A systematic review of hundreds of deaths among children left in hot cars has found a variety of opportunities to prevent them, with the right safety features.
Conducted by researchers at the Children’s Hospital of Philadelphia (CHP), the study published last month examined 354 cases of hot-car deaths in the U.S. to determine whether a variety of technologies could have detected or intervened in 10 “exemplar” cases among them.
An average of 37 American children die in hot cars each year, the researchers say. There have been more than 1,000 such deaths since 1998, and according to reports gathered by PETA, 82 pets have died in extreme heat conditions in 2025 alone, with another 216 rescued alive.
“We have a responsibility to children around the country for making sure these completely preventable deaths never happen,” study co-author Kristy Arbogast said in a release.
In Canada, an average of one child per year dies in a hot car, according to research from Toronto’s Hospital for Sick Children.
In recent years, some jurisdictions have passed laws requiring safety measures and protocols to prevent hot-car deaths, but as the researchers note, those requirements aren’t always specific, leaving the question of which solutions are most effective.
Detect and prevent
Safety features designed to prevent hot-car deaths come in many forms.
Technologies under the CHP study’s scrutiny ranged from basic alerts and reminders to check for passengers at the end of a trip, to sensors in seats and seatbelts, to detectors that track radar, breathing and heartbeats.
If a passenger is detected, another set of safety tools can trigger, including warnings from the car’s audio-visual system, alerts to emergency services or the driver’s phone, and fail-safes like automatically turning on the air conditioning, opening windows or unlocking doors.
Researchers say the effectiveness of these tools “varied drastically” when reviewed in the context of prior cases of hot-car deaths.
For example, a detector on the child restraints might only activate if the passenger is buckled in, and automatically notifying the car’s registered contact number may not account for cases where a child is deliberately left in the car by a caregiver who doesn’t recognize the danger. Honking the horn may alert passersby to intervene, but would be unhelpful in remote areas without other people nearby.
What’s more, the study authors acknowledge that their analysis was limited to the worst-case incidents that resulted in a death, leaving out what they estimate to be hundreds or thousands of non-fatal “near miss” cases that may never be documented.
While some safety measures appeared capable of detecting or intervening in all 10 exemplar cases, researchers say they found no all-purpose solution to ensure a death is prevented on its own.
“When we considered the variety of scenarios in our study, we found that no single technology would have prevented all cases we identified,” said Jalaj Maheshwari, co-lead author of the study, in a release.
“This tragic situation can happen to anyone, and we feel that multiple technologies working in tandem are the best way to reduce the number of vehicular heatstroke deaths that occur.”
Highlighted among the measures were heartbeat sensors, which were found to be successful at detecting a passenger in all cases, automatic 911 calls, which would trigger a potential rescue in all cases, and emergency air conditioning, which the researchers say could have prevented about 80 per cent of the examined cases.
Co-author Arbogast says that beyond the innovations themselves, more attention is needed among automakers and drivers alike to safeguard against tragic mistakes.
“In addition to technology, we need to continue to make sure caregivers are educated about the real danger posed to children by overheating vehicles, and how quickly that danger can build,” she said.
“We need to ensure these life-saving technologies have the support of the automotive industry, policymakers and a variety of stakeholders who can make a difference.”
