Dr Isaac Galatzer-Levy Explains the Importance of Accurately Predicting PTSD After a Traumatic Event
Machine learning algorithms can help clinicians understand risk factors for posttraumatic stress disorder (PTSD) and match them to appropriate individualized treatments, according to Isaac Galatzer-Levy, PhD, assistant professor in psychiatry and bioinformatics, NYU School of Medicine, and vice president of clinical and computational neuroscience, AiCure.
Transcript How can machine learning help forecast the course and trajectory of PTSD or depression?
The causes of PTSD and depression are very heterogeneous, vary between individuals, so you need an algorithm that’s able to accommodate a lot of information and really fit what we call an individualized predictive model. So we need to be able to take in a lot of information to understand what are the unique risk factors that lead to varied outcomes. If we do take that kind of approach, we can then start to match treatments to those risk factors to have a really individualized treatment model for the prevention of posttraumatic stress responses.
Why is it important to accurately predict PTSD in the immediate aftermath of a traumatic event?
It’s very important because the vast majority of people, even in the first world, will be exposed to a life-threatening event across their life course. The epidemiological literature showed roughly 90% of us will be exposed to some sort of event that can cause PTSD or depression, but only 10% will develop the disorder. So if you think of contexts like the emergency room, where large numbers of people come through, or the battlefield, where large numbers of people are exposed to potentially traumatic events, or even post terrorist attack events, it’s very hard to know who needs the resources and what would help them. So if you have algorithms to detect people at risk, you can much more efficiently give resources to those who are most in need.
There’s a large literature showing that if you intervene with people who don’t need help, who are going to be resilient, you can actually increase the probability of them developing PTSD. So to have effective treatment, you really have to know who needs the treatment, you need methods to identify those at risk, and you need methods to identify what are the risk factors that are driving that individual to develop the disorder.