Tracking: Success through Partnerships

1
Tracking Success through Partnerships

• Academic Partners for Excellence in Environmental
  Public Health Tracking

Evelyn Talbott, Dr. P.H. University of
Pittsburgh Dan Wartenberg, Ph.D, University of
Medicine and Dentistry of NJ. Lu Ann White,
Ph.D. Tulane University Jennifer Mann,
Ph.D. John Balmes, MD
University of California at Berkeley

2
(No Transcript)
3
Importance of Collaborations
4
University of Pittsburgh
Transfer of asthma ER Data from a Hospital System
to a Local Health Department
Evelyn O. Talbott, Dr.P.H., M.P.H LuAnn L. Brink,
Ph.D.
5
Asthma Expected Achievements

• Innovative, cost-effective surveillance strategy
• Near real-time surveillance
• Manageable amount of information (50 asthma
  cases/day)
• Based upon a diagnosis, not free-text
• Mechanism to collect other information of public
  health importance

6
What Can Be Achieved with Surveillance

• Asthma Trends over time by zipcode, county, CT (
  i.e. coke oven closes, new technology is put in
  place, etc.) provide a baseline from which to
  follow
• To become alerted concerning a putative hazard in
  the environment
• Identify high risk groups for further follow up
  and public health measures

7
(No Transcript)
8
Cooperation

• Data provided to Allegheny County Health
  Department as part of Public Health Surveillance
• Data provided by the University of Pittsburgh
  Medical Center to improve Public Health
• Only de-identified data will be provided to
  investigators

9

• Via site-to-site VPN,
• UPMC will send
• patient name,
• SSN,
• race
• date of visit,
• time of visit,
• address,
• date of birth,
• gender,
• type of insurance,
• Chief complaint
• ICD9 diagnosis code
• disposition
• For all ED visits with a final diagnosis code of
  493.x from Allegheny County hospitals

Matched respiratory data
ACHD will purge all ED data that do not match a
corresponding discharge code of ICD 9 460-520
Irrelevant data
see full talk Wednesday at 400PM
10
iSOVAT
Spatial OLAP Visualization and Analysis Tool
Bambang Parmanto, PhD Ravi Sharma, PhD Evelyn O.
Talbott, Dr.P.H., M.P.H University of
Pittsburgh In collaboration with Cliff
Mitchell, M.D. John Braggio, PhD Maryland State
Dept of Health
11
Interface
12
iSOVAT

• Provides linkage and integration of data sets
  from various sources, including spatial data (for
  example, industrial or mining locations, rivers
  and lakes), health data (e.g., cancer registry),
  and demographic data (e.g., population, age
  structure, income).
• iSOVAT is capable of integrating all these
  complex data sets into a multidimensional
  database that can be viewed easily from multiple
  angles and in visual forms (maps and charts).

13
Potential collaborations

• Collaborations with Maryland DOH to provide stand
  alone package with following capabilities
• Provide socio-demographic information by state,
  county, Zip code, etc as needed
• Show age specific and both crude and age adjusted
  rates (95 CI) of outcomes
• Provide layering of environmental variables and
  health outcomes

14
An Example of Numbers of Inpatient asthma and MI
primary admissions by County
15
Maryland Portal

• Will link health outcome (asthma and MI) with
  environmental (PM2.5), demographic, and
  socio-economic data.

Inpatient counts due to asthma and MI male and
female by County
16
Inpatient counts due to asthma and MI by race by
County
17
Jennifer Mann, Ph.D.John Balmes, MD University
of California, Berkeleyin collaboration with Tim
Tyner, MS and Fresno Unified School District
School-based asthma surveillance
18
School-based asthma surveillance

• School-based classroom survey
• Given to all 7th and 9th graders in Fresno
• Assesses symptom level among diagnosed
• can be used to triage level of and need for
  intervention
• Both prevalence and severity level could be used
  for EPHT
• Identifies possible asthmatics undiagnosed
  who report severe asthma symptoms

19
School-based asthma surveillance

• Project also develops models that are low-cost
  and do not violate school privacy law (FERPA)
• FERPA does not have public health exclusion so
  researchers can not collect or enter data from
  schools
• Use 12th graders in FUSD who are part of a
  Medical Academy to introduce survey and answer
  questions
• After data processing by UCB, students do their
  own data analysis for report to FUSD.
• Mapped 7th grade data at city block level in
  preparation for linkage with modeled pesticide
  use data for EPHT
• FUSD sent out letter giving everyone right to
  refuse the mapping step to be compliant with
  FERPA

20
School-based asthma surveillance

• In CT, survey results compared to nurse-based
  surveillance
• Additional children were identified by the survey
  as asthmatic
• Currently being tested in Massachusetts by
  tracking partners

21
Tulane Center for Applied Environmental Health

• CDC Tracking Conference
• February 25, 2009

22
Health and Air Quality

• Tulane Missouri partnership
• Demonstration project examining statistical
  methods for linking hospital discharge data and
  EPA air quality monitoring data in Missouri.
• Link PM2.5 data with health indices (MI, COPD,
  and Asthma) and develop a predictive model for
  these health indices
• Data sources
• Missouri Hospital Discharge data (2001-2005)
• EPA air monitoring data for PM2.5 and Ozone
  downloaded from AQS Data Mart (2001-2005)

23
Tulane - Missouri

• Missouri facilitated obtaining the hospital data
• Missouri IRB and other state approvals
• Data access and pulling required fields
• Tulane is conducting the analyses including
• Developed algorithm to interpolate missing sample
  values
• Tested methods to assign exposure - distance
  between the health endpoint and air monitor
• The PM2.5 value was adjusted for distance from
  monitor by dividing observed PM2.5 value/distance
  from monitor.
• Methods of analysis Time series analysis,
  Poisson regression analysis and Case Crossover
  analysis using conditional logistic regression.

24
Gateway Project

• Proof of Concept partnership between Florida,
  Missouri, Washington and Tulane
• EPHT demonstration project to utilize PHIN-MS
  nodes to actually network EPHT sites to each
  other and a "central repository
• Goals
• Test the design and mechanics of the EPHTN Node
  Package
• Test the effectiveness of EPHT metadata template
  search tools

25
Gateway Project

• Partners gained experience creating, exchanging,
  and searching Tracking-based metadata
• Created a resource for grantee portal
• Metadata and Data Repositories for grantee data
  not on the National Portal
• Metadata search and data retrieval for
  grantee-housed data
• Created a multi-node PHIN-MS-based network to
  allow bi-directional data sharing between
  multiple partners

26
In collaboration with several states
Washington, Connecticut, Maine, Massachusetts,
New Hampshire, New Jersey, New York
Using Routinely Collected Surveillance Data
Studies of Births and Environmental Exposures
Daniel Wartenberg, PhD. University of Medicine
and Dentistry of New Jersey
27
Overall Strategy

• Adverse Birth Outcomes/Air Pollution as a Model
  for EPHT
• Relevance
• Health People 2010 Births an indication of the
  nations health
• Use routine surveillance (tracking)
  data/methodology
• Develop collaboration among multiple institutions
• Addressing cutting edge issues
• Role of exposure misclassification in
  interpretation of results
• Consideration of alternative proxies for exposure
• Suggest application of approach to other issues
  in environmental epidemiology
• Studies Underway
• 1 6-state collaboration
• 2 In depth assessment of exposure assessment
  issue

28
The Collaborative Project6-State Study of Air
Pollution and Births

• Goal
• Demonstrate issues in multistate collaboration (6
  NE states)
• Address local/regional/cross border concerns
• Process
• Request data separately from each state
• Analyze individually and jointly
• Assess local and regional patterns and impacts
• Use of 6 states data Increases sample size
  (power) and relevance
• Status
• Analysis of 2 states underway
• Additional data in request process

29
Air Pollution, Birth Outcomes and Maternal Change
of Residence

• Goal
• Assess frequency and consequences of mothers
  change in residence during pregnancy
• Process
• Use mobility data recorded on Washington birth
  certificate
• Compare birth outcomes of stable vs. mobile
  mothers with respect to air pollution exposures,
  adjusted for known risk factors
• Results
• Stable mothers show stronger associations
• May be due to exposure misclassification,
  sociodemographics of movers, stress of moving and
  other factors
• (see full talk Thursday at 8 AM Track 2a)

30
Future Goals of Academic Partners

• Overall strategic goal of the academic partners
  as we move forward in implementation
• Development of more real time Surveillance
  capabilities (secure portals) to create more
  effective environmental public health
  surveillance programs
• Provide enhanced software and tools to use and
  interpret these data in a meaningful way
• Carryout sensitivity analysis of secondary data
  sources (eg. Birth certificates )for their use in
  the tracking network.

31
Thank you