Medical Insurance Claims Data May Help Predict The Likelihood Of Autism In Children

A machine learning model based on children’s medical claims data can predict autism at age 24 months. The model was able to detect more than three quarters of autism cases. When the predictor variables were separated by outpatient and inpatient visits, the model accuracy increased for children with autism as young as 24 months. The model was more accurate than M-CHAT the most commonly used autism screening tool. M-CHAT has very high rates of false positives and false negatives. The model accuracy increased with children’s ages.

The researchers said the model accuracy indicates that before the . . .