Managing food chain risks: the role of uncertainty

Richard Shepherd

University of Surrey

Project partners

Person responsible Organisation

Richard Shepherd University of Surrey

Andy Hart Central Science Laboratory

Gary Barker Institute of Food Research

Simon French Manchester Business School

John Maule Leeds University Business School

Uncertainty

• There are known knowns

– there are things that we know that we know

• There are known unknowns

– that is to say, there are things that we now know we don't know

• But there are also unknown unknowns

– there are things we do not know we don't know

• And each year we discover a few more of those unknown unknowns

Donald Rumsfeld, US Secretary of Defense

Winner of Plain English Campaign ‘Foot in Mouth Award’ 2003

Cynefin model of decision contexts

Knowable

Cause and effect can be determined with sufficient data

The realm of scientific inquiry

Known

Cause and effect understood and predictable

The realm of scientific knowledge

Complex

Cause and effect may be explained after the event.Social

systems

Chaotic

Cause and effect not discernable

Snowden (2002)

Need to communicate uncertainty

• Need for:– Transparency– Openness

• If uncertainty subsequently found it will lead to problems of credibility

• ‘… the need to be open about uncertainty and to make the level of uncertainty clear when communicating with the public’

HM Government Response to the BSE Inquiry (2001)

Presenting uncertainty to the public

• Admission of uncertainty (Johnson and Slovic, 1995)– More honest– Less competent

• Public preferences (Frewer et al. 2002)– Public want information on uncertainty– More accepting of uncertainty when due to

scientific process than lack of interest or action by government

Project objectives

• Develop interactive web-enabled tools for quantitative assessment of risks and uncertainty

• Use participatory methods to ensure web-enabled tools, etc. appropriate for stakeholders

• Develop methods to predict consumer behaviour driven by perceptions of risk and uncertainty

• Develop improved methods for communicating with stakeholders

• Test, evaluate and demonstrate improved approaches in case studies of food contamination and microbiological hazards

Food chain 1

Food chain 2

Food chain 3

Modular food chain models

Each module includes production, processing, storage, retail, cooking etc.

Dietary risk modelling

National diet survey data

2D Monte Carlo engine

Post-processing & analysis

Long-term extrapolation

Participatory Processes

‘Live’ groups• Stakeholder

workshops• citizen juries• focus groups

Web enabled interactions

Surveys

All stages of process: from problem definition to

interpretation, decision-making, communication

Predicting changes in consumer behaviour

Analyse effects of communication and management actions

Predict dietary changes

Case studies

Achemical

contaminant

Amicrobial

contamination

Acrisis

scenario

Specialist user

Decision-making forums

Lay public and stakeholders

Communication and decision support interfaces

Results for technical reports

Test alternate scenarios and management options

Lay user:What if?

Risk to me?

Media

direct via internet

indirect

Models, systems and processes designed and validated with respect to

Modules within the project

Participatory processes

• Participatory methods– Stakeholder workshops – Citizens’ juries– Focus groups– Scenarios to stimulate discussion

• Runs throughout project to:– Inform initial developments and ensure processes and

web-enabled tools appropriate for stakeholders– Test in case studies

Dietary risk modelling

• Develop web-based tools based on CSL probabilistic tool• Probabilistic methods of risk assessment:

– Take account of variability and uncertainty– Usually aimed at specialists

• Hierarchical 2D Monte Carlo to quantify uncertainties• Expand to include:

– Other contaminants and pathogens– Long term exposures– Suitable for non-technical users– ‘What if’ tools

Modular food chain models

• Managing risk across the food chain

• Modularisation of food chain– Production, processing, storage, consumption

• Dependencies across the chain - concentrations of agent a function of: – Control measures– Performance criteria

• Build set of uncertainty distributions

Predicting changes in consumer behaviour

• Impact on consumer behaviour of communication and management actions

• Issues addressed:– Risk information v direct recommendation– Personal relevance of information– Presentation of uncertainty– Numerical/verbal presentation of uncertainty

• Predict consumer behaviour changes

Communication and decision support interfaces

• Communication dependent on how different actors understand and think about risk

• Mental models

• Social representations

• Test using scenarios

Case studies

• Chemical - pesticide– Data available – Amenable to probabilistic modelling

• Microbiological - cross contamination with campylobacter– Undercooked chicken– Mainly caused by cross contamination

• Scenario with unanticipated risk– Hypothetical scenarios– Rapid response

Key audiences

• Natural and social scientists• Stakeholders throughout the food chain

– Producers– Manufacturers– Risk managers and regulators

• General public(s)• Communication through:

– Dissemination activities– Stakeholder workshops

Concluding comments

• Interdisciplinary research– Natural sciences– Social sciences

• Quantitative assessment and modelling of risks and uncertainty across the food chain

• Stakeholder involvement and participatory processes

• Effective communication with the public and stakeholders