Working Paper: Using Data Mining Techniques to Derive Actionable Decisions in Public Health
Using data from The County Health Rankings and Roadmaps Program (Rankings) and a classification/tree analysis (Exhaustive Chi-squared Automatic Interaction Detection (CHAID)) we identify the relevant predictors of key health outcomes at the county level.
The results from classification/tree analyses can be used to design programs and initiatives that can help a county improve its health outcomes using data that is readily accessible such as the Rankings. The aim of our analyses is to implement a straightforward method that may help organizations (county/city officials, community-based organizations) to better design and calibrate actionable and tractable policies to improve population health.
Book Chapter: Padron NA. El buen Gobierno Sanitario. Madrid, Spain: Springer Healthcare Communications Ibérica SL 2009 Nov; ISBN(chapter): 978-84-937496-1-3.
Report: Sepulveda M, Padron NA, Pagan J & Sbodio M. Health in Cities – Use of Multisectoral Data to Improve Public Health Infant Mortality in Los Angeles and Philadelphia
This report covers our experience setting out to investigate how would a multisectoral data approach to policy design could be used in the context of a complex public health challenge: Infant Mortality. Our experience exploring the organizational and technological landscape in two large cities confirms our hypothesis that cities have become the locus of governance accountability, action and innovation. As demographic, labor and market conditions evolve, the need to embed reliable data in policy design is ever more urgent. Are cities ready?
For our exercise we worked closely with city officials, academic experts, nonprofit and community leaders in the cities of Los Angeles and Philadelphia. However, we believe the lessons learned from this work will be useful not only to Philadelphia and Los Angeles, but also to other cities interested in using multi-sectoral data and new analytical tools to address population health challenges.
Our project brought together researchers from IBM-Research (Marco Luca Sbodio) and The New York Academy of Medicine, and was made possible through generous funding from the Robert Wood Johnson Foundation.
Besa C, Lewis S, Pandharipande PV, Chhatwal J, Kamath A, Cooper N, Knight-Greenfield A, Babb JS, Boffetta P, Padron N, Sirlin CB, Taouli B. Abdom Radiol (NY). 2016 Jul 23. [Epub ahead of print. PMID:27448609