Reducing Disparities in Timely Autism Diagnosis through Family Navigation

Background: Emerging evidence demonstrates that Autism Spectrum Disorder (ASD) can be reliably diagnosed by age two, and that early identification, defined as diagnosis before age 3, can improve outcomes. Low-income and minority children with ASD, however, are diagnosed later than their white and more financially advantaged peers. Feasible, culturally appropriate interventions with broad scale-up potential are necessary to reduce this disparity. One promising strategy is patient navigation. Patient navigation has its origins in cancer care and has demonstrated efficacy in improving adherence to follow-up visits after a screening abnormality and decreasing the time from an abnormal screen to diagnostic resolution. We reframed the model from patient to family navigation and focused on a defined episode of care, beginning with an abnormal autism screen and ending with the completion of the diagnostic assessment.

Objectives: To assess the feasibility and acceptability of family navigation and its potential to decrease time to diagnostic resolution among of low-income and minority families of children, ages 15 months to 3 years, who have positive behavioral screens indicating a need for an ASD diagnostic assessment.

Methods: Pilot RCT of 40 children who failed the M-CHAT or whose clinicians have clinical concerns related to ASD and who receive primary care within an integrated urban healthcare network.  Children with confirmed risk for ASD, based on the results of the M-CHAT Follow-up Interview, are randomized to receive either Family Navigation (FN) until diagnostic resolution or usual care (UC).We are examining the following process measures: participant recruitment and flow, delivery of navigator services, collection of follow-up data, and family satisfaction with FN. The primary outcome is time to diagnostic resolution. We will conduct an intention-to- treat analysis and compare time to diagnosis between groups by survival analysis, creating Kaplan-Meier plots and applying a Cox proportional hazards model.

Results: To date, we have screened 50 children of whom 34 met eligibility criteria and 32 were randomized. The FN and UC groups are demographically similar. Children were 59% male; mean age 26 months. Families were 77% non-white and 93% low-income.  All FN families have had at least 2 visits with navigator, with an average of 4 encounters.  To date, 11 of 16 FN families versus 7 of 16 UC families have completed the diagnostic assessment; 4 FN and 4 UC families remain in process, and 1 FN family versus 5 UC families have not completed the diagnostic assessment during the follow-up period. Among the 18 families who have completed the diagnostic process, 10 received an ASD diagnosis; mean time from confirmatory screening to diagnostic resolution among all completers was 77 days (SD 39).

Conclusions: Results support the intervention’s feasibility and promise to increase the proportion of children who complete an ASD diagnostic assessment. They are consistent with an earlier pilot trial in which 19 of 20 FN families versus 10 of 19 UC families completed the diagnostic assessment. This model will be expanded and tested in a recently awarded NIMH R01, Early Identification and Service Linkage for Urban Children with Autism.