Authors: Malvika Pillai; Jose Posada; Rebecca M. Gardner; Tina Hernandez-Boussard; Yair Bannett · Research
Can AI Help Track Quality of ADHD Care for Young Children?
New AI methods could help ensure children with ADHD receive appropriate behavioral treatment recommendations from their doctors
Source: Pillai, M., Posada, J., Gardner, R. M., Hernandez-Boussard, T., & Bannett, Y. (2023). Measuring Quality-of-Care in Treatment of Children with Attention-Deficit/Hyperactivity Disorder: A Novel Application of Natural Language Processing. medRxiv. https://doi.org/10.1101/2023.06.12.23291071
What you need to know
- Only about 30% of young children with ADHD receive recommended behavioral treatment recommendations at their first visit
- Current methods to track quality of ADHD care are limited and labor-intensive
- AI technology can help automatically assess whether doctors are following treatment guidelines
The Challenge of Tracking ADHD Care Quality
Imagine trying to review thousands of detailed medical records by hand to check if doctors are providing the right care recommendations for children with ADHD. This is the current reality for many healthcare organizations, and it’s an enormous challenge. ADHD affects 8-10% of children in the US, making it the most common behavioral disorder in childhood. Guidelines recommend that young children with ADHD should first try behavioral therapy before medication, but tracking whether this happens has been extremely difficult.
A New Approach Using AI
Researchers at Stanford University developed an innovative solution using artificial intelligence (AI) to automatically review doctors’ notes and check if they recommended behavioral therapy. The AI system, based on a technology called BERT (Bidirectional Encoder Representations from Transformers), was trained to understand the complex language in medical records and identify when behavioral treatment was recommended.
What the Study Found
The research team tested their AI system on records from over 400 children aged 4-6 years who were diagnosed with ADHD. They discovered that only about 30% of children received the recommended behavioral treatment at their first ADHD visit. The AI was highly accurate at identifying these recommendations, correctly catching 72-92% of cases where behavioral therapy was mentioned.
Why This Matters
Traditional methods of checking healthcare quality involve either:
- Looking at very basic measures like medication prescriptions
- Manually reviewing a tiny sample of patient records
- Both approaches miss important aspects of care
The AI approach allows healthcare organizations to:
- Review all patient records quickly and accurately
- Track multiple aspects of care
- Provide regular feedback to doctors
- Identify areas where care could be improved
What This Means for You
If you have a child with ADHD or work in healthcare, this research has important implications:
- For Parents: Ask about behavioral therapy options when your child is diagnosed with ADHD, especially for young children (ages 4-6). Evidence shows this should be the first treatment approach.
- For Healthcare Providers: Regular feedback about treatment recommendations can help ensure you’re following best practices for ADHD care.
- For Healthcare Organizations: AI tools could help track and improve the quality of ADHD care more efficiently and comprehensively.
Conclusions
- AI technology can help ensure children with ADHD receive appropriate care recommendations
- Regular monitoring of treatment quality could help improve outcomes for children with ADHD
- This approach could be expanded to track quality of care for other behavioral and mental health conditions