Preliminary Findings BioSense Evaluation Project

1
Preliminary FindingsBioSense Evaluation Project

• Jim Buehler
• October 12, 2007
• ISDS Annual Conference, Indianapolis, IA
• On behalf of
• James Buehler, MD, Ellen Whitney, MPH
• Rollins School of Public Health, Emory University
  
• Alexander Isakov, MD, MPH, Donna Smith, MS
• School of Medicine, Emory University
• Michael Prietula, PhD
• Goizueta School of Business, Emory University

2
BioSense vision
To provide state, local, and nationwide health
situational awareness...before, during, and after
a health event using biosurveillance.
www.cdc.gov/biosense/publichealth.htm
3
Project Objectives

• How do public health departments and healthcare
  providers respond to various emergencies or
  potential emergencies?
• What is role utility of different forms of
  surveillance?
• Establish a method to assess impact of advances
  in surveillance on public health and healthcare
  practice
• Inform ongoing development of syndromic
  surveillance

4
Case Study Interviews

• State local public health officials
• Epidemiologists
• Health officers
• Program managers
• ID control
• Nurses
• PIOs
• Environmentalists
• Emergency managers

• Healthcare
• Infection Control Practitioners (ICPs)
• Hospital epi
• ED ID Physicians
• Emergency managers
• IT
• Interviews recorded transcribed
• Qualitative social network analyses

5
Case Study Interviews

• Course of event respondents role
• Perspective on
• Key landmarks, turning points, decisions
• Utility of various information sources
• Public health Healthcare interactions
• Lessons learned

6
Case Studies

• Georgia
• Salmonella outbreak, 2006
• Tornado, 2007
• Wildfire, 2007
• Influenza, 2006-2007 (inc. local district)
• Florida
• Wildfire, 2007 (part of GA event)
• Hurricane Wilma, 2005 (initial interview only)
• San Diego County
• Influenza, 2006-2007
• HazMat events, 2005-2007
• Environmental biological alert
• Wildfire, 2003

7
Valdosta Salmonella Outbreak, 2006
GA DHR Div of PH
72 cases of Salmonella infection found in Lowndes
County
8
Salmonella OutbreakKey Inputs

• Lab serotype PFGE
• Calls from hospital ICP
• Epi environmental investigations
• Healthcare Lab ? ICP ? Docs
• SS not available
• Syndromic surveillance wouldnt have picked it
  up anyway. it would have been hard to pick it
  upunless the data were broken down by
  serotype.

9
Americus Tornado, 2007
http//www.srh.noaa.gov/
10
Americus Tornado Key Inputs

• Ad hoc queries to ICPs
• Informal links to PH Emergency responders
• EOC Physical virtual
• SS 911 calls
• Monitored extent of injuries possible
  post-event outbreaks
• Without syndromic surveillance I dont think
  that I would have the same level of confidenceto
  be able to say there were no increased incidences
  of disaster related health effectsbecause we
  would have been dependent on the human factor and
  the human factor has limitations SoI would
  have felt comfortable but not at the same exact
  level.

11
Influenza Season, 2006-2007GAState District,
CASan Diego
http//health.state.ga.us/
http//www2.sdcounty.ca.gov/hhsa/
12
InfluenzaKey Inputs

• Calls from hospitals, schools, nursing homes,
  community physicians
• Sentinel networks (increasingly difficult to
  maintain)
• Lab-based monitoring rapid tests
• ED-based SS
• Syndromic surveillance reallywas an added
  enhancementbecause it is more timely and we are
  able to see and pull out those age specific
  groups.
• Healthcare Lab ? ICP ? Docs
• I get information walking down the hall..

13
San Diego HazMat Events, 2005-2007
http//www.sdcounty.ca.gov/deh/hmd/
14
San Diego HazMat Environmental Positives Key
Inputs

• ED-based SS detected or recognized small clusters
  of illness due to chemical exposures
• we came across an increase of chemical
  exposure.. we determined thattwo individuals
  who wereemployees at a waste treatment facility
  had gone to the ERThey hadinadvertently mixed
  some chemicals andgot exposedwe contacted
  HAZMAT HAZMAT backtracked and determined that
  they had failed to notifyCal OSHA
• Close collaboration between PH Env Health/HazMat

15
Preliminary Impressions

• Surveillance occurs in context of relationships
• Public health healthcare providers
• Personal contacts remain essential
• Requires integration of human and technological
  capacities
• Situational awareness is situation dependent
• Redundancy valued

16
When was SS most useful?

• Influenza
• Post disaster
• Disaster consequence monitoring
• Absence of outbreaks
• Injuries
• Impact on people with chronic disease
• Specific HazMat or possible exposure events
• Least useful
• When more detailed information required about
  illness, e.g., serotype
• Use by hospitals underdeveloped

17
Role of BioSense Local Systems

• BioSense Not considered or used in most
  instances
• Real-time hospital not nearby
• DoD VA data not sufficiently timely or
  representative
• Difficult to use and inflexible
• Locally implemented SS systems used first
• Developed in context of local state PH ?
  Healthcare relationships that define surveillance
  practice
• System ownership familiarity
• Exploratory use of BioSense increasing

18
Limitations

• Formal qualitative analyses pending
• Sites events may not be representative
• None of events studied had massive human health
  impacts

19
Next Steps

• Complete interviews and analyses for current
  round of case studies
• Review progress with CDC and 3 other project
  institutions
• Make sure we asking the right questions
• Articulate respondents insights regarding role
  of SS in event situational awareness

20
Acknowledgements

• CDC BioSense Evaluation Cooperative Agreement
• Georgia Division of Public Health
• GA West Central, South, Southeast, East Metro
  Health Districts
• Florida Department of Health
• Duval Co. (Jacksonville, FL) Health Dept.
• County of San Diego Health Human Services
  Agency, Dept of Environmental Health
• City of San Diego EMS
• Hospitals in each of these communities