Underdiagnosis and referral bias of autism in ethnic minorities.

Begeer et al. (2009) asked Dutch pediatricians to read short stories about kids with autism.

Each story was the same except for the child’s ethnic name. Doctors then chose whether to refer the child to an autism clinic.

Half the doctors also filled out a short rating scale before deciding.

Doctors picked more majority-group kids for referral than minority kids when they used only their gut feeling.

When doctors first completed a quick checklist, the gap almost disappeared.

Structured questions helped doctors see autism signs equally in every child.

Kerub et al. (2021) followed real Bedouin toddlers and found the same pattern: equal screen scores, but 43% of Bedouin families never finished diagnosis versus 16% of Jewish families.

Fombonne et al. (2022) looked at Black and White preschoolers already referred to a U.S. clinic and saw no difference in autism severity, showing the bias happens before referral, not after.

Kim et al. (2025) pooled twenty years of U.S. data and confirmed Hispanic and Native American children are still under-identified, proving the 2009 warning still matters today.

If you rely on casual observation, you may miss autism in minority clients. Add a brief rating scale like the M-CHAT or SCQ to every intake. One extra minute can cut your own referral bias and get kids into therapy sooner.