What is the scale for Autism Spectrum Disorder (ASD) screening and what score is considered statistically significant?

Autism Spectrum Disorder Screening Tools and Significant Scores

The Modified Checklist for Autism in Toddlers, Revised with Follow-Up (M-CHAT-R/F) is the most commonly studied and recommended screening tool for Autism Spectrum Disorder (ASD), with a score of ≥3 on initial screening and ≥2 after follow-up interview indicating significant risk for ASD. 1, 2

M-CHAT-R/F Screening Protocol

• The M-CHAT-R/F is a parent-rated scale that consists of an initial questionnaire followed by a structured follow-up interview for children who screen positive 1, 3
• The screening process assesses communication skills, joint attention, repetitive movements, and pretend play 1
• Initial screening with M-CHAT-R/F uses a total score cutoff of ≥3 to identify children requiring follow-up 2, 4
• Children who continue to screen positive after follow-up interview (score ≥2) have a 47.5% risk of being diagnosed with ASD and a 94.6% risk of any developmental delay or concern 2

Scoring Algorithm and Risk Levels

• A three-level risk algorithm is recommended to maximize clinical utility 2:
  • Low risk (0-2 points): No further action required unless surveillance indicates risk for ASD
  • Medium risk (3-6 points): Administer follow-up interview; if positive (≥2 points), refer for diagnostic evaluation
  • High risk (≥7 points): Bypass follow-up interview and refer immediately for diagnostic evaluation and early intervention 4

Statistical Significance and Psychometric Properties

• Positive Predictive Value (PPV) varies significantly based on population risk:
  • Low-risk samples: PPV = 0.06-0.11 without follow-up interview 1, 5
  • Low-risk samples with follow-up interview: PPV = 0.43-0.65 1
  • High-risk samples: PPV = 0.60 without follow-up, improving to 0.76 with follow-up interview 1
• Sensitivity and specificity metrics:
  • Pooled sensitivity: 0.83 (95% CI: 0.75-0.90) 5
  • Pooled specificity: 0.51 (95% CI: 0.41-0.61) 5
  • Sensitivity is higher for screening at 30 months compared to 24 months 5

Age Considerations and Performance

• The M-CHAT-R/F is designed for children aged 18-30 months 1
• PPV is lower in younger children (16-23 months: PPV = 0.28) compared to older children (24-30 months: PPV = 0.61) 1
• Sensitivity is higher when screening is performed at 30 months compared to 24 months 5
• Limited evidence supports use at 18 months in low-risk populations 5

Clinical Implementation

• A positive initial screening should lead to a follow-up interview, which if positive, should lead to a full diagnostic workup for ASD 1
• The M-CHAT-R/F has demonstrated improved utility compared to the original M-CHAT by detecting ASD at a higher rate while reducing the number of children needing follow-up 2
• Implementation of universal screening can lower the age of ASD diagnosis by approximately 2 years compared with recent surveillance findings 2

Important Caveats

• False positives may occur due to developmental concerns that may resolve and behaviors in typically developing toddlers that overlap with ASD deficits 1
• Despite lower specificity for autism at 18 months, PPV for any diagnosable developmental disorder remains high across age groups 1
• The U.S. Preventive Services Task Force notes insufficient evidence to assess the balance of benefits and harms of universal screening for ASD in children for whom no concerns have been raised 1, 3
• Clinicians should account for a child's age and presence of developmental concerns when interpreting M-CHAT scores 5