Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 255
Appendix B
Case Studies for the Digital
Health Infrastructure
THE caBIG® INITIATIVE
Prepared by Ken Buetow (National Cancer Institute)
Overview
The National Cancer Institute (NCI) has developed an informatics
program designed to improve patient care and accelerate scientific discov-
eries by enabling the collection and analysis of large amounts of biological
and clinical information and facilitating connectivity and collaboration
among biomedical researchers and organizations. Called caBIG® (cancer
Biomedical Informatics Grid), this program is developing the foundational
informatics infrastructure to improve health and combat disease emerging
at the intersection of life sciences, information technology, and medicine.
The caBIG® program has three core components: community, connectivity,
and content.
The caBIG® Community
The caBIG® program has been from the start a collaborative endeavor.
Its community has grown dramatically in size and scope since the program
began in 2004. More than 2,200 individuals representing more than 700
different organizations are actively engaged in caBIG®, and participation
is steadily increasing. These individuals include basic and clinical research-
ers, consumers, physicians, advocates, software architects and developers,
bioinformatics specialists, and executives from academe, medical centers,
255
OCR for page 256
256 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
government, and commercial software, pharmaceutical, and biotechnology
companies from the United States and in 15+ countries around the globe. In
support of the development of a Rapid Learning Health System, the caBIG®
participation includes both academic and community cancer centers.
The caBIG® program is organized into workspaces focused on spe-
cific domains such as medical imaging, IT architecture, and clinical trials.
Subject matter experts and developers within each workspace use virtual
conferences and regular face-to-face meetings to work collaboratively on
domain-specific issues and projects. The entire caBIG® community meets
once per year for the Annual Meeting, whose growth in size and scope
year-over-year has mirrored that of the caBIG® program. More than 1,100
individuals representing more than 300 organizations and 13 countries
attended the 2009 meeting, held in Washington, DC, where participants
celebrated the first 5 years of the caBIG® program by planning new ap-
plications and research uses.
While the caBIG® community is highly diverse, its members have simi-
lar needs for data management and analysis.
Through participation in the caBIG® program, they are able to access
the informatics infrastructure required to work productively, advancing
the knowledge of the underlying causes of disease and providing improved
patient outcomes.
Connectivity
From a technology perspective, caBIG® is centered on four key
principles:
• Open development—Planning, testing, validation, and deployment
of caBIG® tools and infrastructure are open to the entire research
community, and contributions from many organizations ensure
applicability to a wide range of common research problems.
Open access—caBIG® is open to all individuals and organiza-
•
tions interested in solving their data management and connectivity
challenges, thus ensuring widespread access to tools, data, and
infrastructure.
Open source—The underlying software code of caBIG® tools is
•
freely available for use and modification by any organization, pub-
lic or private, thus encouraging commercial partnerships.
• Federation—Data and analytical resources can be controlled locally
or integrated across multiple sites. Control of secure access to those
resources is retained by the originating organization. This federated
approach obviates the need for a central authority and reduces data
management overhead.
OCR for page 257
257
APPENDIX B
Together, these four organizing principles ensure the availability of
robustly designed tools that address a wide range of basic and clinical
research requirements. Central to these needs is the requirement for in-
teroperability—the ability to access and make meaningful use of data and
information by multiple systems. By building and deploying IT based on
industry-recognized standards, and providing application programming
interfaces and software development kits for third-party developers, the
process of creating new caBIG®-compatible software or adapting existing
software to become caBIG®-compatible is simplified, encouraging partner-
ships across the IT community.
At the heart of the caBIG® program—invisible to the end user and
customized for the specific needs of biomedical researchers—is caGrid, a
model-driven, service-oriented architecture that provides standards-based
core “services,” tools, and interfaces so the community can connect to
share data and analyses efficiently and securely. More than 120 organiza-
tions are connected to caGrid. The number continues to expand as more
NCI-designated Comprehensive Cancer Centers, as well as academic and
commercial organizations set up nodes on the grid, making caGrid the larg-
est biomedical research network in the world today.
Such caBIG®-enabled connectivity is not limited to organizations
conducting cancer research. In partnership with the American Society of
Clincal Oncology, caBIG® is developing specifications and services to sup-
port oncology-extended electronic health records (EHRs) that are being
deployed in community practice and hospital settings. caBIG® tools and
technology are also being used by researchers working on cardiovascular
health, arthritis, and AIDS. In addition, pilot projects have successfully
connected caGrid to other networks, including the Nationwide Health
Information Network, the CardioVascular Research Grid, and the compu-
tational network TeraGrid.
Ultimately, the vision of the caBIG® program is to provide the techni-
cal infrastructure for a resource called the Biomedical Knowledge Cloud.
The Biomedical Knowledge Cloud is a means of using the Internet to con-
nect massive amounts of individual and organizational biomedical data,
software applications with which to handle and analyze all those data, and
the computational horsepower to do the work. The Cloud is bounded with
patient privacy and other data-sharing protections.
Although the components of the Cloud are not new, the concept of
joining them together to provide seamless, secure access to biomedical
information is just being realized today.
Simply connecting organizations and individuals would be of little
value if they were not able to access and understand the data and informa-
tion created by different laboratories. Currently, many software systems,
either commercial or created in-house to serve the needs of a particular
OCR for page 258
258 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
institution, are based on proprietary data formats and, as a result, cannot
readily exchange data. They lack the ability to be interoperable.
Since its inception, the caBIG® program has worked with key health-
care industry organizations to create or expand standards and common
vocabularies that describe data types, analytical processes, and even clinical
procedures. By developing tools that adhere to these standards and lever-
age these vocabularies, institutions can exchange data seamlessly, much
the same way that agreed-upon interoperable standards enabled the global
banking system to develop the interconnected network of ATMs for con-
sumer use.
caBIG® is based on the belief that strong confidentiality, privacy, and
security measures are both necessary and feasible in any electronic health
information exchange environment, and that the measures can be scaled to
accommodate a broad range of participants without unnecessarily imped-
ing scientific discovery and medical progress. The program has developed
robust computer security measures to ensure that only researchers with the
appropriate credentials have access to data, as well as guidelines to assist
those researchers in determining the sensitivity of their data for sharing.
Content
As scientific understanding of the complexities underlying most diseases
increases, the interconnectedness of biological systems becomes more and
more obvious, and researchers must apply multidisciplinary analysis tech-
niques to large, diverse datasets to make new discoveries. For example, by
correlating the activity of a specific set of genes with observed outcomes
from large groups of patients on the same treatment protocol, cause-and-
effect relationships can be found that would be missed when examining the
expression patterns alone, or when looking at a small group of patients.
By enabling researchers to work collaboratively—leveraging large, di-
verse datasets—all constituents of the health care community reap the ben-
efits. Biomedical researchers can ask and answer more complex questions
that help uncover the underlying causes of disease, speeding the develop-
ment of novel diagnostics and therapeutics. Healthcare providers can stay
current on new treatments and outcome information gathered from large
populations of patients who have similar diseases. This capability allows
them to provide the best treatment options for their patients, ultimately im-
proving patient outcomes. The patients themselves are assured of receiving
optimal treatment regardless of their physical location, since their complete
medical history is secure yet available as needed to guide their treatment.
OCR for page 259
259
APPENDIX B
FDA’S SENTINEL INITIATIVE
Prepared by Judy Racoosin (Food and Drug Administration)
Background
In May 2008, the Department of Health and Human Services (HHS)
and the Food and Drug Administration (FDA) announced the launch of
FDA’s Sentinel Initiative, a long-term program designed to build and imple-
ment a national electronic system (Sentinel System) for monitoring the
safety of FDA-approved drugs and other medical products. The Sentinel
System will function as an active postmarket safety monitoring system,
augmenting FDA’s existing safety monitoring systems. The launch of the
Sentinel Initiative followed passage of the Food and Drug Administration
Amendments Act (FDAAA), which became law in September 2007. Section
905 of FDAAA mandates FDA to develop an enhanced ability to monitor
the safety of marketed drugs using automated healthcare data sources. The
Sentinel Initiative is designed to manage the development and implemen-
tation of the Sentinel System while ensuring that it fulfills the mandates
included in FDAAA.
Once developed and implemented, the Sentinel System will enable FDA
to monitor the safety of drugs and other medical products with the assis-
tance of a wide array of collaborating institutions throughout the United
States. Data partners in the Sentinel System will include organizations
such as academic medical centers, healthcare systems, and health insur-
ance companies. As currently envisioned, participating data partners will
access, maintain, and protect their respective data, functioning as part of a
“distributed system.” Collaborating organizations will also include patient
and healthcare professional advocacy groups, academic institutions, and
regulated industry, among others.
In the active surveillance environment of the Sentinel System, FDA will
prioritize safety questions that have emerged from premarket or postmarket
safety data sources (e.g., clinical trial data, postmarket adverse event re-
ports) and submit them to a Coordinating Center for evaluation by data
partners that are participating in the Sentinel System. Data partners will se-
curely access their databases to evaluate the submitted question and return
Health Insurance Portability and Accountability Act (HIPAA)–compliant
result summaries to the Coordinating Center. The Coordinating Center will
then aggregate and/or summarize these results and forward them to FDA
for their use in assessing the safety question.
OCR for page 260
260 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
Mini-Sentinel
In September 2009, FDA awarded a contract to Harvard Pilgrim
Health Care, Inc. (Harvard Pilgrim) to develop a smaller working version
of the future Sentinel System. The pilot has been dubbed “Mini-Sentinel.”
Harvard Pilgrim’s Mini-Sentinel Coordinating Center (MSCC) is creat-
ing a kind of laboratory, giving FDA the opportunity to access disparate
automated healthcare data sources and test epidemiological and statistical
methodologies in the evaluation of postmarket safety issues. Through this
pilot, FDA will learn more about some of the barriers and challenges, both
internal and external, to establishing a Sentinel System for medical product
safety monitoring. The MSCC is leading a consortium of more than 20 col-
laborating institutions.1
The MSCC and participating data partners are using a common data
model as the basis for their analytical approach. The approach requires the
data partners to transform their data into a standardized format. Based on
this standardized format, the MSCC will write a single analytical software
program for a given safety question and provide it to each of the data part-
ners. This will allow each data partner to run the program on its standard-
ized data. Data partners will conduct analyses behind their existing, secure
firewalls and send only summary results to the MSCC for aggregation and
further evaluation. The MSCC will provide FDA with both the aggregated
results and the summary results from each data partner. The use of a com-
mon analytical program will minimize the potential for differences in results
across data holders resulting from differences in the implementation of an
active surveillance protocol. As this pilot is being implemented, a governance
structure is being developed to ensure that the activity encourages broad
collaboration within appropriate guidelines for the conduct of public health
surveillance activities. In order to accomplish that, the MSCC is developing
a Statement of Principles and Policies that will include descriptions of the
organizational structure and policies related to communication, privacy,
confidentiality, data usage, conflicts of interest, and intellectual property.
Also, with the launch of Mini-Sentinel, a Privacy Panel was formed to
provide expertise regarding patient privacy-related regulations that pertain
1 The collaborating institutions in the consortium include the following organizations:
CIGNA Healthcare; Cincinnati Children’s Hospital Medical Center; Brigham and Women’s
Hospital; Duke University School of Medicine; HMO Research Network sites (includes
Group Health Cooperative, Harvard Pilgrim Health Care Institute, HealthPartners, Henry
Ford, Lovelace Clinic Foundation, Marshfield Clinic Research Foundation, Meyers Primary
Care Institute [Fallon]); HealthCore; Humana-Miami Health Services Research Center; Kaiser
Permanente (includes KPNC, KPSC, KPCO, KPNW, KPG, KPHI, KPOhio, KPmidatlantic);
Outcome Sciences, Inc.; University of Illinois at Chicago; University of Iowa, College of Public
Health; University of Pennsylvania School of Medicine; Vanderbilt University School of Medi-
cine; Weill Cornell Medical College.
OCR for page 261
261
APPENDIX B
to the conduct of Mini-Sentinel. The panel members are independent pri-
vacy experts with extensive knowledge of legal and ethical issues related
to the use of protected health information (PHI), including applicable laws
and regulations, data privacy and confidentiality, and the use of PHI for
public health surveillance activities. The panel is
• Providing expertise on application of relevant laws and regula-
tions governing the privacy and security of health information for
Mini-Sentinel’s purposes. The panel advises specifically on the ap-
plicability of laws, including HIPAA and other relevant laws and
regulations.
• Making recommendations on creation of appropriate policies and
procedures to guide specific uses of PHI in Mini-Sentinel.
• Assisting the MSCC in reviewing documents, agreements, and con-
tracts to ensure that they adequately incorporate the panel’s discus-
sions regarding the issues delineated above.
Federal Partners’ Collaboration
FDA is furthering the science of medical product surveillance by broad-
ening existing pilot programs that use federally held data sources. The
effort, known as the Federal Partners’ Collaboration (FPC), involves the
Centers for Medicare & Medicaid Services (CMS), the Veterans Adminis-
tration (VA), and the Department of Defense. FPC is an expansion of the
SafeRx project, a collaboration between FDA and CMS that uses Medicare
and Medicaid data for medical product safety surveillance. The FPC is
similar to the Mini-Sentinel pilot in that it will use an active surveillance
approach and involves a distributed system. However, unlike Mini-Sentinel,
the FPC will not use a common data model. Rather, the FPC will develop a
common active surveillance protocol, and then each data partner will write
analytical code to run the protocol in their database. Lessons learned from
this pilot will be compared to lessons learned as part of Mini-Sentinel (i.e.,
using a common data model where centralized analytics are employed). In
this way, FDA can compare the potential benefits and drawbacks of every
data partner running a single analytical program based on a common data
model versus each data partner developing its own analytical program
based on a common protocol.
Outreach Related to Active Medical Product Surveillance
Continuing a high level of stakeholder involvement is key to maintain-
ing broad-based support and momentum for effective, responsive, active
medical product surveillance through FDA’s Sentinel Initiative. FDA has
OCR for page 262
262 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
awarded a cooperative agreement to the Brookings Institution (Brookings)
to convene a broad range of stakeholders to explore and address method-
ological, data development, technical, and communication issues related to
active medical product surveillance.
These meetings encompass a range of formats including webinars on
active surveillance–related topics, expert panels on targeted topics central
to the development of the Sentinel System, public meetings on the prog-
ress of the Sentinel Initiative, and meetings intended to interweave lessons
learned by various active surveillance initiatives.
In addition to convening and moderating each meeting, Brookings is
synthesizing findings and making them publicly available in order for other
organizations and individuals to use the information to further develop ac-
tive medical product surveillance methods and systems.
In addition, FDA has sought to foster transparency through the cre-
ation and maintenance of a Sentinel Initiative Website, which provides
information on the background of the initiative, relevant news and events,
presentations, completed deliverables from contracted work, and updates
on ongoing projects and funding opportunities. A docket is open to allow
for public comment on any of this information. The agency also piloted
a web-based discussion room to encourage public comment and promote
transparency. Sentinel Initiative staff also foster transparency by speak-
ing about the Sentinel Initiative to external stakeholder groups including
academia, regulated industry, patient and consumer groups, and medical
professional societies, as well as doing internal outreach to FDA staff.
Contracts and Cooperative Agreements Informing FDA
on the Development of the Sentinel System
The following documents are now available in the FDA docket and
on the Sentinel website http://www.fda.gov/Safety/ FDAsSentinelInitiative/
default.htm.
• Developing a Governance and Operations Structure for the Senti-
nel Initiative, an eHealth Initiative Foundation report.
• Engagement of Patients, Consumers, and Healthcare Professionals
in the Sentinel Initiative, an eHealth Initiative Foundation report.
Defining and Evaluating Possible Database Models, a Harvard
•
Pilgrim Health Care, Inc. report.
Evaluation of Existing Methods for Safety Signal Identification, a
•
Group Health Cooperative Center for Healthcare Studies report.
• Evaluation of Potential Data Sources for a National Network of
Orthopedic Device Implant Registries, an Outcome Sciences, Inc.
report.
OCR for page 263
263
APPENDIX B
• Evaluation of Timeliness of Medical Update for Surveillance in
Health Care Databases, an IMS Government Solutions report.
• Evaluating Potential Network Data Sources for Blood and Tissue
Product Safety Surveillance and Studies, a Pragmatic Data report.
• Evaluation of State Privacy Regulations and Relation to the Senti-
nel Initiative, a Qual-Rx report.
Work on the following projects is ongoing.
Evaluation of Potential Data Sources, a Booz Allen Hamilton report.
•
• Evaluation of Potential Data Sources for Animal Drugs Used in
Veterinary Medicine, an Insight Policy Research, Inc. report.
• Detection and Analysis of Adverse Events to Regulated Products
in Automated Healthcare Data: Efforts to Develop the Sentinel.
Additional Background
FDA’s Sentinel Initiative, launched in 2008, is a long-term program
designed to build and implement a national electronic system (the “Senti-
nel System”) for monitoring the safety of FDA-approved drugs and other
medical products. The Sentinel System will fulfill, and go beyond, the con-
gressional mandate put forth in the FDAAA of 2007 that requires FDA to
create a system for postmarket risk identification and analysis using public
and private automated data sources.
Engaging Patients and the Public
Broad engagement of all stakeholders is essential to developing a gover-
nance framework that addresses the many issues that need to be considered
for the successful implementation of a system that leverages secondary use
of automated healthcare information to improve the public health.
Early activities of the Sentinel Initiative included meetings with a
•
broad range of stakeholders including other government agencies,
potential data partners, patient advocacy groups, professional so-
cieties, academia, and regulated industry.
To ensure continued input of all stakeholder groups, in 2009 FDA
•
awarded a cooperative agreement to the Brookings Institution to
function as a convener on topics related to active medical product
safety surveillance. They have sponsored large public meetings on
the Sentinel Initiative, smaller expert panel discussions on focused
topics, and webinar-style roundtables to discuss new methodolo-
gies and findings.
OCR for page 264
264 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
Paramount to this effort is safeguarding the privacy and security of the
data leveraged for these activities, and ensuring that patients and consumers
understand that their privacy is being protected. This includes compliance
with federal and state laws and regulations and consideration of additional
forms of protection.
As part of the Mini-Sentinel pilot, a privacy panel consisting of
•
established experts in the field, including patient advocates, has
been convened to address privacy and security concerns.
Promoting Technical Advances and Innovation
After careful consideration of all options, FDA concluded that a distrib-
uted system is essential to the success of efforts such as the Sentinel Initia-
tive. The key benefits of this distributed approach include the following:
Patient privacy is maintained by keeping directly identifiable pa-
•
tient information behind local firewalls in its existing protected
environment.
Data partners’ involvement in running analyses ensures an in-
•
formed approach to interpreting results because they are aware
of the changes that have occurred in their healthcare systems that
result in the unique character of their database.
Patient and consumer concerns about potential misuse are greatly
•
reduced. A distributed system can operate efficiently, compared to
a centralized system, by having the data partners adopt a common
data model and then creating the capability for efficient distribu-
tion of queries (executable computer programs) and return of their
output.
Through the Mini-Sentinel pilot, FDA has developed the first stage
•
of such a system, which will ultimately include administrative
claims, EHRs, and registry data.
We must develop methods to link information about patients between
data sources (e.g., inpatient hospital records, outpatient records, and pa-
tient registries) in order to produce a complete longitudinal profile of
patient care. We must train the next generation of statisticians and epidemi-
ologists to ensure that we will have a workforce with the skills to support
active medical product safety surveillance.
OCR for page 265
265
APPENDIX B
Fostering Stewardship and Governance
Governance and stewardship should allow multiple activities, includ-
ing both public health practice and research, to share a common national
resource that allows the use of automated healthcare data for secondary
uses, such as safety surveillance, comparative effectiveness research, and
quality assessment.
HEALTHWISE
Prepared by Cartherine Serio and Don Kemper (Healthwise)
About Healthwise
Nonprofit mission: Help people make better health decisions (since 1975)
110 million user sessions with Healthwise content in 2009
Current Clients:
The top 10 U.S. health plans (Kaiser, Aetna, Cigna, Wellpoint, etc.)
•
300-plus U.S. hospitals (Mass General, Sutter, Sisters of Mercy,
•
Health Partners, etc.)
Leading disease management companies (Health Dialog, Alere,
•
etc.)
Government agencies (Department of Veterans Affairs, Department
•
of Defense, State Medicaid, British Columbia, etc.)
Most large health portals (WebMD, AOL, Yahoo, MSN, Health.
•
com, etc.)
Innovation History
1970s: Medical self-care handbooks and workshops (28 million
•
books sold)
1980s: Wellness books and workshops
•
1990s: Healthwise Knowledgebase for Websites and nurse call
•
centers
o Patient decision aids (now 158)
o Symptom guides, action plans, tests, treatments, and self-care
2000s: Information therapy, interactive conversations, guided
•
self-management
Promoting the consumer as the greatest untapped resource in health
care.
OCR for page 266
266 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
Healthwise Engagement Tips
1. Think like a consumer (patient-centric engagement)
• Understand consumers’ motivation for engagement
• Anticipate consumers’ needs
• Design toward those needs
2. Engagement ingredients
• Establish trust
• Use plain language
• Allow choice
3. Engagement strategies
• ull: How to engage people who are ready to “pull” information
P
from the Web.
• ush: How to engage people by “pushing” information relevant
P
to their needs.
• ay: How to engage even more people through incentives for
P
active patienthood.
4. Engage the consumer’s community
• ocial networking allows consumers to get, and give, emotional
S
support
KAISER PERMANENTE2
Health Information Technology (HIT) at Kaiser Permanente
Kaiser Permanente has been using information technology for more
than 40 years to improve clinical and administrative functions. Its use of
electronic health records (EHRs) dates from the 1990s in some regions.
Building on this experience, and with the active participation of its physi-
cians, Kaiser Permanente in 2003 launched a $4 billion health information
system called KP HealthConnect that links its facilities nationwide and rep-
resents the largest civilian installation of EHRs in the United States. As of
April 2008, the system was successfully implemented in outpatient clinics in
all eight Kaiser regions. Every Kaiser hospital has the essential components
of the system and 25 had implemented all modules as of December 2008.
2 Prepared by Roundtable Staff using the following sources:
Healthcare Information and Management Systems Society (HIMSS) Analytics. Kaiser
Permanente—EHR Adoption Model. http://www.himssanalytics.org/hc_providers/
stage7casestudies_KP.asp.
McCarthy, D., and K. Mueller. 2009. Kaiser Permanente: Bridging the Quality Divide with
Integrated Practice, Group Accountability, and Health Information Technology. The
Commonwealth Fund. http://www.commonwealthfund.org/~/media/Files/Publications/
Case%20Study/2009/Jun/1278_McCarthy_Kaiser_case_study_624_update.pdf.
OCR for page 267
267
APPENDIX B
The EHR at the heart of KP HealthConnect (purchased from vendor
Epic Systems Corp.) provides a longitudinal record of member encounters
across clinical settings and includes laboratory, medication, and imaging
data. HP HealthConnect also incorporates
• Electronic prescribing and test ordering (computerized physician
order entry) with standard order sets to promote evidence-based
care.
• Population and patient-panel management tools such as disease
registries to track patients with chronic conditions.
• Decision support tools such as medication-safety alerts, preventive
care reminders, and online clinical guidelines.
• Electronic referrals that directly schedule patient appointments
with specialty care physicians.
• Performance monitoring and reporting capabilities.
• Patient registration and billing functions.
KP HealthConnect is designed to electronically connect members to
their health care team, to their personal health information, and to rel-
evant medical knowledge to promote integrated health care. For example,
members can complete an online health risk assessment, receive customized
feedback on behavioral interventions, participate in health behavior change
programs, and choose whether to send results to KP HealthConnect to
facilitate communication with their physician.
To more fully engage patients in their care, physicians and staff en-
courage them to sign up for enhanced online services. As a result, more
than one-third of health plan members nationwide (and nearly one-half of
members in Northern California) are using a web portal called My Health
Manager to track selected medical information from the EHR, view a his-
tory of physician visits and preventive care reminders, schedule and cancel
appointments, refill prescriptions, and send secure electronic messages to
their care team or pharmacist. Online laboratory test results—the most
popular online function—include links to a knowledge base of information
on test results and related self-care strategies. A pilot project is testing the
capability for members (initially Kaiser employees) to transfer information
securely from My Health Manager to Microsoft Corporation’s HealthVault
personal health record application.
Physician leaders report that access to the EHR in the exam room is
helping to promote compliance with evidence-based guidelines and treat-
ment protocols, eliminate duplicate tests, and enable physicians to handle
multiple complaints more efficiently within one visit. A study in the North-
west region found that patient satisfaction with physician encounters in-
creased after the introduction of the EHR in exam rooms there. Early
OCR for page 268
268 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
findings from ongoing hospital implementations suggest that the combina-
tion of computerized physician-order entry, medication bar coding, and
electronic documentation tools is helping to reduce medication administra-
tion errors.
Use of the EHR and online portal to support care management and
new modes of patient encounters appears to be having positive effects on
utilization of services and patient engagement. For example, three-quarters
or more of online users surveyed agreed that the portal enables them to
manage their health care effectively and that it makes interacting with the
health care team more convenient. Patients in the Northwest region who
used online services made 10 percent fewer primary or urgent care visits
than before they had online access (7 percent fewer visits compared with a
control group of patients).
HIT in Practice: Care Coordination and Transitions
Having a broad spectrum of services available within one organization
and, in many cases, in one location, makes it easier to coordinate care for
patients. Kaiser Permanente’s integrated model of care focuses not only on
the spectrum of medical care that a patient may need at any one time, but
also on members’ interactions with the organization across time and the
continuum of care—clinic, hospital, home, hospice, or extended care.
The Northern California region uses a population and patient-panel
management strategy to improve care and outcomes for patients who
have—or who are at risk for developing— chronic diseases. This approach
is built on the philosophy that a strong primary care system offers the most
efficient way to interact with most patients most of the time, while recogniz-
ing that some patients need additional support and specialty care to achieve
the best possible outcomes. Patients are stratified into three levels of care:
1. Primary care with self-care support for the 65 percent to 80 percent
of patients whose conditions are generally responsive to lifestyle
changes and medications.
2. Assistive care management to address adherence problems, com-
plex medication regimens, and comorbidities for the 20 percent
to 30 percent of patients whose diseases are not under control
through care at level 1.
3. Intensive case management and specialty care for the 1 percent to 5
percent of patients with advanced disease and complex comorbidi-
ties or frailty.
Level 1 emphasizes a proactive team approach that conserves physi-
cian time for face-to-face encounters by enhancing the contributions of
OCR for page 269
269
APPENDIX B
ancillary staff (medical assistants and also nurses and pharmacists in some
locations) to conducting outreach to patients between visits. The team uses
a population database and decision support tools built into the EHR to
track patients with chronic conditions such as diabetes or heart disease,
develop action plans to engage them in self-care, ensure that they are tak-
ing appropriate medications, and remind them to get preventive care and
other tests when needed.
Outreach to patients with chronic conditions typically occurs as follows:
The physician reserves a weekly appointment slot to meet with his or her
staff and review a computer-generated list of 10 to 20 patients who are not
achieving treatment goals. The physician indicates follow-up instructions for
each patient, such as increasing medication dosage or ordering a test. The
medical assistant or nurse then contacts the patient to relay the physician’s
instructions, using prepared scripts to ensure consistent communication.
Contact is typically made by telephone but may occur by letter in some cases.
At level 2, care managers (specially trained nurses, clinical social work-
ers, or pharmacists) support the primary care team to help patients gain
control of a chronic condition. Interventions may include providing self-
care education, titrating medications according to protocol, and making
referrals to educational classes (e.g., for smoking cessation). The goal is to
move patients back to level 1 after an intervention period of several months
to a year. Successful transitions require that primary care teams be prepared
to follow up with patients and prevent them from relapsing. Care managers
may be part of the local primary care team or may be centrally located at
a medical center, depending on local resources.
An example of intensive case management (level 3) is a cardiac reha-
bilitation program called Multifit for patients with advanced heart disease,
such as those recovering from a heart attack or heart surgery. Nurse case
managers provide telephonic education and support for up to 6 months
to help patients make lifestyle changes and reduce their risk of future
cardiac events. Aided by the EHR and a patient registry, the Colorado
region enhanced the program by adding a telephonic cardiac medication
management service provided by clinical pharmacy specialists, with ongo-
ing follow-up until patients achieve treatment goals and can be transferred
to primary care for maintenance. Results for patients participating in the
Colorado program included the following:
• Cholesterol screening increased from 55 percent to 97 percent of
patients, while cholesterol control has almost tripled from 26 per-
cent to 73 percent of patients.
• Relative risk of death declined by 89 percent among those enrolled
in the program within 90 days of a cardiac event, and by 76 per-
cent for those with any contact with the program.
OCR for page 270
270 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
HIT as an Engine for Continuous Innovation
Developing Improved Modes of Care Delivery
The 21st Century Care Collaborative is using KP HealthConnect to
develop innovations that will transform the ability of primary care teams to
improve patient care delivery and member experience while also promoting
a sustainable work environment for clinicians and staff. A prototype change
package—developed from the experience of several pilot-test sites—is being
spread regionally using a flexible approach that lets facilities and teams test
elements to determine what works best in their circumstances. Principles
and examples include the following:
1. Understand the needs of your population: Design the work and
build the care team to meet the needs; for example, maximize team
roles and optimize team communication.
2. Develop relationship-based care and demonstrate that we know
members; for example, convene member councils, complete after-
visit summaries.
3. Provide alternatives to traditional office visits; for example, offer
telephone visits and group visits, use secure messaging.
4. Embrace total panel ownership; for example, conduct outreach to
patients with chronic conditions, follow up with patients on new
medicines.
5. Engage members in collaborative care planning; for example, use
goal sheet with diabetic patients, convene chronic care support
groups.
These changes have synergistic effects. For example, replacing face-
to-face visits with telephone visits saves time and increases convenience
for members. It also frees time for the care team to conduct proactive
panel-management activities, address urgent-care needs, and look for other
opportunities to make things easier for patients, such as by calling those
on the appointment schedule to resolve problems over the phone. Pilot
sites reported improved quality and increased satisfaction for members
and staff.
Pursuing Advances in Medicine
In Northern California, Kaiser Permanente’s Division of Research con-
ducts epidemiological and health services research to improve the health
and medical care of members and the population at large. A major cur-
rent project is assembling one of the world’s largest biobanks of genetic,
OCR for page 271
271
APPENDIX B
environmental, and health data. The biobank will enable research on the
causes of diseases that eventually may lead to advances in diagnosis, treat-
ment, and prevention. Almost 400,000 Northern California members have
volunteered to participate in the program by completing a health survey and
are being asked to contribute saliva samples for DNA analysis.
Lessons Learned
The Right People at the Table
It’s imperative that clinicians play a significant role in the planning,
design and implementation of an EHR system. They use the system day in
and day out, so they need to be involved in the decision-making process. If
not, you end up with just a fancier version of the paper record. In designing
HealthConnect, hundreds of stakeholders and IT experts worked together
for months to figure out the functions the system needed to best serve its
members.
Training Is Integral to Success
A large portion of costs were attributable to training and workflow
re-design. A great deal of time and energy was spent to accommodate the
ramping-up process after the system was implemented. Kaiser has con-
tinued with the training and exchanges of best practices and believes it must
be an ongoing process.
Don’t Underestimate the Desire to Do the Right Thing
It would be unrealistic to say that every doctor switched over to elec-
tronic records without any issue. The transition was much more of a culture
shock for doctors who had been using paper records for 30 or 40 years.
Some were more resistant to change than others, which can be expected in
a project of this size. At the end of the day, though, clinicians understood
that what was being done was in the best interest of the patient.
OCR for page 272
272 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
THE SMART GRID3
Conception
The Smart Grid was mandated at the federal level. Title XII of The
Energy Independence and Security Act of 2007 stated that the National
Institute for Standards and Technology has “primary responsibility to co-
ordinate development of a framework that includes protocols and model
standards for information management to achieve interoperability of smart
grid devices and system.”
Vision
As conceived, the Smart Grid will
• Enable active participation by consumers
• Accommodate all generation and storage options
• Enable new products, services, and markets
• Provide power equality for the digital economy
• Optimize asset utilization and operate efficiently
• Anticipate and respond to system disturbances (self-heal)
• Operate resiliently against attack and natural disaster
Governance
At the recommendation of the Federal Energy Regulatory Commission,
a profit-neutral organization known as an independent system operator or
regional transmission organization will be charged with coordinating and
managing the operations of the grid. The scale of these organizations is
variable (e.g., local, state, regional).
Consumers Energy
In response to a request from Consumers Energy’s president and chief
executive officer, Dave Joos, a small company team started investigating
the smart grid in early 2007. Since then, Consumers Energy has created
the Smart Services Learning Center, a smart-grid testing and demonstra-
tion facility, to assess vendor products and provide product and integration
testing. The company performs product field tests by using strategically
deployed off-grid meters that are tied back wirelessly to the center.
Consumers Energy’s initial testing and assessment has revealed a clear
3 Prepared by Roundtable staff.
OCR for page 273
273
APPENDIX B
lack of product standardization and integration throughout the industry.
With this in mind, the company believes more standardization must be
achieved before beginning a full-scale smart-grid system implementation.
Company officers, managers, and directors understand the risks involved
with deploying systems too quickly and support efforts to further assess
and improve vendor products.
Consumers Energy is working with other industry experts to help estab-
lish standards and guidelines for an integrated, secure smart-grid systems
environment. For example, the company helped create the SAP Lighthouse
Council, a group of leading utilities and vendors that collaborate with SAP
to develop standardized software and interfaces for smart-grid systems and
devices.
Employees also are helping ensure that standardized interfaces are built
into customer and grid-based devices to allow for easy connectivity with new
utility systems and devices. They have been actively involved with other utili-
ties, vendors, standards organizations, and regulators to ensure that appro-
priate security capabilities are built into the systems, and that systems can
be updated easily when new security threats arise. Upgradability, standard
interfaces, and vigorous testing are the best methods for minimizing risks,
avoiding product obsolescence, and lowering product costs.
The Model
Fundamentally, the Smart Grid is a long-term, complex systems devel-
opment project of nationwide scale and implications. It uses the engineer-
ing approach of accommodating a wide variety of legacy nodes that are
organic—constantly growing and evolving like a biological system. This
continuous evolution is desirable, so that the Smart Grid’s architecture can
preserve, and indeed encourage, the capacity of every node to innovate
locally and deal with complexity in a way that suits local and grid needs.
The Smart Grid development methodology is not based on compre-
hensive internal design and operating standards for each node on the grid
to follow. There is no need for consensus among the nodes on how they
should operate within local boundaries. Instead, the approach accommo-
dates highly diverse nodes connecting to the Smart Grid using open data
translation protocols that standardize information management, rather
than using the internal workings of each node. The grid becomes a com-
munications bus to which each node must be able to write, and from which
each node must be able to read. This architecture preserves capacities for
local operating autonomy and innovation throughout the Smart Grid. It
is also manages a standardized communications capacity among complex,
and otherwise noninteroperable, legacy nodes on the grid. These features
are all characteristics of ultra-large-scale (ULS) software intensive systems.
OCR for page 274
274 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
VISA AND THE “CHAORDIC” GOVERNANCE MODEL4
By Chaord, I mean any self-organizing, adaptive, non-linear, complex sys-
tem, whether physical, biological, or social, the behavior of which exhibits
characteristics of both order and chaos or, loosely translated to business
terminology, cooperation and competition.—Dee Hock
Background
In the 1960s, the nascent credit card industry witnessed rapid growth,
quickly outgrowing the governance structure of most companies. Although
membership and participation were booming, companies were losing
money. As the decade neared a close, a lack of organizational regulation
and control left the industry with estimated losses in the tens of millions.
Foundations of a New Model
In an attempt to regain control of the industry, Bank of America formed
a small committee to devise solutions to operational problems. In his role
as chair of this committee, Dee Hock began to construct a new governance
model. According to Hock, what emerged from the meeting was a set of
several principles, framed as “what if” questions:
• What if the organization were cooperatively and equitably owned,
with all relevant and affected parties eligible to participate in func-
tions, governance and ownership?
• What if power and function were distributive, with no power
vested in or function performed by any part that could reasonably
be exercised by any more peripheral part?
• What if it were self-organizing, with participants having the right
to self-organize at any time, for any reason, at any scale, with ir-
revocable rights of participation in governance at any greater scale?
4 Prepared by Roundtable Staff using the following sources:
Fowler, J. Gone chaordic: Dee Hock, the mastermind behind VISA, has some ideas about
reorganizing health care. Health Forum Journal.
Hock, D. W. 1995. The chaordic organization: Out of control and into order. World Business
Academy Perspectives 9(1).
Waldrop, M. M. 1996. Dee Hock on organizations. Fast Company. http://www.fastcompany.
com/magazine/05/dee3.html.
Waldrop, M. M. 1996. The Trillion-Dollar Vision of Dee Hock. Fast Company. http://www.
fastcompany.com/magazine/ 05/deehock.html?page=0%2C0.
OCR for page 275
275
APPENDIX B
• What if governance were distributive, with no individual, institu-
tion, or combination of either or both, particularly management,
able to dominate deliberations or control decisions at any scale?
• What if it could seamlessly blend cooperating and competing, with
all parts free to compete in unique, independent ways, yet could
yield self-interest and cooperate when necessary to the inseparable
good of the whole?
• What if it could be infinitely malleable, yet extremely durable, with
all parts capable of constant, self-generated, modification without
sacrificing its essential nature, thus releasing human ingenuity and
spirit?
VISA—A Novel, Chaordic Organization
As envisioned by Hock, the governance model that emerged from that
meeting, and would become the structure behind VISA, defied previously
well-established tenets of corporate organization. VISA was a nonstock,
for-profit membership corporation with ownership in the form of nontrans-
ferable rights of participation. VISA was highly decentralized and highly
collaborative and functioned, as Hock believed, as “an enabling organiza-
tion” above all else. Similar to a Jeffersonian governmental structure, every-
thing possible (authority, initiative, decision making, wealth) was relegated
to the periphery, with only standards and the most large-scale operational
issues remaining under centralized control. The center-and-periphery model
posed a logical solution to a fundamental problem within the credit card
industry: member financial institutions were all competitors (issuing their
own cards) but, in order to have a sustainable system, needed to all cooper-
ate (regardless of which bank issues cards, all VISA cards must be accepted
by all merchants).
According to Hock, the success of this business model is best demon-
strated by the fact that, although VISA is an enormous corporation, few
know of its organizational structure. However, at the same time, the core of
the enterprise has no knowledge of or authority over a vast number of the
constituent parts. No part knows the whole, the whole does not know all
the parts and none has any need to. The entirety is largely self-regulating.
Core Tenets of Chaordic Organizations
Hock has long insisted that the VISA model cannot be transposed suc-
cessfully onto any other industry. It is the notion of a chaordic governance
model, not anything specific to VISA’s history, that can be used across
institutions. However, inherent in toeing the line between chaos and order,
designing and implementing such a governance model is an organic and dy-
OCR for page 276
276 DIGITAL INFRASTRUCTURE FOR THE LEARNING HEALTH SYSTEM
namic process unique to the institution attempting it. Although the product
will be drastically different in each case, there are certain core tenets of the
chaordic model. Some examples include
• Maximize human ingenuity. Hock argues that the most abundant,
least expensive, most underutilized, and frequently abused resource
in the world was human ingenuity; the source of that abuse was
archaic, Industrial Age institutions and the management practices
they spawned.
• Organizations must have clarity of a shared purpose, common
principles, and strength of belief. According to Hock, organizations
are merely conceptual embodiments of a very old, very basic idea—
the idea of community. An organization’s success has enormously
more to do with clarity of a shared purpose, common principles
and strength of belief in them than to assets, expertise, operating
ability, or management competence.
• Push all possible operations to the periphery. No function should
be performed by any part of the whole that could reasonably be
done by any more peripheral part, and no power vested in any part
that might reasonably be exercised by any lesser part.
• Foster and tolerate evolution. The organization must be adaptable
and responsive to changing conditions, while preserving overall
cohesion and unity of purpose. The governing structure must not
be a chain of command, but rather a framework for dialogue, de-
liberation, and coordination among equals.
