2015 APHA Annual Meeting & Expo

Home Events & Meetings Annual Meeting Schedule & Program

Recorded Annual Meeting Presentations (RAMP) are now available for purchase.
Only $90 for presenters, $140 for registered attendees and $200 for non-attendees.

Online Program

335140
Healthier homes with predictive analytics to identify risks of lead poisoning


Tuesday, November 3, 2015 : 2:50 p.m. - 3:10 p.m.

Eric Potash, PhD, Center for Data Science and Public Policy, University of Chicago, Chicago, IL
Rayid Ghani, MS, Center for Data Science and Public Policy, University of Chicago, Chicago, IL
Joe Walsh, PhD, Center for Data Science and Public Policy, University of Chicago, Chicago, IL
Raed Mansour, MS, Chicago Department of Public Health, City of Chicago, Chicago, IL
Jay Bhatt, DO, MPH, MPA, FACP, Chicago Department of Public Health, Chicago, IL
Audio (mp3) recording Recorded Presentation (mp4)
Lead has been banned from most consumer products for over 35 years, but it continues to poison thousands of children in Chicago with large social, economic, health, and educational costs.  We present a model designed to help the Chicago Department of Public Health reduce childhood lead poisoning by estimating the risk that a child will be poisoned or that a dwelling poisons a child.  Using historical blood lead level tests, Cook County assessment records, American Community Survey data, vital statistics, and a linear-kernel support vector machine, the model ranks children and dwellings by risk so inspectors can focus on the most dangerous cases first, potentially improving lives while saving millions of dollars and years of work. This prioritization of inspections and allocation of inspectors are one half of improving prevention measures. The other half incorporates of these risk modelling scores into EMR may improve prevention of lead poisoning in children at the point of care by physicinas by identifying addresses during prenatal care as well as screening tests required before entry into schools. These innovations in data science may occur within the constructs of existing policies.

Learning Areas:

Communication and informatics
Environmental health sciences
Epidemiology
Other professions or practice related to public health
Public health or related research
Systems thinking models (conceptual and theoretical models), applications related to public health

Learning Objectives:
Demonstrate how to improve public health practice in prioritizing lead inspections using risk scores calculated through predictive modeling. Demonstrate how to improve public health practice in allocating lead inspections using risk scores calculated through predictive modeling. Identify effective traditional and non-traditional data sources for predictive modeling used to calculate risk. Explain how risk scores can be integrated within EMR to trigger blood lead testing or lead inspections to prevent lead poisoning. Discuss the use of multidisciplinary collaboration to build capacity and sustainability in the Healthy Homes Program for lead inspections. Articulate the formulation of strategies involving innovations in data science within the constructs of existing policies.

Keyword(s): Data Collection and Surveillance, Environmental Health

Presenting author's disclosure statement:

Qualified on the content I am responsible for because: I am an PhD mathematician working on applications of statistics and machine learning to public policy. For the past year I have collaborated with the Chicago Department of Public of Public Health on predictive analytics for lead poisoning.
Any relevant financial relationships? No

I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines, and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed in my presentation.

Back to: 4321.0: Understanding Aspects of Predictive Analytics in Advancing Public Health