Public Health Data
Computational modeling and simulations depend on empirical data for model parameterization and testing. The Vaccine Modeling Initiative is committed to advancing access and use of public health data for modeling and policy making. As part of the VMI, Project Tycho™ aims to change the paradigm of data sharing and use in public health in the US and globally. As separate effort, we are applying innovative incentive based network models to create computer representations of public health data reporting systems for a better understanding of bottlenecks and data quality. Project Tycho™ has four main focus areas of activity to accelerate the availability and use of public health data.
1. Acquisition of new data
We completed the digitization of all US weekly nationally notifiable disease surveillance reports published from 1888 to the present and aim to continuously include additional public health data from around the world.
2. Data infrastucture
We are conducting active research on new algorithms to digitize, standardize, integrate, and store public health data efficiently using combinations of automated processes and manual verification. Rigorous quality control systems will be essential to maintain the highest quality standards.
We are collaborating with international partners from a large variety of scientific disciplines to create innovative analytical approaches to add value to public health data.
Above right: History of US disease surveillance. Each concentric circle represents a decade, starting with 1888 in the center. The type of disease reports changed over time as indicated by different colors: red represents reports of death, and other colors represent case reports for a number of disease subcategories: black: 1, green: 2, blue: 3, and orange: >3.
Many barriers have been identified that currently limit the availability of public health data for research and policy. Project Tycho™ is actively engaged in advocacy for better data availability. Recently, we conducted a literature review to compile all evidence on real and potential barriers to data sharing in public health (submitted).
Public health data reporting systems often experience challenges in timeliness and completeness of reporting. For example the Global Polio Eradication Initiative greatly depends on accurate detection and reporting of suspected polio cases. In a collaboration between the University of Pittsburgh and Carnegie Mellon University, we use incentive based network models to create a computer representation of data reporting systems in public health to identify bottlenecks and data quality constraints. With these models, we aim to develop tools for the assessment of reporting systems and to better understand incentives that enable accurate reporting and high data quality.