An evaluation of the diagnostic validity of the structured questionnaires of the adult Asperger's Assessment.

Researchers tested three questionnaires commonly used to diagnose autism in adults in England's health system. While these tools were good at identifying people who do have autism, they were poor at correctly identifying people who don't have autism. This means the questionnaires might incorrectly suggest autism in people who don't actually have it, which could lead to unnecessary assessments or misdiagnosis.

This study evaluated the diagnostic validity of three structured questionnaires used in the Adult Asperger's Assessment (AAA) by an NHS diagnostic service in England: the Adult Autism Quotient (AQ), Empathy Quotient (EQ), and Relative's Questionnaire (RQ). Using Receiver Operating Curve analysis, researchers found that while these questionnaires showed good sensitivity for detecting positive autism spectrum disorder diagnoses, they had poor specificity for excluding individuals without ASD. Binary logistic regression analysis revealed that even when combined, the questionnaires demonstrated limited diagnostic validity. The findings suggest these widely-used clinical tools may have significant limitations in the diagnostic assessment process.

These findings suggest current structured questionnaires may need revision or supplementation with additional assessment tools. Clinicians should interpret questionnaire results cautiously and rely more heavily on comprehensive clinical evaluation rather than questionnaire scores alone for autism diagnosis in adults.

Sample size not reported. Study type unclear. Limited details about participant characteristics or diagnostic criteria used. No information about inter-rater reliability or comparison with gold standard diagnostic methods.

The Adult Autism Quotient (AQ), the Empathy Quotient (EQ) and the Relative's Questionnaire (RQ) were used as part of the Adult Asperger's Assessment (AAA) by a diagnostic service for adults without an intellectual disability with suspected autism spectrum disorder (ASD). This service is part of the National Health Service (NHS) in England. Little is known about the utility of these structured questionnaires despite wide use in clinical practice. It was investigated whether the questionnaires could discriminate between individuals with and without a diagnosis of ASD.

Receiver Operating Curve analysis showed good levels of sensitivity to detect a positive diagnosis, but the specificity to exclude those without a diagnosis was poor. A binary logistic regression showed that a combination of the questionnaires also showed limited diagnostic validity. These findings have clinical implications in reviewing the efficiency of the assessment process.