Document Sample
Mason Powered By Docstoc
					 Linking Data from the
Departments of Health and
  Education in Florida
    Craig A. Mason, Ph.D.
     University of Maine
    Developmental Epidemiology
•   Keith G. Scott – U Miami
•   Derek Chapman – Virginia Commonwealth Univ
•   Russell Kirby – U Alabama: Birmingham
•   Craig A. Mason – U Maine
•   Shihfen Tu – U Maine
•   Rick Urbano – Vanderbilt
•   Marygrace Yale – U Miami
• Mid 1970’s
  – K. Scott coordinating developmental research at
    University of Miami-Mailman Center
  – Focus on developmental disabilities
  – Collaborations with both DOE, regarding special
    education issues, serve as a resource
  – Collaborations with DOH, regarding early
    childhood risk factors, particularly psychosocial
• Late 1970’s
  – K. Scott approached by Department of
    Education regarding impact of low birthweight
    on mental retardation
  – Considering large-scale intervention
  – Degree to which low birthweight results in
    higher rate or increase in the number of cases of
    mental retardation
  – Potential impact of intervention aimed at
    reducing low birthweight, on population rates of
    mental retardation
• Concept of community versus individual Risk
  – Attributable fraction and public health
    epidemiological perspective
  – An uncommon risk factor that has a large effect on
    the person who experiences it may have a small
    impact on population rates of a disorder
  – A common risk factor that has a small effect on the
    person who experiences it may have a large impact
    on population rates of a disorder
  – At the time, a new concept within Education and
    traditional child development fields
Community vs. Individual Effects
• Instrumental in creation of a new research
   – Public health framework applied to traditional child
     development, rather than clinical model
   – Estimate prevalence of things for which there were
     previously no solid estimates
   – Autism…
   – Public health surveillance model
• Developmental Epidemiology and Biobehavioral
  Developmental Epidemiology
  and Biobehavioral Informatics
• Begin with hand linkage of records
• 1990 donation of $100,000 HP computer
• 1992 begin annual downloads of
  approximately 2.5 million statewide DOE
• Birth certificates from 1979 (app 200,000
  per year)
  Developmental Epidemiology
  and Biobehavioral Informatics
• Expanded to incorporate additional datasets
  – WIC, Public Defenders/Juvenile, Child Find
• Base for additional projects: Autism
• Bring in additional researchers and
  – Kirby and Tu research in ACC
• Various collaborations with CDC
  – Promoting utility of education data
• Methodological developments
Why Was This Able to Succeed?
• A variety of factors served to help the
  success of this fairly unique program
• Some were simply fortuitous, some were
• Legal, political, pragmatic, as well as
  external issues completely beyond anyone’s
           External Factors
• Florida was ahead of the curve in creating
  statewide, centralized education records
• High-level officials in both DOE and DOH
  were interested in making this happen
• Technological revolution
• University had independent partnerships
  – Already acting as an agent of both departments
• No direct legislative action, but operated
  through existing legislated University
  Centers tied to DOE
  – Legislatively mandated to provide information
    to DOE and various policy makers regarding
    educational issues
  – Similar to MEPRI in Maine
• Activities related to improving educational
  and classroom experience
• Everyone respected each others areas
  – As researchers, our interest was always in
    facilitation and access
  – Methodologists interested in developing strategies
    and protocols
  – Ownership and control can doom linkage projects
• Security
  – Data maintained in secure room in secure
    building on limited access server removed from
    any network connections
• IRB requirements
           Practical Benefits
• Provide additional research support to both
  – Help with various reports, etc.
  – Use as opportunities for theses, dissertations
• Facilitate expanded linkages
  – Courts/public defenders
  – Ideally build on existing partnerships
    Avchen, Scott, and Mason (2001)

• Note dramatically increasing risk with lower birth
  weight, combined with large decrease in numbers
Avchen, Scott, and Mason (2001)
    Yale, Mason, and Scott (2005)
  Smoking                              EMH

                   Low Birth
         1.43%      Weight

• Balance reflects a small interaction between
  smoking and low birth weight
• Significant portion of the total effect passes
  through low birth weight
    Yale, Mason, and Scott (2005)

  Smoking                          SED/EH

                  Low Birth
       1.06%       Weight

• Nearly all impact on community rates is
  through processes other than low birth

Shared By: