From Recognition to Diagnosis: Caregiver Response, Help-Seeking Pathways, and Access-Related Factors Associated with Autism Diagnostic Delay in Jordan

Background: Autism spectrum disorder (ASD) is often diagnosed well after developmental concerns first emerge, and evidence on factors associated with the duration of the recognition-to-diagnosis pathway remains limited in the Middle East. This study aimed to identify factors associated with the overall recognition-to-diagnosis interval and potential areas for earlier recognition, referral, and access to appropriate assessment. Methods: This multisite cross-sectional survey of 384 caregivers of children with confirmed ASD was conducted across Jordanian governorates. Diagnostic timing was known for 338 participants. Five sequential nested ordinal logistic regression models were fitted on a common complete-case sample (N = 299), successively adding background characteristics, recognition, caregiver response, help-seeking route, and access/professional response variables. Robustness was assessed via grouping-specific binary models, multiple imputation, and bootstrap resampling. Results: Caregivers reported first concerns at a median age of 2.0 years. Among those with known diagnostic timing, 60.7% were in the “6 Months to 1 Year” delay category or longer, 46.2% were in the “1–2 Years” category or longer, and 24.0% were in the “More than 2 Years” category. The background characteristics model showed limited explanatory capacity (Nagelkerke R2 = 0.021), and adding recognition variables did not improve model fit (p = 0.562). Fit improved significantly with caregiver response (p = 0.001), help-seeking route (p = 0.008), and access/professional response (p < 0.001). The final model reached a Nagelkerke R2 = 0.178, indicating modest overall explanatory capacity. Longer diagnostic delay was independently associated with caregivers who reported that early signs had initially not been acted upon because they were interpreted as part of normal development (AOR = 2.21). It was also associated with first contact via a speech/learning center (AOR = 2.27) or other service (AOR = 2.72) rather than a pediatrician. Caregiver-reported previous professional reassurance that the child did not have ASD was also associated with longer delay (AOR = 2.18). Professional reassurance was the most consistent correlate across sensitivity analyses. Definite appointment difficulty showed a significant Yes-versus-No contrast (AOR = 1.81), although the appointment difficulty variable was not statistically significant in the global test. Sociodemographic factors showed no independent association. Among 11 exploratory barriers, only prior misdiagnosis survived multiplicity correction (AOR = 2.21). Conclusions: In this Jordanian cohort, the length of the recognition-to-diagnosis interval was associated with factors operating after developmental concerns were first recognized, rather than with the timing or breadth of recognition itself. Caregiver response, entry route into care, and professional response emerged as potentially important pathway markers. However, the modest explanatory capacity of the final model indicates that substantial variability in diagnostic delay remains unaccounted for by the measured variables. These findings support provider- and system-level measures, including clearer referral pathways, explicit follow-up when reassurance is provided, improved appointment access, and expanded diagnostic capacity, complemented by caregiver-facing information and support.