Content uploaded by Steven Merahn
Author content
All content in this area was uploaded by Steven Merahn on Aug 02, 2019
Content may be subject to copyright.
126 www.greenbranch.com | 800-933-3711
PRACTICE MANAGEMENT
In a 2006 article in the New England Journal of Medicine,
“Educational Strategies to Promote Clinical Reasoning,”
Bowen notes that the organizing principles of medical
knowledge may have unfavorable eects on the capac-
ity to apply that knowledge in problem solving1; she further
notes that such curriculum-driven structure may limit the
development of problem representations that incorporate
the necessary complexity and context to support adequate
clinical solution development.2 ese insights have addi-
tional and important implications for the new competen-
cies required by the emerging operating model for care
delivery now generally referred to as “population health
management.”
POPULATION HEALTH MANAGEMENT
Population health management shifts the focus from caring
for patients who self-select for care to taking on transcendent
responsibility for the health status of a cohort or population
of patients, and has come to be associated with evolutionary
trends in payment or compensation for healthcare services
known variously as “accountable care” or “value-based
care.” e goal of value-based care is to deliver high-level
quality-related performance across a population or cohort
of patients using the lowest levels of resources and services
required to eectively reach those quality goals.3
Under value-based care, professional and provider sys-
tem compensation is based not on fees for specic services
provided to individual patients, but on the “accounting”
of the system-level performance toward population level
goals for clinical quality and resource utilization, and set by
the value (economic and otherwise) associated with those
goals (e.g., reducing predicted primary hospitalization
rates by closing evidence-based gaps in care).
Population management is a form of systems-based
practice and requires implementing and mastering an
operating manual that is separate and distinct from
traditional care delivery,4 including a specialized infra-
structure with its own functional requirements and an
associated set of operating capabilities. Although analyt-
ics are often considered the foundation of population
management, improving both the quality of care and the
quality of health of target populations requires new forms
of interaction design and goal-directed collaboration
within and between the systems—human, information,
and environmental—that make up the patient’s health
resource community.
We have observed a pattern in health systems and phy-
sician networks at various stages of the evolution toward
value-based population health management in which the
traditional organizing principles of care planning limit
the capacity of these organizations to develop problem
representations that are appropriate and relevant to the
population management operating model, subsequently
impeding their ability to envision models of care delivery
that extend beyond traditional models of medical practice.
Knowledge Representation and
Care Planning for Population Health
Management
Steven Merahn, MD*
The traditional organizing principles of medical knowledge may be insufcient to
allow for problem representations that are relevant to solution development in
emerging models of care such as population health management. Operational
classication and central management of clinical and quality objectives and as-
sociated strategies will allow for productive innovation in care design and better
support goal-directed collaboration among patients and their health resource
communities.
KEY WORDS: Care plan; quality measures; population health; accountable care; value-
based care.
*Chief Medical Ofcer, US Medical
Management, 500 Kirts Blvd., Troy, MI
48084; phone: 917-689-8954; e-mail:
smerahn@usmmllc.com.
Copyright © 2015 by
Greenbranch Publishing LLC.
www.greenbranch.com
| 800-933-3711
Merahn | Knowledge Representation and Care Planning 127
CLINICAL STRATEGIES AND
CARE PLANNING
Traditionally, the clinical strategies and associated goals
for any one patient were a responsibility distributed among
all of a patient’s providers. This often resulted in lack of
sharing of resources and knowledge, redundancies in
tests, and, sometimes, adverse events due to conicts in
treatment strategies.5 e movement toward the patient-
centered medical home (PCMH) model resulted from the
desire to create a centralized point of management for the
patient’s diverse care plans in order to coordinate care,
identify potential conicts, reduce redundancies, and im-
prove communication.6
In population health, the care plan is the framework for
establishing specic and measurable objectives for clinical,
quality, operational, and outcomes-related aspects of an
individual’s health. In practice, however, the frameworks
for care planning commonly used under PCMH and popu-
lation health are almost exclusively dened by condition,
professional discipline, or patterns of utilization, restricting
the capacity of organizations to consider alternative care
delivery models.
is issue has historical roots. In 1983, Dr. Robert Gor-
don, then Special Assistant to the Director of the National
Institutes of Health, wrote a brief but breakthrough report
critiquing the traditional approach of classifying chronic
illness prevention strategies based on “origins of disease”
and proposing a new framework for “operational classica-
tion” of patient-focused clinical strategies.7
Gordon sought to dene prevention strategies by their
predictable outcomes, targeted to those “for whom the
measure is advisable on a cost-benet basis.” Although his
approach generated some interest within the substance
abuse community,8 these prescient recommendations
were not widely adopted by or socialized in mainstream
medicine.
The core of the Gordon classification describes three
categories of prevention:
Universal measures, recommended for essentially
everyone;
Selective measures, advisable for population subgroups
distinguished by age, sex, occupation, or other evident
characteristics; and
Indicated measures—those that should be applied
only in the presence of a demonstrable condition that
identies the individual as being at higher than average
risk for the future development of a related or comorbid
disease (or event).
In the context of current evolutionary trends in health-
care delivery, the Gordon classication could be consid-
ered the foundation of an approach to transforming clinical
objectives into a rational operational classication scheme,
and is presented here as a universal framework for care
planning in population health management.
THE GORDON CARE PLAN
FRAMEWORK
We congured the Gordon classication into a care plan
framework reecting a spectrum of operational and clinical
goals associated with population management initiatives,
expanded to include wellness/lifestyle objectives at one
end, and prognosis-based compassionate care at the other
(Table 1).
In its practical application, the revised Gordon classi-
cation care plan framework (Gordon Care Plan) recatego-
rizes and aggregates the judgments and recommendations
of all the human and information systems that participate
in a patient’s health resource community. This creates
a continuously updated central repository— the single
“source of truth”— and performance driver shared across
all members of a patient’s health resource community.
Organizations can then determine the most ecient meth-
ods to achieve progress toward goal; responsibility for goal
achievement can be assigned or distributed among the
patient’s resource community.
Given the population-level obligations of health systems
under value-based care, the potential objectives of any
single care plan will fall into one of two categories: system-
driven objectives and person-driven objectives.
System-Driven Objectives
System-driven objectives are those elements of account-
ability between the provider system and those contracting
for care; most system-driven objectives align with clinical/
quality needs of patients but may include operational or
utilization-related objectives as well as qualitative mea-
sures associated with patient experience management
(e.g., physician communication skills, patient satisfaction).
In population management, the provider system analyzes
and straties the population to determine which members
meet criteria for one or more system-driven measures, and
those measures then become part of those patients’ care
plans. Whenever possible, system-driven objectives should
accrue to the benet of the patient. If there is a potential
conict between a system-driven objective (e.g., reducing
hospital readmission) and a patient’s needs, the patient’s
needs should overshadow the system-driven objective.
Person-Level Objectives
Person-level objectives are those elements of care or health
status improvement that may not be part of the value-based
contractual obligations of the provider, but are either neces-
sary or aspirational for the patient. Necessary person-level
objectives may be either correlated to system-driven objec-
tives or simply called for by standards, evidence, or best
128 Medical Practice Management | September/October 2015
www.greenbranch.com
| 800-933-3711
practices. Aspirational objectives often are health status,
functional, or outcomes-related goals for the patient that
may be unrelated to any quantiable criteria.
COLLABORATION AS KEYSTONE TO
POPULATION MANAGEMENT
Shared decision-making is increasingly considered a best
practice, and a measure of quality for value-based care.9
Consolidating the patient’s open issues into a single op-
erating framework via the Gordon Care Plan can simplify
and facilitate patient involvement in their action plan for
health and wellness, increasing the likelihood of high-level
patient engagement and involvement in the achievement
of care plan goals.10 is includes better understanding of
the conceptual foundation of goals being set by profession-
als, providing a personal context to the plan via person-
level goals, priorities, and preferences, and establishing a
mutual foundation for their relationship with, and perfor-
mance by, their health resource community.
Along with shared decision-making, the Gordon Care
Plan framework supports five critical success factors for
population health management (Table 2).
REMAPPING QUALITY MEASURES
AS CARE PLAN GOALS
Table 3 illustrates a practical application of the Gordon
care plan framework, showing a mapping of the 33 current
Accountable Care Organization (ACO) quality measures
for the Medicare Shared Savings Program11 (administered
by the Centers for Medicare & Medicaid Service) against
the Gordon care plan framework (excluding person-level
objectives).
Table 1. A Gordon Classication-Based Care Plan Framework
Clinical
Strategy
Wellness Universal
prevention
Selected
prevention
Indicated
prevention
Condition
management
Compassionate
care
Strategic
Target
Individual focus Community-
based
Characteristic-
based
Condition or risk-
based
Diagnosis or
event-based
Prognosis-based
Strategic
Framework
Targets whole
population to
achieve or sustain
levels of physical,
mental and social
well-being
Targets whole
population
(nation, local
community) and
aims to prevent
or delay universal
health risks or
conditions
Targets groups
or individuals
whose risk is
above average;
subgroups may
be distinguished
by traits such
as age, gender,
family history,
geography
Targets groups
or individuals
with an existing
condition or other
data or identiers
indicating
condition-related
risks
Targets groups
or individuals
with conrmed
diagnosis or
other condition
that is unstable,
whose quality of
condition-related
care does not
meet evidence-
based standards
or whose health
status is poor
Patients with any
serious illness who
have physical,
functional,
psychological, or
spiritual distress
as a result of their
conditions and/
or associated
treatments
Examples
of Care
Plan Goal
or Action
Improve
nutritional
decision-making
Flu vaccine;
seat belt usage;
weight screening;
dental hygiene
Mammograms
in women over
50; apnea screen
when BMI>30 +
snoring
Reduce CV risk
in DM (add
ACE inhibitor/
ARB); reduce
osteoporosis risk
with COPD Rx;
readmission risk
reduction; cancer
survivorship
Blood sugar
management in
DM; medication
management in
CHF; surgery for
appendicitis
Palliative care
hospice
ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; BMI, body mass index; CHF, congestive heart failure; COPD, chronic
obstructive pulmonary disease; CV, cardiovascular; DM, diabetes mellitus; Rx, prescription.
Table 2. The Gordon Care Plan Framework Supports Critical
Success Factors for Population Health Management
Clinical
integration
Serves as a central repository and universal
categorization scheme across all specialties
and disciplines and information systems
Knowledge
sharing
Goals inherently contain information about
patient’s condition and health status, serving
as a curated, goal-oriented view of the
patient’s entire data set
Real-time
exibility
As new data elements become available,
goals may be resolved, or new conditions may
be identied, requiring additions to the plan
Collaboration Represents a set of common goals across
the patient’s health resource community and
supports collective determination—shared
responsibility, authority, and accountability—
for achieving results
Resource-
managed care
delivery
Creates new opportunities to envision
alternative paths to goal achievement through
communications tools, community xtures, or
nonclinical personnel and drive innovation via
its application as a universal metadata solution
for care management resources
www.greenbranch.com
| 800-933-3711
Merahn | Knowledge Representation and Care Planning 129
Operationally, centralized care planning is a functional
requirement in the population health operating model,
so using the Gordon care plan framework is a matter of
configuration with minimal disruption. Current models
of evidence-based practice and order set development
provide the content basis of categorizing clinical goals and
determining the “standing orders” for the population under
management.12 Although there may be potential categoriza-
tion conflicts, the framework itself provides the rules for
resolution: weight management may be an independent
wellness goal for one patient, but the presence of related
conditions may make it a trigger for indicated prevention
in another. Although the Gordon Care Plan can be imple-
mented manually, the use of information technologies to
create and manage one is easily accomplished by congura-
tion of current electronic health record systems, analytics,
clinical decision support, and care management platforms.
CONCLUSION
Foreseeing Bowen’s 21st-century eorts to align learning
design with desired capacity for problem representation,
Bruner’s seminal 1960 work e Process of Education dis-
cussed the importance of structure as part of a learning
process he referred to as the “transfer of principle,” which
permits the “recognition of subsequent problems as special
cases of an idea originally mastered.”13
“The teaching and learning of structure, rather than
simply the mastery of facts and techniques, is at the cen-
ter of the classic problem of transfer . . . If earlier learning
is to render later learning easier, it must do so by pro-
viding a general picture in terms of which the relations
between things encountered earlier and later are made
as clear as possible.”13
As Bowen and subsequent authors have pointed
out, the structure of knowledge is essential for problem
Table 3. 2015 Centers for Medicare and Medicaid Services Medicare Shared Savings ACO Measures
Mapped Against Gordon Care Plan Framework
Wellness/
Lifestyle
Management
Universal
Prevention
Selected
Prevention
Indicated
Prevention
Condition
Management
Compassionate
Care
Individual Focus Community-
Based
Characteristic-
Based
Condition or
Result-Based
Diagnosis or
Event-Based
Prognosis-Based
EXAMPLE: 2014 ACO Medicare Shared Savings Program Quality Measures
System-Driven
Objectives:
Clinical*
Health
Promotion/
Education (5)
Health Status/
Functional Status
(7)
Flu Vaccination
(14)
Adult Weight
Screening (16)
Adult Tobacco
Use Assessment
(17)
Adult Depression
Screening (18)
Adult Blood
Pressure
Screening (21)
Falls Risk
Screening (13)
Pneumococcal
Vaccine (15)
Colorectal Cancer
Screening (19)
Mammography
Screening (20)
Documentation
of current
medications in
medical record
(39)
All Condition
Readmissions (8)
COPD/Asthma
Admissions (9)
CHF Admissions
(10)
CAD+DM/CHF:
% of patients
on ACE or ARB
therapy (33)
Skilled nursing
facility 30-day
all-cause readmis-
sions (35)
All-cause
unplanned
admissions for
patients with:
• diabetes (36)
• heart failure (37)
• multiple chronic
conditions (38)
DM: Eye exam
(41)
DM: HgbA1c<8%
(22)
DM: % of
patients
HgA1c>9% (27)
HTN: % of
patients
BP>140/90 (28)
IVD: % of
patients on
ASA or other
antithrombotic
(30)
LVSD: % of
patients on beta-
blocker therapy
(31)
CAD: % of
patients on Rx
to lower LDL
cholesterol (32)
Depression
Remission at 12
months (40)
Data from Centers for Medicare & Medicaid Services Quality Measure Benchmarks for the 2015 Reporting Year (Release Notes/Summary
Changes February 2015); (X) = CMS ACO Measure Number.
*Excludes patient experience measures 1, 2, 3, 4, 34.
ACE, angiotensin-converting enzyme; ACO, Accountable Care Organization; ARB, angiotensin receptor blocker; ASA, acetylsalicylic acid (aspirin);
BP, blood pressure; CAD, coronary artery disease; CHF, congestive heart failure; COPD, chronic obstructive pulmonary disease; DM, diabetes
mellitus; HgbA1c, glycated hemoglobin; HTN, hypertension; IVD, ischemic vascular disease; LDL, low-density lipoprotein; LVSD, left ventricular
systolic dysfunction; Rx, prescription.
130 Medical Practice Management | September/October 2015
www.greenbranch.com
| 800-933-3711
representation and its subsequent application to clini-
cal reasoning and problem solving.14 Eorts to rene the
structure of medical knowledge will be critically important
as the healthcare community seeks to imagine and design
new models of care delivery that may break from tradi-
tional discipline-driven models.
Although clinical analytics can provide risk stratifica-
tion, predictive modeling, utilization management, and
identification of evidence-based gaps in care across the
population, the true keystone of successful population
management is the essential triad of care planning, care de-
livery, and collaboration, because that is where patients are
engaged in productive change toward health and wellness.
e structure of the Gordon Care Plan provides an opera-
tional framework for collaborative, resource-managed care
design by the people, platforms, programs, and services that
make up the patient’s health resource community.
Y
REFERENCES
1. Bowen JL. Educational strategies to promote clinical diagnostic rea-
soning. N Engl J Med. 2006;355:2217-2225.
2. Chang RW, Bordage G, Connell KJ. e importance of early problem
representation during case presentations. Acad Med. 1988;73(Suppl):
S109-S111.
3. Porter ME. What is value in health care? N Engl J Med. 2010;363:
2477-2481.
4. Mostashari F, Sanghavi D, McClellan M. Health reform and physician-
led accountable care: the paradox of primary care physician leader-
ship. JAMA. 2014;311:1855-1856.
5. Schoen C, Osborn, R, Squires D, etal. New 2011 survey of patients
with complex care needs in eleven countries nds that care is often
poorly coordinated. Health A (Millwood). 2011;30:2437-2448.
6. Jackson GL, Powers BJ, Chatterjee R, etal. e patient-centered medi-
cal home: a systematic review. Ann Intern Med. 2013;158:169-178.
7. Gordon R. An operational classication of disease prevention. Public
Health Rep. 1983;98):107–109.
8. Mrazek PJ, Haggerty RJ. Reducing Risks for Mental Disorders: Frontiers
for Preventive Intervention Research. Washington, DC: National Acad-
emy Press; 1994
9. Barry MJ, Edgman-Levitan S. Shared decision making—the pinnacle
of patient-centered care. N Engl J Med. 2012;366:780-781.
10. Hibbard JH, Greene J. What the evidence shows about patient activa-
tion: better health outcomes and care experiences; fewer data on
costs. Health A (Millwood). 2013;32:207-214.
11. Medicare Shared Savings Program Quality Measure Benchmarks for
the 2014 and 2015 Reporting Years. Washington, DC: Centers for Medi-
care & Medicaid Services. 2013; www.cms.gov/Medicare/Medicare-
Fee-for-Service-Payment/sharedsavingsprogram/Downloads/MSSP-
QM-Benchmarks.pdf . Accessed June 18, 2014.
12. Ballard DJ, Ogola G, Fleming NS, etal. e impact of standardized
order sets on quality and nancial outcomes. In: Henriksen K, Battles
JB, Keyes MA, etal., editors. Advances in Patient Safety: New Direc-
tions and Alternative Approaches. (Vol. 2: Culture and Redesign).
Rockville, MD: Agency for Healthcare Research and Quality; 2008.
13. Bruner J. e Process of Education. Cambridge, MA: Harvard Univer-
sity Press, 1960.
14. Mutnick A, Barone M. Assessing and remediating clinical reasoning.
In: Kalet A, Chou CL, eds. Remediation in Medical Education. New
York: Springer; 2014; 85-101.
