Embed Size (px)
Citation preview
Global status of the evaluation of obesity prevention
Boyd Swinburn
Professor of Population Nutrition and Global Health
University of Auckland
Alfred Deakin Professor
Deakin University
Obesity prevention evaluation symposiumAdelaide, December 2015
Overview
• Global scan
• Full prevention house
• 3 generations of intervention/evaluation– Package testing
– Capacity building
– Integrated systems
• What would SA need to shift into systems interventions and evaluations?
• Conclusions
Obesity prevention: systems-oriented and at scale
• Australia
– Healthy Together Victoria, WHO-CC & South Coast
• New Zealand
– Healthy Families NZ, Healthy Auckland Together
• UK
– Public Health England – 4 pilot studies (RFP)
• US
– TX CORD, Somerville
– In forward thinking
The Full Prevention House
• Leadership & governance
• Information & intelligence
• Finances & resources
• Networks & partnerships
• Workforce development
• Health in all policies
Service
delivery,
programs,
policies
High Level Policies
Political commitment
Specific actions
– people and
food & PA envs
System & capacity building blocks
Systems dynamics
Apply arrows everywhere
Systems change
Capacity change
Predisposition change
Environment change
Behaviour change
BMI change
G1 – package testing
G2 – community capacity buildingG3 – integrated
systems
Moving interventions upstream: 3 generations of approaches
G1-3 characteristicsG1 – Package testing G2 – Capacity Building G3 – Integrated Systems
Researchquestion
Does this intervention package work?
Does capacity building work?
Do systems approaches work? (or ‘how’?)
What is the intervention?
Package of individual & env interventions
Building community capacity
Activating / re-orienting existing systems
Who develops it? Content experts Local & content experts Local experts
Role of community
Consulted & support implementation
Engaged, co-design & implement
Design & implement
Role of experts Create & implement Co-design, support implementation
Support design & ?implementation
Resources applied
Directly to interventions
Community organisations to deliver interventions
To support change agents to leverage systems
Evaluation methods
RCT, cluster RCT Cluster RCT, quasi-experimental
Quasi-exp, systems methods, SNA, CQI
Fidelity Package implementation
Process and relationships Process and relationships
Validity Internal (delivery),weak external
Internal (processes), strong external
Internal (processes), very strong external
Application to ‘at scale’ action
Marginally relevant Proof of principle Directly applicable
Barwon-South Western region
Geelong
Regional population ~300,000
Geelong population ~200,000
3 Whole-of-community intervention programs• 3 years duration• preschool, primary
school, secondary school
• Community capacity building approach
Colac
Logic model for interventions
1 Intervention dose is either 1 or 0 (intervention, control) or $$ (economic input – all schools)2 Capacity is leadership, skills/knowledge, structures, resources3 Relevant environments are schools, homes, neighbourhoods, churches4 Weight, BMI, BMI-z, waist, waist:height, %fat, prevalence of o/w+obesity
InterventionDose1
Δ Community capacity2
Δ Environments3
Δ Knowledge, attitudes, beliefs, perceptions etc
Δ Behaviours Δ Anthropometry4
INPUTS
POPULATION MEDIATORS
INDIVIDUAL MEDIATORS
OUTCOMES
MODERATORS
Ethnicity, socio-cultural factors, gender, age, SES
Δ QoL
Δ QALYs gained
Δ Policy
= Measured = Modelled
(Geelong) <5s
2004-‘08
1.8% (2y/o) & 2.7 %(3.5y/o)
over 3 y
$100k for 12,000 children
Δ behaviours and environments
Δ state prevalence(de Silva-Sanigorski Am J Clin Nutr 2010)
(Colac) 4-12y
2002-’06
~1kg, 3cm waist over 3y
Greater effect in lower SES
children
No Δ ‘safety measures’
Sustained & ?spreading
influence(Sanigorski et al Int J Obesity 2008)
(E Geelong) 13-18y
2004-‘08
5.8 % prevalence over 3 y
Δ community capacity
Δ in school environments
No Δ behaviours (Millar et al Obes Rev 2011)
Investment during & after a 3y intervention program in Colac (vs comparison region)
Swinburn et al Ped Obesity 2014
Changes in overweight & obesity prevalence
Swinburn et al Ped Obesity 2014
Pre-systems thinking approaches
• Interventions were systems building blocks
• Organisational argy-bargy may have been an important sign of systems change
• Quasi-experimental designs with standard epidemiological tools were used
• Low cost interventions eg policies, training
• Problem of measuring behavioural change
• ‘Obesity prevention virus’ spreading along networks
• Cultural barriers to change - ?need systems lens
Systems approaches
• At community level
– $$ investment: rapid capacity & stimulates action
– External $$ can be pulled (eg NPAPH) – ‘project-like’
– ‘Boot-strap’ with existing community resources
– ‘Coalition of the willing’ – mandate, capacity, interest
• At state level
– ‘Backbone’ for coordination, training, creating platforms for interaction, supporting networks and leaders, communications, monitoring & evaluation
– Tools to support a community through a process of activating a cascade of systems change
– Supportive state-level policies, leadership, advocacy
Deakin University CRICOS Provider Code: 00113B
GROUP MODEL BUILDING
15
• Uses system dynamics to develop a causal map/diagram
• Community driven participatory research — core modelling team
• Involves community members in the process of building model (STICK-E)
→ Ownership, collaboration, insight, foundation for action plans, monitor progress
© Steven Allender, Deakin University 2015
Deakin University CRICOS Provider Code: 00113B© Steven Allender, Deakin University 2015
Connection circle for ‘What is making
children fat in our community?’
Deakin University CRICOS Provider Code: 00113B© Steven Allender, Deakin University 2015
Deakin University CRICOS Provider Code: 00113B© Steven Allender, Deakin University 2015
Deakin University CRICOS Provider Code: 00113B
Access to processedfood/quick food
Level of physicalactivity
Parents who cook
Overeating
Ability/time tocook mealsKids raised by people
other their parents
Mechanization/advancesin technology
Precious parenting
Consumption ofsoft drinks
Marketing ofprocessed foods to
kids
Screen time
Advertising/sponsorshipof fast and processed
food
Fear and riskaverse society
Social connection
Access to buyfresh produce
Participation insports
Culture of foodas reward
Car use
Family income
Eating to feelgood
Fundraising with junkfood/fast foodsponsorships
+
+
Perceived waterquality
-
-
Junk foodconsumption
Health literacy
Influence ofadvertsing
-
+
+-
Consumption ofwater
-+
+
+
+
Clubs funds
-
+
+
+
Clubs' healthpromoting policies
-
-
+
+
+
+
+
Preparation forbad weather
+
Available time
-
+
+
+
Price of softdrinks
-
Need to purchasebottled water
-
+
Exposure to fast foodadvertising in sports
+
Fast foodconsumption
Kids asking forfast food
+
Parentspurchasing fast
food
++
+
Influence of parentson children's PA
Parents' value of PAfor children
+
+
Parents' value ofhealthy eating
+
Influence of otheradults on children's PA
+
+
Other adults value ofPA for children
+
Other adults value ofhealthy eating
-+
+Healthy food
options marketed
+
+
<Healthy foodoptions marketed>
-
-
+
<Ability/time tocook meals>
-
Junk food eatingaddiction
+
+
-
<Fear and riskaverse society>
-
Kids bike riding,walking, etc.
+
+
Participationfees
+
-
-
+
+
+
+
-
<Health literacy>
<Mechanization/advancesin technology>
<Marketing ofprocessed foods to
kids>
+
-
<Car use>
+
-
-
Time parents havewith children
Time other adultshave with children
+
+
Influence of parents onchildren's healthy eating
behaviours
Influence of other adultson children's healthy
eating behaviours
-
-
<Time other adultshave with children>
<Time parents havewith children>
+
+
Schools withhealthy curriculum
-
Normalizinghealthy culture
+
Healthy eating+
Normalizing PAculture
+
+
Healthy policies(regional/state level)
+
+ Demand forfruit/vegies
+
+
Distance togrocery stores
-
Breast feedingrates
Kids with healthyweight
-
+
Kids using hunger ascues for eating
+
Sugar consumption
+
Feeling satisfied
-
+
-
Value offast/processed food
for the money
+
Available incomefor food
+
+
Cost of fresh foods-
Quality of freshfoods
Buying fresh food
+
+
Shift work
-
Trading hours ofsupermarkets
-
Work fatigue
+
+
Quality of openspace
+
Access to recreationinfrastructure "parks
and paths"
+
Design of town forsupporting active transport
"ie cycling overpass"
+
Communityconnectedness+
Self esteem
Positive bodyimage
+
+
+
++
Involving kids inplanning
Parents availableto volunteer
+
Pick-up drop offculture
-
-
Available incomefor sports
+
-
++
Stores stockingfresh produce
++
+
Kids wantingmore food
-
-
+
+
Cooking classes
+
<Parents purchasingfast food>
-
+
-
Attractiveness ofparks and open
spaces+
+
+
+
Portion sizenorm
+
<Family income>+
+
+
+
+
+
-
<Shift work>
-
<Work fatigue>
+
-
+
Adults buyingjunk/processed foods
-
+
+
<Mechanization/advancesin technology>
+
<Available time>
-
+
<Kids using hunger ascues for eating>
-
Water taste
+
<Advertising/sponsorshipof fast and processed
food>
+
+
+
<Fast foodconsumption>
-
<Value offast/processed food
for the money>
+
+
-
<Fear and riskaverse society>
+
+
<Available time>+
+
Community Response to Obesity Map: Portland, VIC (Version 3) | 17 July 2014
Fast food & junk food
Participation in sport
General Physical Activity
Social influences
© Steven Allender, Deakin University 2015
Critical importance of monitoring systems (progress, benchmarking, intervention)
• Childhood obesity– MCH / pre-school
– Need for school-based monitoring
– WITH OPT-OUT CONSENT! ([email protected])
• Key behaviours
• Food and PA environments
• Government policy implementation
• Private sector activities
England’s National Child Measurement Program
Deakin University CRICOS Provider Code: 00113B© Steven Allender, Deakin University 2015
Community A
Community B
Community CCommunity D
Community E
Community F
Heat map of local risk patterns
NZ Food-EPI
• Positives: international standard in 6
• Stronger infrastructure than specific policies
• Major gaps in implementation
– Marketing to children
– Fiscal policies
– Comprehensive plans & funding
Conclusions
• Time is right to transition to sustainable, at-scale, systems-oriented approaches to obesity prevention
• Communities – bootstrap, coalition of the willing (mandate, capacity, interest)
• States/territories – backbone functions, political leadership, tools
• Development of systems tools
• Importance of monitoring systems
