TOP 37 - Analysis of near-miss pediatric vehicular heatstroke data in the US

Publication Number: 2788.TOP 37

Alexis H. Kinzer, University of Alabama School of Medicine, Hoover, AL, United States; Emma Sartin, University of Alabama at Birmingham, Birmingham, AL, United States

Background: On average, there is 1 death attributed to pediatric vehicular heatstroke (PVH; i.e. children overheating in vehicles) every 10 days in the US (Safe Kids, 2023). Importantly, there is no known information about non-fatal PVH events in the US (i.e., “near-miss cases”). These near-miss cases are valuable to study as they may give us insight into who is at higher risk for potential PVH and where action can occur from a public health standpoint to prevent cases of PVH in the future. They also likely happen much more frequently than fatal cases; one Australian nonprofit reported more than 5,000 children are rescued from cars annually.

Objective: Our objective of is to analyze 2 novel sources of near-miss cases: 1) a dataset created and maintained by the nonprofit Kids and Cars Safety (KCS), and 2) a dataset comprised of 911 calls from Shelby County, Alabama (AL). Our goal is to validate and analyze legitimate cases to characterize near-miss PVH events, their outcomes, frequency, and how they differ or relate to the currently published PVH fatality data.

Design/Methods: KCS Dataset: KCS created this dataset by collecting news stories, self-reports, police reports, and social media posts. We are currently validating 2015-2023 data against “source” or evidence for the event. Validation will be complete by December 2025, and analyses of legitimate cases will be complete by January 2026.

Data Source #2: Shelby County 911 is a consolidated 911 center that provides processing/dispatch services for law enforcement and fire/EMS agencies. The center began tracking instances of occupants unattended or locked in vehicles in 2019; as of the end of 2024, they have logged >600 cases.
Both datasets contain information about the context of the event (e.g., child gained access independently, child has autism, caregiver locked keys in car with child), information about the type of vehicle involved (e.g., color, make, model), the type of responder dispatched/on scene (fire, law enforcement), location of event (e.g., drugstore, residental living), and the duration of event from the initial 911 call to the resolution. For our analyses, cases must be an event where a child (under 18 years old) was left in a car unsupervised (for any amount of time) and did not die from their injuries. For both datasets, we will use historical data to incorporate outdoor (if not reported in notes) temperature information. We will conduct descriptive analyses to identify preliminary trends in non-fatal PVH cases (e.g., seasonality, demographics, location, etc.).