Sign Up Now for a FREE Enrollment Consultation!

We respect your family's privacy.

First-of-its-kind study uses machine learning algorithms to predict risk of childhood PTSD

July 26, 2017 0 Comments

There is growing evidence of the effectiveness of machine learning (ML) algorithms in health care. It has found application in the areas of heart disease and diabetes along with some rare diseases. However, the predictive ability of ML has not been tested in identifying the risk of childhood post-traumatic stress disorder (PTSD). Identifying PTSD risks among children and adolescents may provide opportunities for preventive treatments and improved outcomes. Studies are increasingly highlighting the importance of providing early intervention programs to prevent PTSD among children.

A new proof-of-concept study published in the journal BMC Psychiatry on July 10, 2017, is the first to use an ML-based algorithm to identify risk factors for childhood PTSD. The algorithm also explains the factors which cause PTSD after trauma, such as the child’s pain and parents’ stress levels. The findings suggest that potential preventative treatment for PTSD targeting a child’s specific needs should include better pain treatment for children and increased support for parents during the period immediately following the trauma.

Past research shows that 26 percent of American children are likely to experience a traumatic event before the age of four. The American Psychological Association (APA) notes that in some community samples, more than two-thirds of children experience a traumatic event by the age of 16, although, not all such children develop PTSD symptoms. An estimated 4 percent adolescents aged 13-18 have a lifetime prevalence of PTSD, whereas 1.4 percent of this population experience the severity. Conventional methodologies for predicting childhood PTSD have met with limited success.

Identifying causal pathway to PTSD is helpful in preventive care

To achieve their objective, the researchers applied ML predictive classification models to 163 patients, aged between 7 and 18, who were hospitalized after suffering injuries. The injured children were evaluated immediately after their hospitalization (within hours or days) and were re-evaluated three months after being discharged. The predictive models encompassed 105 variables collected during the hospitalization, which spanned a wide array of bio-psychosocial areas such as genetic factors, parents’ symptoms, stress, the extent of the injury, and child symptoms and functioning.

After three months, it was found that 7 percent of the evaluated patients were suffering from high levels of PTSD. The researchers used five ML classification methods to build a predictive classification model with a “considerable predictive signal.” They went on to explain that the signal obtained from the ML predictive models is much stronger than the performance exhibited by conventional classification methods.

The researchers also used ML “causal discovery” feature selection methods to shortlist 10 “most stable” variables that determine “causal pathway to PTSD.” The variables include the history of the disease, severity of pain experienced by the child, mutations in specific genes and severity of parental distress levels among others.

Child’s recovery after trauma

Glenn Saxe, professor of child and adolescent psychiatry at New York University (NYU) Langone Health’s Child Study Center and lead researcher, explains that PTSD is an interplay of complicated interactions between several processes. The discovery of a set of causal variables may offer opportunities for designing precautionary measures. For instance, preventive medication may be used to target biological processes encoded in the specific genes identified in the set of causal variables. Similarly, the severity of pain experienced by the child and severity of parental distress levels are also capable of being directly treated. The researchers also discovered that the child’s history of breastfeeding during infancy and regularity of attending religious services were protective factors, highlighting that attachment, community and spirituality play important roles in a child’s recovery after trauma.

Without timely interventions, the after-effects of trauma can have a lasting impact on the mental health of children and adolescents. As one of the leading therapeutic boarding schools, White River Academy aims to provide the required help to teenage boys aged between 12 and 17, to recover from their mental health problems, including PTSD childhood trauma. Call our 24/7 helpline number or chat online with one of our experts to know about the best PTSD treatment centers in your vicinity.

What do you think?

Leave a Reply

*

View More

Reviews

We would like to thank all the wonderful staff at WRA for the great parent weekend. We enjoyed it and felt that we learned valuable insights on Positive Peer Culture and the values we must have and the importance of family commitment to each other...