Policies with Agent-based Models Douglas A. Luke June 26, 2017 - - PowerPoint PPT Presentation
Studying Implementation of Evidence-based Policies with Agent-based Models Douglas A. Luke June 26, 2017 Academy Health Annual Research Meeting Enhancing Implementation Science: Applying Systems Models to Address Complexity Goals Present
Present rationale for systems
Argue for systems science
Demonstrate how agent-based
Complex Systems – Daniel Ferreira-Leites Ciccarino
Poverty
Gun-violence
Climate change
Obesity
Tobacco control
Healthcare access
Implementing evidence-based practices in health settings
Many sectors/actors
Problem embedded across multiple biological, social, organizational levels
Incomplete knowledge
High economic/political stakes
Interconnectivity with other problems
Solution unclear or undefined
Complex systems are:
System behavior:
Challenge Description
Researchers prefer to examine individual policies in isolation, but the reality is that new policies are added to a thick, complex ‘policy soup’ (Kingdon).
Policy effects can be most accurately measured if exactly the same policy is implemented exactly the same way across multiple contexts—this is never the case in reality. Instead, policies tend to be tailored (adapted) to meet local needs.
Policy changes always lead to push back from various constituencies, organizations, commercial entities, etc. Sometimes researchers call these ‘unintended consequences.’
Inspired by Glass & McAtee, 2006, SSM
From Luke & Stamatakis, 2012, ARPH
A bottom-up simulation approach that is used to study
Emphasizes
Similar to microsimulations
NetLogo
Explore some NetLogo models
https://duncanjg.wordpress.com/2012/09/24/a-simple- meta-population-model-in-netlogo/
Separation-avoid crowding neighbors
Alignment-steer towards average heading of neighbors
Cohesion-steer towards average position of neighbors
https://www.youtube.com/watch?v=ozLacy8t3gw
Longest history of ABMs in public health is in the modeling of
Large-scale models (often using synthetic populations of entire nations or
Used by policymakers, federal governments, industry
Examples
http://www.epimodels.org/ http://fred.publichealth.pitt.edu/ https://www.youtube.com/watch?v=ECJ2DdPhMxI https://mattbierbaum.github.io/zombies-usa/
Agent Properties Agent Actions Agent Rules Time Environment
Hammond, R. (2015) IOM Report
R21 CA172938 – NCI U01 CA154281 - NCI (With Ross Hammond, Brookings Institution; Kurt Ribisl, UNC; Lisa Henriksen, Stanford)
Decrease availability Increase search cost of obtaining Decreases visibility of
Changes social norms, reduces
Reduces “Tobacco Swamps”
From Luke, et al, 2011, Am J Prev Med
Licensing
Proximity to schools
Retailer proximity
Retailer type (e.g., pharmacies)
Major Model Components Age Adults Town Type 1) Urban Rich; 2) Urban Poor; 3) Suburban Rich; 4) Suburban Poor Smoker Type Light Smoker vs. Heavy Smoker Transport Mode Walk, Bike, Car Actions
the other half at home in the evening Rule
rational, two-phase, and learning Time 1 day with 2 periods (morning - work & evening - home) Environment
Warehouse, Tobacco), Workplace, School, Population
age distribution, land area covered (10 square miles) Outcome
From Luke, et al. (2017). American Journal of Public Health
Different policies designed to reduce retailer density
Context-dependency of policies, important for
‘Layering’ of policies may be more effective than
Poor Urban Rich Urban Poor Suburban Rich Suburban Retailer cap
++ ++
Store type
++ +
School buffer
++ +
Proximity buffer
++ +
Multiple policies
++ ++ ++ ++
Not just for epidemics!
Can study policy interactions, and how policy implementation plays out over time
Interdisciplinary, use and influence empirical work
See IOM Report for evaluation framework and guidance
Exciting work in tobacco control, obesity, substance use, violence, among others
theory
incompatible with available data
vice versa
From Epstein, J. M., 2008, JASS, Why Model?
From Epstein, 2008; Why Model? http://www.santafe.edu/media/workingpapers/08-09-040.pdf