Can Telehealth Help Diagnose ADHD in Children?

What you need to know

• Online questionnaires completed by parents and teachers can help identify ADHD in children with 82% accuracy when analyzed using machine learning.
• Core ADHD symptoms reported by parents and teachers were the most useful information for predicting an ADHD diagnosis.
• Autism symptoms can complicate ADHD diagnosis, as children with both conditions tend to have more severe ADHD symptoms reported.

Background on ADHD Diagnosis and Telehealth

Attention-deficit/hyperactivity disorder (ADHD) is a common neurodevelopmental condition that affects children’s ability to focus, control impulses, and regulate activity levels. Accurately diagnosing ADHD requires gathering information from multiple sources, including parents, teachers, and direct observation of the child by clinicians.

Recently, there has been growing interest in using telehealth methods to aid the diagnostic process for conditions like ADHD. Telehealth refers to providing healthcare services remotely using digital technologies. For ADHD diagnosis, this could involve having parents and teachers complete online questionnaires about a child’s behavior, rather than filling out paper forms in a doctor’s office.

However, it’s not yet clear how well information collected online matches up with in-person clinical assessment for ADHD diagnosis. This study aimed to explore that question using machine learning, a type of artificial intelligence that can find patterns in large amounts of data.

How the Study Worked

The researchers looked at data from 326 children and adolescents (ages 3-16) who were evaluated for suspected ADHD at a clinic in Italy. As part of the evaluation process:

1. Parents and teachers filled out standardized online questionnaires about ADHD symptoms and other behavioral/emotional issues.

2. Children completed in-person IQ testing.

3. Clinicians conducted full in-person evaluations and made final diagnostic decisions.

The researchers then used a machine learning technique called a decision tree to analyze all the questionnaire data. A decision tree is a series of if-then rules that can be used to classify data into categories - in this case, ADHD or not ADHD. They compared the decision tree’s classification to the actual clinical diagnoses to see how well it performed.

Key Findings

Accuracy of Online Questionnaires

The decision tree algorithm was able to correctly classify children as having ADHD or not having ADHD with 82% accuracy, based solely on the online questionnaire data from parents and teachers. This suggests that carefully designed online screening tools could potentially help identify children who need further evaluation for ADHD.

Most Important Information for Diagnosis

The machine learning analysis found that parent and teacher ratings of core ADHD symptoms were the most useful information for predicting an ADHD diagnosis. This included things like:

• Difficulty paying attention
• Hyperactivity
• Impulsive behavior

Other behavioral issues like oppositional behavior were also relevant, but not as crucial as the core ADHD symptoms.

Role of Autism Symptoms

About 15% of the children in the study had both ADHD and autism spectrum disorder (ASD). The researchers found that children with both conditions tended to have more severe ADHD symptoms reported by parents and teachers.

This created some challenges for the machine learning algorithm. Children with high ratings of autism-related social difficulties were more likely to be classified as having ADHD, even if they didn’t actually meet full criteria for an ADHD diagnosis.

This finding highlights the complexity of differentiating ADHD and autism, as the two conditions can have overlapping symptoms. It suggests that screening tools may need to specifically account for autism traits to avoid overdiagnosing ADHD in children who primarily have autism.

Potential Benefits of Telehealth Screening

The study results suggest several potential benefits of incorporating online questionnaires and machine learning into the ADHD diagnostic process:

1. Faster initial screening: Online questionnaires could help quickly identify children who need more thorough evaluation, potentially reducing wait times.

2. Broader reach: Telehealth methods could make initial ADHD screening more accessible, especially for families in rural areas far from specialty clinics.

3. More efficient use of clinical resources: By using technology to handle initial data collection and analysis, clinicians could focus more of their time and expertise on complex cases and treatment planning.

Limitations and Cautions

While the results are promising, there are some important limitations to keep in mind:

1. This study only looked at initial screening, not final diagnosis. In-person clinical evaluation is still crucial for accurate ADHD diagnosis and treatment planning.

2. The sample was from a single region in Italy and may not represent all populations.

3. Online questionnaires may be challenging for parents with lower education levels or limited access to technology.

4. Machine learning algorithms can sometimes perpetuate biases present in training data, so careful validation is needed before clinical use.

Conclusions

• Online questionnaires analyzed with machine learning show promise for initial ADHD screening, with 82% agreement with clinical diagnosis in this study.

• Core ADHD symptoms reported by parents and teachers are the most important information for predicting diagnosis.

• The presence of autism traits can complicate ADHD assessment and may need special consideration in screening tools.

• While promising, telehealth methods should be seen as a supplement to, not a replacement for, thorough clinical evaluation in ADHD diagnosis.

This research represents an early step in exploring how technology might improve the ADHD diagnostic process. Further studies with larger, more diverse populations are needed to validate these findings and develop clinically useful screening tools. As telehealth continues to evolve, it has the potential to increase access to care while allowing clinicians to focus their expertise where it’s needed most.