What is your approach to evaluating Tardive Dyskinesia in nonverbal patients?   

Is there an equivalent monitoring alternative to an AIMS score? Example: Evaluating nonverbal ASD patient on long term antipsychotics with random movements that may or may not be involuntary.