Language delay is not predictable from available risk factors.

Area of Science:

• Child development
• Pediatric health
• Speech and language pathology

Background:

• Early identification of language delay is crucial for timely intervention.
• Existing predictive models for language delay at 30 months require further investigation.
• The role of universal health contacts in screening for developmental disorders needs evaluation.

Purpose of the Study:

• To identify factors associated with language delay in 30-month-old children.
• To assess the efficacy of using pre-existing variables for predicting language delay.
• To determine if active family contact is necessary for identifying language delay.

Main Methods:

• A pilot universal 30-month health contact was utilized.
• A two-item language screen was administered by health visitors.
• Data from 315 children were analyzed, with 33 exhibiting language delay.

Main Results:

• Seven variables were initially associated with language delay.
• Logistic regression analysis identified five significant predictive variables.
• The predictive model had a sensitivity of 89% but a low positive predictive value of 15%.

Conclusions:

• Predictive risk factors alone cannot reliably identify children with language delay at 30 months.
• Defining an "at risk" population based on demographic factors like male gender proved unreliable due to low specificity.
• Direct clinical contact with all families is necessary for accurate identification of language delay, an ESSENCE disorder.