Transcript MedInfo 2004 Poster - Columbia University

Reasoning for Public Health Emergency Risk Communication
Jacqueline Merrill, RN, MPH 1, Suzanne Bakken, DNSc 1,2 , Vimla Patel, PhD 2
1School of Nursing and 2Department of Biomedical Informatics, Columbia University, New York, NY
SCOPE OF THE PROBLEM
RESULTS
DISCUSSION
• Public health risk communication in emergencies and urgent
events is often sub optimal because established risk
communication theory and principles can be overlooked
PRELIMINARY ANALYSIS
• The routine steps of public health risk message development
are affected by the nature of emergencies and urgent events
(illustrated in model below)
• Message development steps were applied in different order
• Informatics methods can supply decision support tools to aid
development of effective theory-based public health
emergency risk messages
• A foundational step to developing such support is clear
understanding of the message development process
Concept Model--Public Health
RISK
• Subjects demonstrated sound conceptual knowledge of risk
communication principles and theory
than recommended by experts, and steps skipped
• Subjects used message templates and heuristics
• Message development relied strongly on collaboration and
conformed to model of group decision making under stress
Characteristics of Naturalistic
Decision Making (Klein, et al., 1993)
Characteristics of Risk Communication
in PH Emergencies/Urgent, High
Concern Events
ill- structured problems
incomplete information, missing facts
uncertain dynamic environment
uncertain dynamic environment
shifting ill - defined goals
evolving nature of urgent events
action feedback loops
agency, public and partner feedback
time stress
need for immediate information
high stakes
high event visibility
Risk Communication in Emergencies
ASSESSMENT
CHARACTERIZATION
COMMUNICATION
Public Health Culture
• Collaborative
High Concern Event
MESSAGE DEVELOPMENT
• Time Sensitive
• Factual, Deliberate
• Incomplete Knowledge
• Process Oriented
• Urgent need for info
• Target Population Focus
• Broad Population Focus
multiple players
multiple stakeholders & levels of
government
organizational goals/norms
consultative, collaborative PH model
Decision support
needed here
OBJECTIVES
• Identify theory-based principles and expert
recommendations for emergency public health risk
communication
• Analysis focused on sequence of activities and steps
involved
• Goals are not ill defined (although problem may be)
• Process phases (or themes) and sub-themes were identified
• Process phases were cyclical or collaborative, sometimes
both
• Points of redundancy were message heuristic, standardized
template and a strong collaborative approach
• Identify features of emergency risk message development in
high concern situations
• Examine how practice compares with theory-based
recommendations
Phases of emergency message development
Phase (Themes)
Sub-themes
1. Info receiving
Comments
Cyclical
Collaborative--other
counties, epi staff, SHD &
CDC
• Identify process steps in emergency message development
METHODS
2. Info seeking
Cyclical
State HD and CDC
• Think aloud protocol conducted with 3 experienced public
health information officers
3. Meeting
Who is in charge
What’ s the message
Who speaks
Collaborative
Cyclical
• Subjects asked to describe actions performed while
preparing public health risk communication to address a
paper based emergency scenario
4. Message content
What’ s the message
Who speaks
Cyclical
Use heuristics
Collaboration
• Preliminary analysis coded responses to published
recommendations of emergency risk communication experts
• Secondary analysis used open and axial coding to reveal
massage process steps
• Factors associated with sub optimal group decision making
in the Naturalistic Model include self censorship and
undermined confidence
• Sub optimal decisions are likely to contribute to sub-optimal
messages
• Example: all subjects skipped the message evaluation
step recommended by experts, a step intended to
expose message weaknesses
• Secondary analysis revealed a structured process with well
defined goals that could be automated
•
• Decision support needed, such as partially automated text
authoring that will test message completeness and
conformance to risk communication theory
• Decision support for public health emergency risk message
development must address institutional culture and process
evaluation
FUTURE STUDY
• Use rapid ethnography to explore contribution of public
health organizational culture to message development
process during actual urgent event
SECONDARY ANALYSIS
MESSAGE DELIVERY
• Naturalistic Decision Making Model conflicts with deliberate,
consultative style of non-emergency decision making typical
used in US public health agencies.
5. Standardize
communication
What’ s the message
Who speaks
Message template guides
6. Give out info
Who’s in charge
Who speaks
Means of distribution:
website, radio, newspaper,
etc.
•Model the workflow of message authors to develop
algorithms to guide semi-automation of authoring process
• Discourse analysis of existing risk messages to identify
typical message structures and components of effective
messages
•Incorporate health communication theory into stages of
automated message construction
REFERENCES
[1] Association of State and Territorial health Officials.
Communication in Risk Situations. Wash. DC: Author, 2002
[2] Reynolds, B. Crisis and Emergency Risk Communication.
Atlanta: Centers for Disease Control and Prevention, 2002.
[3] Klein, GA, Orasanu, J, Calderwood, R. and Zsambock, CD.
Decision making in action: models and methods. Norwood, CT:
Ablex, 1993.
[4] Hinds, P.S., Vogel, R.J., Clarke-Steffen, L. The possibilities
and pitfalls of doing a secondary analysis of a qualitative data
set, Qualitative Health Research, 1997, vol. 7(3): 408-24.
ACKNOWLEDGMENTS
This work was supported by a National Institute of Nursing
Research Training Grant, T32 NR007969. The authors thank
Susan Ledlie, RN, DNSc for consultation on secondary data
analysis