Title: Using System Dynamics (SD) Methodology for Strategic Planning in VA QUERI Programs
1Using System Dynamics (SD) Methodology for
Strategic Planning in VA QUERI Programs
- David B. Matchar, MD
- Kristen Hassmiller Lich, PhD
- Jack Homer, PhD
- For the Stroke QUERI
2Identify problem
Collect data
Evaluate alternatives
Select solutions
Implement
3Difficulties with standard approaches
- Challenges to effective, sustainable translation
of research into action in the real world (our
QUERI mission!) - Limited resources. funding does not cover
development and evaluation of policies and
clinical interventions. Furthermore, mistakes in
strategic direction are costly. - Numerous policy options. It is difficult to
develop a single strategic plan from the large
and diverse evidence on stroke. - Multiple stakeholders, multiple visions. When
dealing with complex problems, stakeholders often
operate from conventional and often narrowly
focused wisdom about how to improve systems of
care that all limit their ability to see new ways
of operating. - Absence of a forum for integration. Multiple
stakeholders are key to successful and
sustainable implementation. There is a lack of
existing linking structures in which key
participants can come together to make change
happen.
4(No Transcript)
5Workshop
- (IA) An overview of SD methodology and (IB) its
application in a current Stroke QUERI - (II) Discussion of strategic planning problems in
other QUERIs that may be addressed using the SD
approach and - (III) Consider ways the SD approach may be
utilized broadly in the QUERI program.
6(IA) AN OVERVIEW OF SD METHODOLOGY
7Brief Background on System Dynamics Modeling
- Compartmental models resting on a general theory
of how systems change (or resist change) often
in ways we dont expect - Developed for corporate policies in the 1950s,
and applied to health policies since the 1970s - Concerned with understanding dynamic complexity
- Accumulation (stocks and flows)
- Feedback (balancing and reinforcing loops)
- Used primarily to craft far-sighted, but
empirically based, strategies - Anticipate real-world delays and resistance
- Identify high leverage interventions
- Modelers engage stakeholders through interactive
workshops
Forrester JW. Industrial Dynamics. Cambridge,
MA MIT Press 1961. Sterman JD. Business
Dynamics Systems Thinking and Modeling for a
Complex World. Boston, MA Irwin/McGraw-Hill
2000.
8System Dynamics Health Applications1970s to the
Present
- Disease epidemiology
- Cardiovascular, diabetes, obesity, HIV/AIDS,
cervical cancer, chlamydia, dengue fever,
drug-resistant infections - Substance abuse epidemiology
- Heroin, cocaine, tobacco
- Health care patient flows
- Acute care, long-term care
- Health care capacity and delivery
- Managed care, dental care, mental health care,
disaster preparedness, community health programs - Health system economics
- Interactions of providers, payers, patients, and
investors
Homer J, Hirsch G. System dynamics modeling for
public health Background and opportunities.
American Journal of Public Health
200696(3)452-458.
9Model Uses and Audiences
- Set Better Goals (Planners Evaluators)
- Identify what is likely and what is possible
- Estimate intervention impact time profiles
- Evaluate resource needs for meeting goals
- Support Better Action (Policymakers)
- Explore ways of combining policies for better
results - Evaluate cost-effectiveness over extended time
periods - Increase policymakers motivation to act
differently - Develop Better Theory and Estimates (Researchers)
- Integrate and reconcile diverse data sources
- Identify causal mechanisms driving system
behavior - Improve estimates of hard-to-measure or hidden
variables
10Community CV control model
11Simulations for Learning in Dynamic Systems
Multi-stakeholder Dialogue
Morecroft JDW, Sterman J. Modeling for learning
organizations. Portland, OR Productivity Press,
2000. Sterman JD. Business dynamics systems
thinking and modeling for a complex world.
Boston, MA Irwin McGraw-Hill, 2000.
12Our key challenge dynamic complexity
- System complexity
- A moving target
13Dynamic complexity arises because systems are
- Dynamic
- Tightly coupled
- Governed by feedback
- Nonlinear
- History dependent
- Self organizing
- Adaptive
- Evolving
14Effective models of complex systems
- Causal (not correlational)
- Dynamic (not equilibrium)
- Grounded in empirical tests (econometrics,
ethnography) - Broad boundaries (not limited to one disciplinary
domain)
Engage stakeholders who develop ownership
15(IB) USE OF SD FOR STROKE QUERI STRATEGIC PLANNING
16Our approach
- This project uses System Dynamics (SD) modeling
to help key stakeholders of the Stroke QUERI
achieve a comprehensive understanding of the
complex systems involved in stroke prevention and
treatment and provides a tool to support
effective stakeholder communication and the
establishment of strategic actionable priorities.
17The process
- Met with key system stakeholders represented on
the Stroke QUERI Executive Committee - established a shared conceptual framework of the
continuum of stroke in the VA - Identified key classes of interventions under
consideration - After several iterations of feedback by
stakeholders, the framework was transformed into
a stock and flow simulation model.
18Conceptual framework
19Stock and flow simulation
20Technical issues
- Simulates veteran enrollees between 2008 and 2028
- Separating enrollees into mutually exclusive
states (stocks) based on - Event (TIA or stroke, with post-stroke enrollees
separated by modified Rankin score) - High or low risk group (gt or lt one risk factor
smoking, DM, HTN, AF). - Outcome variables defined by stakeholders
(events, DALYs, cost) - Model parameters based on VA data when possible
alternatively, scientific literature and expert
opinion. - Programmed using Vensim software (www.vensim.com)
- CAVEAT The current model is preliminary,
intended to provide a credible foundation for
further improvement, working in close
collaboration with a larger group of system
stakeholders and content experts.
21Key constants
22Scenario variable definitions
- Community awareness Probability of
appropriately responding to stroke sxs. - Fraction of non-event population at higher risk
Fraction of VA population without prior TIA or
stroke with a current modifiable risk factor. - Quality of first event (TIA, stroke) prevention
Intensity of efforts to target individuals who
have not had a TIA or stroke but have risk
factors, with medications and lifestyle change,
that could prevent TIA or stroke or other
cardiovascular event. (When Quality 1,
prevention interventions are used when
appropriate and to their optimal effect.) - Quality of TPA use Fraction of enrollees
experiencing an ischemic stroke who are eligible
for and receive tPA correctly in the acute care
setting. - Quality of recurrent event (TIA, stroke)
prevention Intensity of post-acute efforts that
would prevent a recurrent stroke or TIA (e.g.
carotid endarterectomy, discharge planning) and
diligence/adherence of long-term caregivers. - Quality of office MD response to non-hospitalized
TIA For patients with TIAs who do not go to
hospital, additional intensity of outpatient
response. - Quality of stroke rehabilitation Quality of
rehabilitation efforts for first 90 days after
stroke intended to prevent permanent loss of
functioning.
23(No Transcript)
24(No Transcript)
25Relative Risk
Absolute risks for No event, lower risk ( per
thousand population per year) TIA 3.47 Stroke
3.71 Non-stroke death 23.601.
26Illustrative flow calculation
Stroke rate for high risk with prevention is
stroke rate for high risk absent prevention x
(1-quality quality x (1 ability)) So, if
quality 1, then Stroke rate for high risk
with prevention is stroke rate for high risk
absent prevention x (1 ability) Note, 1 -
ability stroke rate with prevention/stroke rate
absent prevention RR
27Preliminary results
28Next Steps
- Work with a larger group of system stakeholders
and stroke experts to refine and validate model
assumptions and parameter estimates. - Analyze the model to identify leverage points
for interventions that have the greatest
potential to improve stroke care. - Use the model as a flight simulator to try out
various policy scenarios in an interactive
workshop with system stakeholders. - Based on insights/discussion, create an action
plan that is feasible (e.g., through leveraging
existing resources), and sustainable (e.g., by
accounting for barriers and undesirable ripple
effects of interventions).
29Our objective
- To achieve a humane, effective, and sustainable
health care system - In our lifetime
16 years
30II and III Implications for QUERI
- Does your QUERI have
- a handle on
- The size of the relevant population (current and
projected)? - Range of plausible policy and clinical actions?
- Potential impact of these actions?
- A strategic plan based on the above?
- Stakeholder/decision maker participation/buy-in/co
mmitment to action? - Is this relevant to the QUERI broadly?