Methods development & practices

One of the missions for the Academic Partners for Excellence for EPHT is to support the national program with a focus on methods development.  Examples of methods that are needed to build the EPHT network are:

• Methods to make environmental data more useable by health officials such as making misaligned data sets linkable
• Mapping techniques which preserve the confidentiality of the data
• Statistical methods for spatial and temporal analysis

The Tulane APEX addresses these needs through several objectives designed to develop and/or adapt methods for implementing the EPHT network in state and local health departments.  This in turn will lead to tools and best practices that can be used by health agencies for environmental public health tracking.

    1.   Evaluate current health, exposure and environmental data collection and surveillance   
          (tracking) methods and develop best practices that increase efficiency, and decrease time
          and cost for data collection and management.

Examine and characterize prototypical examples of environmental and health databases and surveillance systems and describe uses, benefits, limitations and barriers to its use. Under this task, we propose to conduct an evaluation of datasets and surveillance systems relevant to asthma and air monitoring (and mercury in fish) available from collaborating partners, to identify methodological issues that impede inclusion of these data in an integrated tracking system, and to make recommendations for improving the efficiency and quality of data incorporated into air quality and asthma tracking systems.  This will include an evaluation of potential alternative data sources (traditional and non-traditional) that may supplement or replace data that is frequently used for these purposes.  The principles of creating normalized databases for data collection and management will be incorporated in best practices that state and local health departments can use to collect and manage high quality data on environmental hazards, exposure and health outcomes. 

    2.   Evaluate and adapt existing methods and develop new methods as necessary to  
          facilitate effective and efficient data linkage in Environmental Public Health Tracking.

The ability to meaningfully integrate existing, yet disparate, data sources is a cornerstone of EPHT.  Such integration requires standard methods for efficiently and effectively linking records from one source to records in another source.  Whether this is done for individuals or for aggregated data, certain issues must be considered in linking the data: accuracy of the linkage, security of protected health information and maintaining confidentiality, ability to apply appropriate analytic techniques, onetime vs. ongoing linkage, ability to include data regarding control variables, and temporal as well as spatial linkage issues.  For aggregated data, an additional issue to consider is the resolution at which the data are collected, stored, and ultimately linked and analyzed. This objective uses the data compiled in objective 1 to examine the issues of data quality and timeliness, integration of health and environmental data, utility of surveys and other information, use of protected health information, data completeness, integration of information regarding covariates and confounders, as well as the development of standards for EPHT or compliance with existing standards. 

    3.   Apply, adapt, and evaluate statistical algorithms that will be used by state and local 
          EPHT partners to analyze trends, and detect patterns and relationships among hazard, 
          exposure and health effects in populations.

The focus of this objective is to apply, adapt, develop, and evaluate algorithms and analytic methods that can be incorporated into state and local health and environmental agencies for trend analysis and aberration detection in asthma exacerbations related to air quality.  Methods to be analyzed range from simple mapping techniques to complex multi-level modeling. 

Methods for trend analysis and aberration detection will be enumerated and reviewed for appropriateness in analyzing asthma and air quality data sets and their ease of application in the state health departments.  Automated solutions for applying these methods will be developed for commonly used database and statistical software (SAS, SPSS, MS Access) and evaluated for accuracy and ease of implementation in state health departments. The capacity of state health departments to conduct, interpret and utilize results from these analytic techniques will be evaluated.  Furthermore, the need for tools to foster the use of these methods in state health departments should be elucidated.