Screening for Autism Spectrum Disorder in a Naturalistic Home Setting Using the Systematic Observation of Red Flags (SORF)

Background: While early detection of autism spectrum disorder (ASD) is critical for timely initiation of intervention, development of more effective screening tools is necessary to warrant a recommendation for routine universal screening (Campos-Outcalt, 2011). Time and resource constraints also limit screening procedures (Daniels et al., 2014), leading to delays in diagnosis until 4 ½ years of age (Baio, 2014). This is further exacerbated for non-white children in low-resource areas (Daniels & Mandell, 2013) due to limitations in accessibility of clinical services. Improved screeners are needed that not only effectively identify children at risk, but also offer increased feasibility for widespread use across community settings.

The Systematic Observation of Red Flags (SORF) is an observational measure that can be used by non-experts (e.g., paraprofessionals, undergraduate students) to detect risk for ASD in young children (Dow et al., 2016). Studying this measure in the home context could improve community-viable screening options accessible to more families, especially those with more limited resources.

Objectives: To examine psychometric properties of the SORF home measure for 18-24 month olds, including item-level performance and overall sensitivity, specificity, and recommended cutoff scores.

Methods: Preliminary analyses were conducted on a sample of 61 participants (31 with ASD, 14 developmental delayed, 16 typically developing) recruited by the FIRST WORDS® Project at Florida State University. A sample size of 217 will be completed at the time of presentation. The sample was randomly selected, stratified by race to optimally represent minority groups (i.e., 13% Black, 11% Hispanic, 9.5% Biracial). Receiver operating characteristic (ROC) curves and analysis of variance (ANOVA) were used to evaluate individual items and summary scores.

Results: Preliminary results based on the Composite summary score composed of best performing items suggest good discrimination between ASD and nonspectrum groups (AUC=.89), with high sensitivity and specificity (sensitivity=.90, specificity=.83). Group differences with large effect sizes were found between ASD and nonspectrum groups for all summary scores and social communication domain scores. Best performing items include showing and pointing (AUC=.83), eye gaze (AUC=.81), nonverbal communication (AUC=.81), and more interest in objects than people (AUC=.81).

Conclusions: Preliminary results support the utility of the SORF as an effective observational screening tool for ASD that can be used in a naturalistic home setting by non-experts. Data on individual item utility and required length of observation will be further analyzed in an effort to create a briefer measure to improve feasibility. Future studies will also examine the SORF’s utility in a younger sample at 12-18 months, and in combination with parent report. Applying the SORF to the home context could provide a community-viable screening option accessible to more families, aiming to improve services for minority and low resource children and decrease delays in diagnosis of ASD.