ISDC 2002 - Palermo, Italy

Jack HomerHomer Consulting

Voorhees, New Jersey, USA

Bobby MilsteinCenters for Disease Control and Prevention (CDC)

Atlanta, Georgia, USA

Communities with Multiple Afflictions:A System Dynamics Approach

to the Study and Prevention of Syndemics

AbstractThe term syndemic was coined by the medical anthropologist Merrill

Singer to describe the mutually reinforcing nature of health-related problems such

as substance abuse, violence, and AIDS, that disproportionately impact inner city

neighborhoods burdened by economic hardship, deteriorated infrastructure, social

disruption, malnutrition, and inadequate health care.

Public health scholars agree that effective responses to the intertwined

afflictions within communities throughout the world require comprehensive,

system-wide interventions. To strengthen the scientific foundation for such

initiatives, the CDC is supporting research on syndemics, including the use of

system dynamics modeling to investigate how and why syndemics develop, and to

evaluate alternative approaches to intervention for particular community contexts.

At present, a generic (not yet case-specific) model has been developed

based on the literature and expert observations. This preliminary model is available

as a web-based game at: http://broadcast.forio.com/sims/syndemic. We present the

model and some results, and outline plans for carrying the work forward.

A Syndemic is...

The spread and persistence of mutually reinforcing health-related problems such as substance abuse, violence, and AIDS, typically found in inner city neighborhoods burdened by economic hardship,

deteriorated infrastructure, social disruption, malnutrition, and inadequate health care.

Singer M. 1994. AIDS and the health crisis of the US urban poor: The perspective of critical medical anthropology. Social Science and Medicine 39(7): 931-948.

Singer M. 1996. A dose of drugs, a touch of violence, a case of AIDS: Conceptualizing the SAVA syndemic. Free Inquiry in Creative Sociology 24(2): 99-110.

A Web of Afflictions

General Community Conditions

Ties

Afflictions

Public Health Goals

• Preventing disease and injury

• Prolonging life

• Reducing overall burden of illness

• Avoiding activity limitation

• Maintaining emotional balance

• Eliminating health disparities

• Enhancing life satisfaction

Standard Practice Falls Short• Public health scholars agree: Effective responses

to the health problems of the urban poor and other communities require system-wide interventions.

• However, most public health agencies continue to act as if each affliction can be prevented individually by understanding its unique causes and developing narrowly targeted interventions.

• This compartmentalized approach is engrained in the agencies’ financial structures, scientific frameworks, and statistical models.

A Change in Perspective

A syndemic public health orientation:

• Places multiple afflictions in context, and identifies systematic links among them

• Assesses the influence of community conditions

• Assesses the capacity of community organizations to direct health and social policy change

• Brings together the sciences of epidemiology and system dynamics with the action agenda of community leaders

Core Public Health Functions Under a Syndemic Orientation

System Dynamics

SocialNavigation

POLICYDEVELOPMENT

ASSESSMENT

ASSURANCE

NetworkAnalysis

SyndemicOrientation

Afflictionprevalence

Generalcommunityconditions

Out-migration ofnon-afflicted

Affliction contagionand cross-impacts

Investments incommunity

Affliction-related death

Community effortsto fight affliction

Community effortsto improve general

conditions

Outside assistanceto build capacity

R

R

R

R

R

B

B

B

Communitycapacity

Outside assistanceto fight affliction

Outside assistance toimprove general

conditions

Availablecommunitycapacity

R

Disruption due tooutside assistance

R

Model Overview

Blue arrow: positive linkGreen arrow: negative linkRed text: decision leverCircled R: reinforcing loopCircled B: balancing loop

Population Stocks & Flows:3 Affliction Types (A, B, C)

Popnnonaf f licted

Popn A0

Popn B0

Popn C0

Popn AB0

Popn BC0

Popn AC0

Popn ABC

Net f low 0 to B0

Net f low 0 to A0

Net f low 0 to C0

Net f low B0 to BC0

Net f low A0 to AB0

Net f low A0 to AC0

Net f low AB0 to ABC

Net f low BC0 toABC

Net f low AC0 toABC

Net f low C0 to AC0

Net f low B0 to AB0

Net f low C0 to BC0

Death A0

Death C0

Death B0

Death AB0

Death AC0

Death BC0

Births

Death ABC

Death 0

Net outmigration

Incidence & Recovery Logic

“Popn A0”: afflicted with A, but not B or C“Popn AB0”: afflicted with A and B, but not CMagenta text: constant<Green text>: variable defined elsewhere in model

Popn A0 Popn AB0

Incidence Bfrom A0

Recovery Bfrom AB0

Incidence rate Bfrom 0

<Generalcommunityconditions>

Fraction of popnat risk for B

<Community effortagainst B>

Effect of having Aon incidence B

<Prevalence of B>

Noncontagiousincidence rate B

Contagion rate B

Effect of having Aon recovery B

Incidence rate Bfrom A0

Recovery rate Bfrom AB0

Recovery rate Bfrom B0

Community effortrelative to

prevalence BNet flow A0 to AB0

Websim Introductory Screen

Three mutually compounding afflictions—let’s call them A, B, and C—have just been introduced to a community. The general community conditions are only mediocre and not supportive of healthy living, so the threat of a growing "syndemic" is quite real. Your goal is to minimize the community’s burden of affliction over a twenty year period.

The community is already making an effort to improve general conditions, but its internal capacity to do so is limited, and more could be done with backing from government and philanthropies. Government agencies and foundations could also step in with programs to help fight the afflictions directly by developing policies and services that reduce rates of incidence and boost rates of recovery. A third type of assistance would be leadership training and organizational development to build up the community’s internal capacity for action of all sorts.

The government and philanthropic organizations have sufficient resources to fund all three types of assistance, but only for a limited number of years. You must decide (1) when to initiate each type of assistance, (2) how widely the community will be involved in externally-funded programs, and (3) how heavily each of the affliction types will be weighted in the community’s allocation of effort.

http://broadcast.forio.com/sims/syndemic/

A Severe Syndemic

For ease of presentation, in all simulation runs, the three afflictions are assigned identical parameter values for incidence and recovery, incl. both self-contagion and cross-impact.

In this run, after 20 years, 59% of the population have at least one of the three afflictions: 24% with one of them, 18% with two, and 17% with all three.

Base conditions=.50, No outside assistance

12,000 people1 index

6,000 people0.5 index

0 people0 index

0 2 4 6 8 10 12 14 16 18 20Years elapsed since first incidence of afflictions

Popn total : Base50 peoplePopn afflicted : Base50 peopleCommunity effort against affliction : Base50 indexGeneral community conditions : Base50 indexCommunity capacity : Base50 index

A Milder Syndemic

In this run, improved general conditions of .70 are sustained by a higher assumed natural rate of community investment.

After 20 years, 21% of the population have at least one of the three afflictions: 16% with one of them, 4% with two, and only 1% with all three.

Base conditions=.70, No outside assistance

12,000 people1 index

6,000 people0.5 index

0 people0 index

0 2 4 6 8 10 12 14 16 18 20Years elapsed since first incidence of afflictions

Popn total : Base70 peoplePopn afflicted : Base70 peopleCommunity effort against affliction : Base70 indexGeneral community conditions : Base70 indexCommunity capacity : Base70 index

Focusing on a Single Affliction

In this run, all community effort (still without any outside assistance) goes to fighting affliction A, in the hope that such focus may prevent the spread of all three mutually reinforcing afflictions. (In the base run, all three afflictions were given equal weight.)

The strategy does not work. It does reduce the prevalence of A, but allows B and C to grow further, resulting in an even greater fraction of people with at least one affliction.

All weight on fighting A, No outside assistance

0.8

0.6

0.4

0.2

0

0 2 4 6 8 10 12 14 16 18 20Years elapsed since first incidence of afflictions

Affliction prevalence : AllWeightA fractionAffliction prevalence : Base50 fractionA prevalence : AllWeightA fractionA prevalence : Base50 fraction

In this and all subsequent runs, baseline community conditions = .50 (as in ‘Severe Syndemic’.)

Outside Assistance - Affliction

In this run, outside assistance to fight the afflictions is provided for the first 10 years.

The strategy works to slow the spread of affliction, but only for as long as the assistance is provided. The assistance disrupts the community’s normal political balance, thereby hurting its capacity to act unassisted. Ten years after the end of assistance, affliction prevalence is back to where it was in the base run.

Assistance to fight affliction for 10 years

0.8 fraction1 index

0.4 fraction0.5 index

0 fraction0 index

0 2 4 6 8 10 12 14 16 18 20Years elapsed since first incidence of afflictions

Affliction prevalence : AssistF10 fractionAffliction prevalence : Base50 fractionCommunity effort against affliction : AssistF10 indexCommunity effort against affliction : Base50 indexCommunity capacity : AssistF10 indexCommunity capacity : Base50 index

Outside Assistance - Capacity

In this run, outside assistance to build community capacity is provided for the first 10 years.

The strategy gradually helps the community to fight affliction more effectively, even after the assistance comes to an end. Greater capacity also leads to better general community conditions, which improve from .50 to .56 by the end of the run.(not shown).

Assistance to build capacity for 10 years

0.8 fraction1 index

0.4 fraction0.5 index

0 fraction0 index

0 2 4 6 8 10 12 14 16 18 20Years elapsed since first incidence of afflictions

Affliction prevalence : AssistC10 fractionAffliction prevalence : Base50 fractionCommunity effort against affliction : AssistC10 indexCommunity effort against affliction : Base50 indexCommunity capacity : AssistC10 indexCommunity capacity : Base50 index

Outside Assistance - Conditions

In this run, outside assistance to improve community conditions is provided for the first 10 years. It is implemented in a way that is inclusive and does not disrupt the community’s political balance. This avoids undermining capacity, but somewhat diminishes the direct effectiveness of the assistance.

The strategy gradually helps the community to fight affliction somewhat more effectively, even after the assistance comes to an end. General conditions improve from .50 to .54 by the end of the run (not shown).

Assistance to improve general conditions for 10 years

0.8 fraction1 index

0.4 fraction0.5 index

0 fraction0 index

0 2 4 6 8 10 12 14 16 18 20Years elapsed since first incidence of afflictions

Affliction prevalence : AssistGI1 fractionAffliction prevalence : Base50 fractionCommunity effort against affliction : AssistGI1 indexCommunity effort against affliction : Base50 indexCommunity capacity : AssistGI1 indexCommunity capacity : Base50 index

Outside Assistance - Combined

In this run, assistance is provided for the first 10 years to (1) build capacity, (2) fight affliction, and (3) improve general conditions. #2 and #3 are implemented in a way that does not disrupt the community’s political balance. This avoids undermining #1, but somewhat diminishes the direct effectiveness of #2 and #3.

This strategy is the best available, given the community’s poor initial conditions and the temporary nature of the assistance. General community conditions improve from .50 to .59 by the end of the run (not shown).

Combined (non-disruptive) assistance for 10 years

0.8 fraction1 index

0.4 fraction0.5 index

0 fraction0 index

0 2 4 6 8 10 12 14 16 18 20Years elapsed since first incidence of afflictions

Affliction prevalence : AssistCGFI1 fractionAffliction prevalence : Base50 fractionCommunity effort against affliction : AssistCGFI1 indexCommunity effort against affliction : Base50 indexC ommunity capacity : AssistCGFI1 indexCommunity capacity : Base50 index

http://www.cdc.gov/syndemics

Syndemics

Prevention Network

Web-based simulator & model background:http://broadcast.forio.com/sims/syndemic/

Syndemics Prevention Network: U.S. Members 2001-2

International

• Argentina

• Australia

• Canada

• Grenada