Kids at Risk of Lead Poisoning can be Identified Using Machine Learning

Machine learning can help public health officials recognize kids most at risk of lead poisoning, allowing them to focus their limited resources on preventing poisonings rather than remediating homes only after a child suffers elevated blood lead levels, reports a new study. The findings of the study are published in the journal JAMA Network Open. Chicago Department of Public Health (CDPH) has executed an intervention program based on the new machine learning model. Rayid Ghani, Distinguished Career Professor at Carnegie Mellon University's Machine Learning Department and Heinz College of Information Systems and Public Policy, reported that the Chicago hospitals are in the middle of doing the same. Other cities also think of replicating the program to address lead poisoning, which remains a significant environmental health issue in the US.

The study report that their machine learning model is nearly twice as accurate in identifying kids at high risk than earlier, simpler models, and equitably identifies kids regardless of their race or ethnicity.

Lead-based paint in older housing is a common source of lead poisoning. Yet the standard public health method has been to wait until kids are identified with raised lead levels and then adjust their living conditions.

"Remediation could help other kids who will live there, but it doesn't help the child who has already been injured," stated Ghani, who was the lead author of the study. "Prevention is the only means to deal with this problem."

The study also revealed that the machine learning model identified these high-risk kids equitably. That's an issue with the current system, where Hispanic and Black kids are less likely to be tested for blood lead levels compared to white children, Ghani said.

Source-Medindia