Examining Disparities in Duration of Screening-to-Diagnosis Time in a Multistage, Early Intervention-Based Screening Protocol for ASD

Background: There are long-standing health disparities in early diagnosis of autism spectrum disorder (ASD) for children from poor, racial and linguistic minority families (Durkin et al., 2010; Mandell et al., 2009; Zuckerman et al., 2013). The implementation of universal screening protocols is one attempt to overcome these inequalities by engaging in regular developmental screening of all children, a practice recommended by the American Academy of Pediatrics (2006) and Centers for Disease Control (2012). It is not known whether the implementation of multistage universal screening procedures reduces health disparities in the duration of the screening, referral, and diagnostic evaluation process.

Objectives: To determine whether the implementation of a multistage universal screening and assessment procedure in Early Intervention (EI) settings will reduce disparities in age of receipt of ASD diagnosis by shortening the time between the onset of screening and the receipt of diagnosis.

Methods: Participants are families of children enrolled in EI who are between ages 14-33 months (N = 125, M_age= 24.7 months at Stage 1 of screening, 83% male). Children were evaluated as part of a 2-stage screening procedure administered by their EI specialists to identify children at risk for ASD; those at risk were referred for diagnostic assessment. Parents self-identified as racially (57% identify as racial minorities), linguistically (46% Non-English primary language) and economically (66% of families earn <$45,000/year) diverse. Parent demographics including self-identified race, English language proficiency, and household income were of interest. Duration of the screening and assessment process was measured as the time in days required to proceed from Stage 1 (parent completion of EI-administered ASD screening questionnaires), through Stage 2 (EI-administered play-based observation screener for ASD), and ending at Stage 3 (diagnostic evaluation). Analyses examined whether demographics predicted duration of screening protocol from Stage 1 to diagnostic evaluation.

Results: Mean duration of the screening and assessment process was 2.8 months (SD = 2.0). A linear regression was conducted to predict duration of the screening and assessment protocol based on demographic factors of race (dichotomized into White and non-White), English language proficiency (5-point scale), and household income (10-point scale). The overall model was significant, F(3,121)=3.08, p=.03, R² = .07. Income was the only significant predictor (ß=-.215, t=-2.26, p=.03); parent race and English proficiency did not predict duration. For families with incomes ≤$45,000, mean duration was 3.2 months, whereas for families with incomes >$45,000, time between initial screening and diagnostic evaluation was 2.1 months.

Conclusions: In a diverse, urban sample, children from lower income households were found to be at particular risk of delayed diagnosis, even after being identified at-risk. Despite utilization of a free, universal screening and diagnostic assessment procedure, inequity still exists within the screening system. Nonetheless, for poor families, overall duration was markedly shorter than the nearly eleven-month delay reported in other studies (Mandell, Novak, & Zubritsky, 2005). This finding has implications for professionals engaging in screening of young children, as professional biases and/or family stressors may be impacting fair and equitable use of universal screening protocols.