In the United States, about 185 people die every day from a drug overdose. In this case, the medical examiner registers the death as accidental (unintentional) or intentional (suicide).
It is important to know whether death was accidental or not so that we can develop better strategies to prevent suicide. But often officials cannot be sure that death was intended.
A person could intentionally overdose without telling anyone about this intention or leaving a note, which makes it difficult to determine whether they intend to commit suicide. If a medical examiner cannot be certain that death was intended, he is classified as "undetermined".
In the United States, district official deaths are recorded, and the standards for recording deaths vary from district to district.
In many cases, the only way to determine whether a death was intended is to do a "psychological autopsy": family, friends, co-workers, and others are interviewed to find out the deceased's mental state. Unfortunately, this method is very expensive and time consuming.
But thanks to the youthful scientist Daphne Liu from Utah and a team of other scientists, we may be able to come closer to some solutions.
We met Daphne for the first time when she and her scientific partner Mia Yu won first place in the NIDA Addiction Science Award at the 2018 International Science and Technology Fair. Her brother had lost a friend through overdosing and felt this idea right for them.
Daphne and Mia developed software that teaches computer systems how to review details of overdose deaths and how likely it is that death was intended. This is a form of artificial intelligence called "machine learning".
Dr. Paul Nestadt of Johns Hopkins University in Baltimore read about the award and contacted Daphne's science teacher to suggest that she work with his team and refine their machine learning concept to improve its accuracy.
Daphne, now a junior in high school, says the scientists will continue to modify their computer algorithm to assess fatal overdoses from other countries. They hope to develop machine learning systems that can make psychological autopsies more cost-effective, time-consuming and more accurate.
Now Daphne's work could make a difference across the country. By explaining more exactly how and why someone died, doctors and scientists can better help people who have drug problems – before it's too late.
More information: How to get help when you or a friend feel suicidal.