Más contenido relacionado La actualidad más candente (18) Similar a iHT2 Health IT Summit in Phoenix 2013 – Terhilda Garrido, VP, HIT Tranformation & Analytics, Kaiser Permanente Case Study "Big Data, Little Data: Value and Transformation stemming from KP's HIT" (20) Más de Health IT Conference – iHT2 (20) iHT2 Health IT Summit in Phoenix 2013 – Terhilda Garrido, VP, HIT Tranformation & Analytics, Kaiser Permanente Case Study "Big Data, Little Data: Value and Transformation stemming from KP's HIT"1. Terhilda Garrido, VP HIT Transformation and Analytics
iHT2 Summit
January 2013
Big Data, Little Data:
Value and Transformation Stemming
from
Kaiser Permanente's HIT
HARVESTING VALUE
HIT Transformation & Analytics – National Quality & Care Delivery Excellence
Copyright © 2013
2. Value
About Who we are
Kaiser Permanente Our HIT investment
Clinical Evidence / Guidelines
Organizational metrics
Data –
Patient population
From Big to “Little” MD Feedback
Patient empowerment
Heading to Big(ger) The Opportunities and Challenges
Data
Copyright © 2013
3. About Kaiser Permanente
Nation’s largest nonprofit health plan
Integrated health care delivery
3 Organizations in one: Health insurer, hospital system, physician partnerships
9 million members
16,658 physicians
172,997 employees
Serving 9 states and DC
37 hospitals
611 medical offices
$47.9 billion annual revenues (2012)
Copyright © 2013
4. Our investment in technology:
Kaiser Permanente HealthConnect ®
The world‟s largest and one of the most advanced civilian
deployments of an electronic health record
KP HealthConnect was implemented 2004-2010
specifically to transform care and service delivery
Our greatest benefits are the resulting improvements in
quality and effectiveness of patient care
Copyright © 2013
5. Integrating health care across the
continuum
Inpatient and outpatient clinical care:
documentation, ordering, decision support
Hospital and outpatient practice
management:
billing, scheduling, registration, ADT
(admission, discharge and transfer)
Ancillary systems:
pharmacy, labs, radiology, etc.
Online access to personal health record:
My Health Manager on kp.org
One data model underlying all EHR
components
40 million records
3.7 petabytes of stored information
Copyright © 2013
6. I LOVE data
About For our clinicians
Kaiser Permanente For our system
Clinical Evidence / Guidelines
Organizational metrics
The Miracle of Data Patient population
MD Feedback
Patient empowerment
Standardization & A Goldmine of Data
The Advent of A Hotbed of Innovation
Big Data The Blue Sky Vision
Copyright © 2013
7. What do you know?
Not everything
that can be
measured
counts, and
not everything
that counts can
be measured.
A. Einstein
Copyright © 2013
8. Data is a necessity in our care model
DATA
REPORTABLE
INFORMATION
AMOUNT
GAP
GAP
KNOWLEDGE
CLINICAL UTILITY
TIME
Source: Michael N. Liebman, PhD
Executive Director Windber Research Institute
Copyright © 2013
8
9. “Conclusions: Our findings indicate no consistent
association between Electronic Health Records and
Clinical Decision Support and better quality.”
Archives Internal Medicine
January 24, 2011
Copyright © 2013
10. Kaiser Permanente compared to U.S. market
29 total top rankings – 13 Medicare & 16 commercial – the
most No. 1 rankings for health plans in the country
563 applied for Star ratings
11 plans across the country received 5 stars
6 of those were Kaiser Permanente plans:
California, Hawaii, Northwest, Colorado, Mid-Atlantic
States, Ohio
17 Top Hospitals
33 Hospitals among safest in the nation
Mail-order pharmacy programs rated at the highest level
of customer satisfaction in the entire country four years
J.D. Power in a row
& Associates Highest ranking in member satisfaction among health
plans in all five of our larger regions
Copyright © 2013
11. Information is power at many levels
Information is power: 1. Health care knowledge level
Power to know what the research / guidelines
problem is
Power to know where to act 2. Organization level – Big Q
Power to know what to change
Information driving 3. Patient population
transformation: level
Five levels of information work
4. MD level – performance
together to drive improvement in
feedback
the health care system. They are all
enabled and dependant on data.
5. Patient level –
empowerment
Copyright © 2013
12. 1. Health care
knowledge level
research / guidelines
Copyright © 2013
13. 1. Health Care Knowledge Level:
A tale of two studies
Tale of Study 1 – It was the best of times, it was the worst of
times for researchers in the city of San Antonio…
ALLHAT* – a classic RCT – randomized control trial
42,418 patients
$120 million
Conducted over 8 years
Results – Thiazide-type diuretics are most
effective at controlling hypertension
Given the low cost of this diuretic class, they
postulated $3.1 billion savings over 10 years
ALLHAT* – Antihypertensive , lipid lowering treatment to prevent health attack trial. Sponsored by the national heart, lung and blood institute
Begley, Sharon, "The Best Medicine," Scientific American 305, no. 1 (2011):50-55
Copyright © 2013
14. 1. Health Care Knowledge Level:
A tale of two studies
Tale of study 2 – Meanwhile in Denver, David Magid,
MD had an idea….
ALLHAT – a classic RCT – randomized Magid* study
control trial • 4,000,000 patients
42,418 patients • $200,000
$120 million • Conducted over 1.5 years
Conducted over 8 years
Results – Thiazide-type diuretics are • Results – For the 60% of patients that were not
most effective at controlling under control / complete hypertension control,
hypertension ACE Inhibitors and beta blockers are effective
as 2nd line medications.
*Magid, David J.; Shetterly, Susan J.; Margolis, Karen L.; Tavel, Heather M.; O’Connor, Patrick J.; Selby, Joe V.; Ho, P. Michael, “Comparative Effectiveness of Angiotensin-Converting Enzyme Inhibitors
Versus β-Blockers as Second-Line Therapy for Hypertension”, Circ Cardiovasc Qual Outcomes, 2010;3:453-458.
Copyright © 2013
15. 1. Health Care Knowledge Level:
A tale of two studies
RCTs are the traditional gold standard for health care research. EHRs and their vast
store of data (in depth and breadth) offer an alternative with pros and cons.
ALLHAT Study Magid Study
Statistical Randomized Control Trial Retrospective observational
Design
- Eliminate Patient Bias - Adjust for patient bias:
Matched case control.
N.b. needs a large sample
- Limited target group - Cross-sectional population
- Controlled conditions - Real World conditions
Results $120,000,000 over 8 yrs $200,000 over 1.5 yrs
42,418 patients 4 million patients
Copyright © 2013
17. 2. Organization Level: Big Q
Unlike other industries,
where the common
denominator is often
profit, health care requires
a balanced scorecard
approach.
Clinical
Effectiveness
Safety
Patient satisfaction
Cost efficiency
Equity in care
Copyright © 2013
17
19. 3. Taking accountability for patient
populations
Clinical Outcomes in Southern California
Metric Improvement Lives Saved Per Decade1
Blood Pressure Control 38.9% 5,341 Lives Over
Colorectal cancer screening 30.2% 4,788 Lives
Cholesterol Control 21.8% 1,751 Lives 14,000
Blood sugar control 11.5% 1.088 Lives
Smoking Cessation 17.0% 955 Lives Lives
Breast Cancer Screening
Cervical Cancer Screening
11.4%
5.9%
570 Lives
59 Lives Saved1
Copyright © 2013
21. 3. Patient population tools
The Panel Support tools have been shown to be effective in improving patient
quality of care and reducing “care gaps” by up to 21%
Copyright © 2013
22. 4. MD level –
performance feedback
Copyright © 2013
23. 4. MD Level:
Performance Feedback Loops
Analysts extract data
MD leaders and craft report to
identify info reflect focus
needs Count of Untouched Messages by Location /
Department / Recipient
6 Department Message Recipient Count
XYZ MEDICAL OFFICE DEPT FIRSTNAME, LASTNAME 3
MD
DEPT FIRSTNAME, LASTNAME 15
MD
FIRSTNAME, LASTNAME 3
MD
FIRSTNAME, LASTNAME 2
MD
XYZ MEDICAL OFFICE DEPT FIRSTNAME, LASTNAME 4
MD
Monitoring MD inbaskets DEPT FIRSTNAME, LASTNAME
MD
28
resulted in a 70-90% decrease DEPT FIRSTNAME, LASTNAME
MD
9
Untouched Abnormal Lab Results InBasket Messages > 14 Days
in ‘untouched abnormal lab
results 14 days+’
900
800
700
Dept MD
600 Chiefs share
Count
with MDs
500
400
300
200 MD maintains / improves
100
0
8/09 9/09 10/09 11/09 12/09 1/10
practice. Reports track
August 2009 - January 2010
progress
Copyright © 2013
24. 4. MD Level:
Performance Feedback Loops
MD leaders Analysts extract
identify info data and craft
needs report to reflect
focus
Count of Untouched Messages by Location / Department / Recipient
6 Department Message Recipient Count
XYZ MEDICAL OFFICE DEPT FIRSTNAME, LASTNAME MD 3
DEPT FIRSTNAME, LASTNAME MD 15
FIRSTNAME, LASTNAME MD 3
FIRSTNAME, LASTNAME MD 2
XYZ MEDICAL OFFICE DEPT FIRSTNAME, LASTNAME MD 4
DEPT FIRSTNAME, LASTNAME MD 28
DEPT FIRSTNAME, LASTNAME MD 9
Copyright © 2013
25. 4. MD Level:
Performance Feedback Loops
Monitoring MD inbaskets resulted inLab70- InBasket Messages > 14 Days
Untouched Abnormal a Results
90% decrease in ‘untouched abnormal lab
results 14 days+’ 900
800
MD maintains / 700
600
improves
Count
500
practice. 400
300
Reports track 200
progress 100
0
8/09 9/09 10/09 11/09 12/09 1/10
August 2009 - January 2010
Dept MD Chiefs
share with MDs
Copyright © 2013
27. I feel much more
I try to learn as much as
confident, and I also feel
possible about diabetes. I
I feel more in control over my closer to my physician that
have been controlling it with
medical condition. I have access I had before. I don’t feel
diet and exercise for 3 yrs.
to information, access to people I intimidated anymore.
Dr. says I am doing a good
need to consult with, so it puts me job, thanks to your site.
back in charge. KP NW Member, 2004 Online survey NCal 2010
The website has improved my
relationship with my doctors. It’s so
I was curious what they had to say. I
easy to communicate with them, and
have a heart condition, and the
they are all willing to respond quickly.
words they use are so big that I
I am very impressed with this method
wanted to go back and be able to
and feel it has made taking care of
actually look at what it was. It was a
myself easier and more palatable. I
good thing for me to look at it and to
feel supported in this way because
get it to stick with me that I need
the computer is available 24/7.
to take better care of myself when
I saw it in black and white.
I think it’s empowering. As It’s made it better because I feel
individuals we need to take care of there is a closer partnership. It’s I feel more in touch with
ourselves, and we have to like instead of not seeing him the sum and substance
assume responsibility and it for six months, we have of my medical condition.
gives us that. communicated in between.
Copyright © 2013
28. 5. Patient Engagement Rewards:
Loyalty & Quality of Care
Member Retention
My Health Manager users were 2.6 times more likely
to choose to remain members than were those who
did not use it.1 Following membership tenure and
illness burden, MHM use was the third strongest
predictor of remaining a member. The newer the
member; the stronger the impact of MHM.
Quality of Care Improved!
In a matched-control analysis published in Health
Affairs, secure email was associated with a 2.0 to
6.5% improvement in performance on HEDIS
measures, such as glycemic (HbA1c), cholesterol, and
blood pressure screening and control2
High Patient Satisfaction!
85% rated encounters 8 or 9 on a 1-9 scale3
1 Turley, Marianne; Garrido, Terhilda; Lowenthal, Alex; Zhou, Yi Yvonne, “Association Between Personal Health Record Enrollment and Patient Loyalty,” Am J Manag Care. 2012;18(7):e248-e253 (web exclusive)
2Zhou, Yi Yvonne; Kanter, Michael H; Wang, Jian J; Garrido, Terhilda, “Improved Quality at Kaiser Permanente Through E-Mail Between Physicians and Patients,” Health Affairs, Vol 29, No 7 (2010); 1370-1375.
3 Internal KP study, “Harvesting Value: Early Findings from Kaiser Permanente HealthConnect™” presented to Center for Informati on Therapy by T Garrido, C Serrato, J Oldenburg (1/15/2008)
Copyright © 2013
29. Value
For our clinicians
Connectivity For our system
All info; all patients; all the time
Redefining Access / Empowering patients
The Miracle of Data Tackling Care Gaps
A Backbone for Reliability
Standardization & The Challenges
Heading to Big Data
Copyright © 2013
31. Layers of data
Exogenous Data - e.g. SES (socio economic),
Credit Card Purchases, Weather
Patient Reported Outcomes, Surveys, Fit Bit…
External Data – Claims, Biomed Device..
KP HealthConnect
Copyright © 2013
32. Data, data everywhere
and not a thought to think
The „Big Data‟ era will be characterized by:
Variety Volume
Velocity Veracity
Time is passing and expectations are rising
Copyright © 2013
33. Care Models – Past, Present, Future
Stratified Patient-Centered
Reactive Care
Care Care
Individual is their
Focus Clinical Encounter Care of Cohorts
own „control‟
„Exhaust‟ of A cornucopia of
Data EHR / Clinical data
encounter data sources
Offline data
„Real-time learning
Bench to Bedside „17 years‟ analysis
health care system‟
-months-
Wellness &
$ FFS Value-based care
Prevention
„Wellness
Theme “Illness Cured” „Disease Stratified‟
Quantified‟
Copyright © 2013
34. Our Vision for the Future: “Blue Sky”
Home as the Hub
The home, and other settings, will grow significantly as a
locale of choice for some care delivery (diagnostics).
Integration & Secure and
Leveraging seamless
IT functionality transitions
enables us to “Warm Handoffs”
leverage scarce or The human skill
specialized sets and operational
expensive clinical processes to deliver
resources - MDs, RNs care and service
and other clinical staff. effectively,
We use data to focus efficiently, and
care delivery and compassionately.
create evidence
based practices.
Customization
Occurs at any level of the members‟ journey with Kaiser
Permanente (choosing health plans, cost sharing,
individual care pathways, and communication modalities).
Brentari, Ruth; Garrido, Terhilda; Mittman, Robert; Liang, Louise; Weiland, Allan; Wiesenthal, Andrew; Cordova, Richard (2003) “Blue Sky Care Delivery 2015, Part I,” The Permanente Journal, Vol 7, No. 4, Fall 2003, pgs 47-50
Brentari, Ruth; Garrido, Terhilda; Mittman, Robert; Liang, Louise; Weiland, Allan; Wiesenthal, Andrew; Cordova, Richard (2004) “Blue Sky Part II: The Health Encounter,” The Permanente Journal, Vol. 8, No. 1, Winter 04, pgs 93-97
Brentari, Ruth; Garrido, Terhilda; Mittman, Robert; Liang, Louise; Weiland, Allan; Wiesenthal, Andrew; Cordova, Richard (2004) “Blue Sky Part III: The Chronic Care Encounter,” The Permanente Journal, Vol. 8, No. 2, Spring 04, pgs 71-75
Copyright © 2013
Notas del editor Learn about KP's investment in the EHR and its transformative value Learn how data and access to information has impacted clinical operations, the patient experience and how we approach research Learn how this data is more patient centric and patient empowering. If you assume that the library of congress has 34 terabytes ; then KP has over 100 library of congress’s currently stored. medical malpractice experience between January 2003 through December 2009 shows that delays in diagnosis and treatment comprise 77% of claims medical malpractice experience between January 2003 through December 2009 shows that delays in diagnosis and treatment comprise 77% of claims medical malpractice experience between January 2003 through December 2009 shows that delays in diagnosis and treatment comprise 77% of claims