "SEX" IS = 'M'
DEFINE "CHOLESTEROL LAST" AS LAST "CHOL TESTS" [BEFORE "DATE"] EXIST
IF "CHOLESTEROL LAST" WAS AFTER 5 YEARS AGO
THEN EXIT CHOLESTEROL
ELSE CONTINUE
IF "MALE_PT" EXISTS AND "AGE" WAS >= 35
THEN REMIND Male patients over the age of 35 should have their total
cholesterol measured every 5 years.
AND ORDER CHOLESTEROL (SCREEN)
AND SAVE "UNO" AS "CHOL_SCN" {INTEGER}
AND EXIT CHOLESTEROL
END BLOCK CHOLESTEROL
Sample CARE Rule Block
© Hadi Kharrazi @ JHSPH-HPM
28
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS
<?xml version="1.0" encoding="UTF-8"?>
<rule xmlns="http://regenstrief.org/xcare">
<let name="AGE">
<and>
<filter>
<reference name="BIRTH"/>
<and>
<not-equal op="NE" component="VALUE">
<literal type="integer" value="0"/>
</not-equal>
<less-than op="LT" component="VALUE">
<literal type="time" value="now"/>
</less-than>
</and>
</filter>
<division>
<subtraction>
<literal type="time" value="now"/>
<reference name="BIRTH"/>
</subtraction>
<literal type="real" value="365.25"/>….
Intermediate XML
© Hadi Kharrazi @ JHSPH-HPM
29
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS
SELECT id,sys_id_value,numeric_value,time_value,text_value,boolean_value,
code_value,data_type,institution_id,patient_id,service_code,primary_time
FROM(
select t_from.* from (SELECT
id,sys_id_value,numeric_value,time_value,text_value,boolean_value,code_value,data_type,institution_
id,patient_id,service_code,primary_time ,
myrank
FROM(
SELECT * FROM
(SELECT
id,sys_id_value,numeric_value,time_value,text_value,boolean_value,code_value,data_type,institution_
id,patient_id,service_code,primary_time, rank() OVER
(PARTITION BY INSTITUTION_ID,PATIENT_ID ORDER BY primary_time DESC)
myrank FROM (select TO_NUMBER(NULL) as id,
To_NUMBER(NULL) as sys_id_value,
1 as numeric_value,
DATE_OF_BIRTH as time_value,
TO_CHAR(NULL) as text_value,
To_CHAR(null) as boolean_value,
TO_CHAR(NULL) as code_value,
'T' as data_type, …
SQL Statement (adaptable for new schemas)
© Hadi Kharrazi @ JHSPH-HPM
30
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS

Limitations

The process does not start directly with free text CPG. Knowledge
Representation languages such as CARE or Arden are required.

If originating CPG does not conform to a standard terminology,
lexical integration is required again.

Fuzzy or probabilistic representational languages are currently
not incorporated.

Mismatched data granularity or ignored data sets will create
database integration complicated.

Each CDSS SQL is mutually exclusive of other CDSS thus limiting
the merger of SQLs that can result in a combined CDSS.
© Hadi Kharrazi @ JHSPH-HPM
31
Automating Population-Based CDSS …
Previous Work > Developing Renewable CDSS

Application Areas

Clinical: CDSS reminder rules (e.g., CPOE integration)  popup
CDSS alerts for providers

Population-based:
o
Healthcare quality measures (e.g., NQF, ACO)  Quality
monitoring dashboard
o
Consumer Health Informatics (e.g., tethered PHRs)
o
Public Health Informatics (PHI) (e.g., notifiable complex
conditions)
o
Transition of care (e.g., CCD-based CDSS for inpatient to
outpatient)
© Hadi Kharrazi @ JHSPH-HPM
32
Automating Population-Based CDSS …
Previous Work > CDSS Application in EHR
CLINICAL_VARIABLE
PATIENT
INSTITUTION
CARE KRL
42,666,350 records
453,518 (unique)
118 providers
74 Rules (SQL)
RMRS sample database was used in this research
a subset of the INPC database
© Hadi Kharrazi @ JHSPH-HPM
33
Automating Population-Based CDSS …
Previous Work > CDSS Application in EHR
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
34
Automating Population-Based CDSS …
Previous Work > CDSS Application in EHR
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
35
© 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)
55
44
41
56
43
40
49
29
17
32
30
5
15
48
114
46
23
119
107
21
16
14
18
125
13
22
123
102
110
109
31
116
9
7
103
2
47
20
120
124
118
1
25
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
CHOLESTEROL_1.sql
CHOLESTEROL_2.sql
PREVENT_STOOL_2.sql
PREVENT_PAP_1.sql
MEDICINE_2.sql
OBFORM_5.sql
OBFORM_2.sql
Academic Hospitals
800,000
36
© Hadi Kharrazi @ JHSPH-HPM
OBFORM_3.sql
OBFORM_1.sql
0
11
Automating Population-Based CDSS …
Previous Work > CDSS Application in EHR
70000
60000
50000
Institute 1: All CDS Hits (regardless of missed proportion)
40000
30000
20000
10000
0
1,000,000
Total Hits
900,000
700,000
600,000
500,000
400,000
300,000
200,000
100,000
institutes
Number of CARE rule hits against the RMRS sample database based on each
healthcare provider / institute (zero-hit institutes are removed)
Academic/Primary hospitals have the highest rates of hits
This peak might be simply because of their higher population of patients (needs population adjustment)
Automating Population-Based CDSS …
Total Hits
Previous Work > CDSS Application in EHR
1,000,000
900,000
800,000
700,000
600,000
500,000
400,000
300,000
200,000
55
44
41
56
43
40
49
29
17
5
32
30
15
48
46
114
23
21
16
119
107
14
18
13
22
125
123
102
110
109
9
31
116
7
2
103
47
20
120
124
1
25
118
0
11
100,000
Adj Population
100.00%
Other Hospitals
90.00%
80.00%
70.00%
60.00%
50.00%
40.00%
30.00%
20.00%
55
44
41
56
43
40
49
29
17
32
30
5
15
114
48
46
23
119
107
21
16
14
18
125
13
22
123
102
110
31
109
116
9
7
103
2
47
20
120
124
118
1
25
0.00%
11
10.00%
institutes
Number of CARE rule hits against the RMRS sample database based on
ever healthcare provider institute (population adjusted percentage)
Percentage of rule hits is suddenly smaller in the left side of the diagram (i.e., primary hospitals)
Percentage of rule hits is larger in the right side of the diagram (i.e., other hospitals)
Percentage > creates higher patient risk at a certain institute
Raw number > forms internal policies / economical road maps
© Hadi Kharrazi @ JHSPH-HPM
37
Automating Population-Based CDSS …
Previous Work > CDSS Application in EHR
rules
institutes
Institute versus Rules (Raw Numbers)
(zero rules / zero institutes are removed)
© Hadi Kharrazi @ JHSPH-HPM
38
Automating Population-Based CDSS …
Previous Work > CDSS Application in EHR
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
39
© 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)
Automating Population-Based CDSS …
Previous Work > CDSS Application in EHR
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%
40
© 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)
Automating Population-Based CDSS …
Previous Work > CDSS Application in EHR
Rules
Missed %
institutes
rules
Missed rate by institute
The effect size of the missed rules depends on the population size.
© Hadi Kharrazi @ JHSPH-HPM
41
Automating Population-Based CDSS …
Previous Work > CDSS Application in EHR
Web Interface
(http://aurora.regenstrief.org:8080/ClinicalKnowledgeHub/)
© Hadi Kharrazi @ JHSPH-HPM
42
Automating Population-Based CDSS …
Previous Work > CDSS Application in EHR
Applying all CARE rules against the RMRS sample database
Preventive Care Quality Improvement  Reminder Forms
© Hadi Kharrazi @ JHSPH-HPM
Copyright Regenstrief Institute
43
Automating Population-Based CDSS …
Previous Work > CDSS Application in PHR
© Hadi Kharrazi @ JHSPH-HPM
44
Automating Population-Based CDSS …
Previous Work > CDSS Application in PHR
Dynamically
generated
CDSS SQL
commands
CHI Framework (Tx Group)
© Hadi Kharrazi @ JHSPH-HPM
45
Automating Population-Based CDSS …
Previous Work > Population-based CDSS
Item
Traditional
CDSS-based
Focus
n-Patient-at-a-time
1-Patient-at-a-time
Model
Statistical (probabilistic)
CPG (rule-based)
Updatability
Delayed (Hard coded)
Immediate (Dynamic)
HIE Integration
Possible
Already Exists
Portability
Major Change Req.
Minimal Change Req.
Data Source
EHR Subsets (limited)
Complete EHR, PHR, …
Contextualization
Complex
SQL Driven
Traditional PubHI versus CDSS-based PopHI
© Hadi Kharrazi @ JHSPH-HPM
46
Automating Population-Based CDSS …
Previous Work > Population-based CDSS
PopHI will
reveal the
causes, and…
Simple statistics based on numerical statistics
Consider the potentials with Population-based CDSS integration
© Hadi Kharrazi @ JHSPH-HPM
Copyright Regenstrief Institute
47
Automating Population-Based CDSS …
CPHIT / Future Work
© Hadi Kharrazi @ JHSPH-HPM
48
Automating Population-Based CDSS …
CPHIT / Future Work
The Johns Hopkins
Center for Population Health Information Technology
(CPHIT, or “see-fit”)

The mission of CPHIT is to improve the health and wellbeing of
populations by advancing the state-of-the-art of Health
Information Technology and e-health tools used by public health
agencies and private health care organizations.

CPHIT’s focus will be on the application of EHRs, PHRs and other
digitally-supported health improvement interventions targeted at
communities, special need populations and groups of consumers
cared for by integrated delivery systems.

Director: Dr. J. Weiner

www.jhsph.edu/cphit
Copyright Dr. J. Weiner
© Hadi Kharrazi @ JHSPH-HPM
49
Automating Population-Based CDSS …
CPHIT / Future Work
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
50
Automating Population-Based CDSS …
CPHIT / Future Work

Meaningful Use
Stage 3
Stage 2
2009 HITECH
Policies
Stage 1
HIT-Enabled Health Reform
2011 Criteria
(Capture & Share)
2013 Criteria
(CDSS)
Population
Health
Informatics
© Hadi Kharrazi @ JHSPH-HPM
2015 Criteria
(Improved
Outcome)
51
Automating Population-Based CDSS …
CPHIT / Future Work
Population-based HIE
PHRs
EHRs
Claims
MPI
Registries
LABs
PACS
Population-based CDSS across databases
Federated Consistent Databases
(1) Data gathered centrally in separate physical files, “mirrors” of remote sites;
(2) Standardized at the time it comes in.
© Hadi Kharrazi @ JHSPH-HPM
52
Automating Population-Based CDSS …
CPHIT / Future Work
Provider
Rx
Labs
Insurance
MU3
HIE
MU1
EHR
EHR
EHR
ACO
EHR
Care
Manage.
MU2
CDSS
Risk
Assess.
Imaging
CPOE
Tethered
CPG
PHR
PDSS
Patient
Sensors…
© Hadi Kharrazi @ JHSPH-HPM
Predictive
Modeling
Quality
Measure
Pay for
Perform.
Population Health Informatics 2.0
Outpatient
Public Health
53
Automating Population-Based CDSS …
CPHIT / Future Work
 CPHIT’s initial R&D Priorities:

Natural language processing (NLP) and pattern recognition that improve
population based interventions.

Effective approaches for integration of consumer-focused m-health/PHRs/
e-health with provider-focused EHRs.

Development of “e-measures” of population health status, risk,
safety/quality using EHR/HIT systems.

EHR based tools to identify high risk populations for preventive and/or
chronic care outreach.

Approaches for integrating current functions of public health agencies into
interoperable EHR networks.

Approaches for applying HIT to population-based evidence-generation /
research.
Copyright Dr. J. Weiner
© Hadi Kharrazi @ JHSPH-HPM
54
Automating Population-Based CDSS …
Thank you
Q&A
© Hadi Kharrazi @ JHSPH-HPM
55