P2.061: Autism Screening in Pediatric Resident Continuity Clinic: An Evaluation of Streamlined Pathway to Rapid Diagnosis and Analysis of Interrater Reliability Between Trainee and Expert

Background: The Rapid Interactive Screening Test for Autism in Toddlers (RITA-T) is a novel level 2 screening tool for 18–36 month-olds. It has been shown to reliably expedite early identification of Autism Spectrum Disorder (ASD) and reduce historically long wait times to diagnosis​​. The RITA-T has a positive predictive value of over 95% for medium to high-risk ASD, and has high reliability even when administered by other providers of ASD care such as speech and language pathologists, occupational therapists, and early interventionalists​​. However, there is no data on the reliability of RITA-T scores administered by resident trainees, who provide primary care to nearly one million children in the United States annually​4. Therefore, we examined the inter-reliability of RITA-T scores administered by pediatric residents in a continuity clinic compared to subsequent evaluations by a Developmental Behavioral Pediatrician (DBP).

Methods: Trained pediatric residents administered the RITA-T to 18–36 month-olds referred by a primary care resident physician for ASD evaluation. Participants scoring medium or high-risk were referred to a DBP for confirmatory RITA-T (medium risk=13-17 points; high risk=17-30 points) and additional ASD testing. Wait times to DBP evaluation were recorded, and the reliability of the RITA-T scores were analyzed via Pearson's correlation and Cohen’s Kappa.

Results: The RITA-T was administered to 24 patients over 21 months, the majority of whom were non-Hispanic Black, male, and with public insurance (Table 1). The mean age was 23.92 months (SD=5.26). The average resident RITA-T score was 19.43 (SD=3.89) with 87.5% (n=21) receiving subsequent referrals to the DBP. 19.1% of these patients scored medium-risk (n=4) and 80.9% scored high-risk (n=17) on the resident RITA-T. The average wait time for DBP appointment was 87.7 days (SD=49.8). Of those referred, 100% (n=21) received an ASD diagnosis and 85.7% (n=18) were placed in the same risk category by both the resident and DBP (κ=0.35). For participants seen by a DBP within 90 days, the mean score difference between resident and specialist was 2.54 points (SD=2.62) with a 90.1% agreement in risk stratification between the two raters. Resident RITA-T scores correlated moderately with DBP-conducted diagnostic tests, including ASD-PEDS (r=t 0.47) and CARS-2 (r=0.39).

Conclusion: Resident trainees may reliably administer the RITA-T to screen for ASD, thus improving triage of medium and high-risk patients. This was particularly true when patients were evaluated by a DBP within 90 days, highlighting the importance of timely follow-up given rapid development during this period. Our results support expanded use of the RITA-T in resident continuity clinics to expedite patient access to resources and accommodations for their ASD. Future efforts will examine larger samples to boost the reliability of these findings, while exploring other aspects of pediatric resident training for the care of patients with ASD.