Investigating the clinical usefulness of the Social Responsiveness Scale (SRS) in a tertiary level, autism spectrum disorder specific assessment clinic.

Doctors tested the Social Responsiveness Scale (SRS) in a busy autism clinic. They gave the 65-item parent form to the children already sent for possible ASD.

The team compared each SRS score with the final clinic diagnosis. They wanted to know if the cut-off could rule-in or rule-out ASD on its own.

The SRS caught almost every child later diagnosed with ASD (high sensitivity). Sadly, it also flagged many who did not have ASD (very low specificity).

Using the published cut-off would have over-diagnosed 45 percent of the sample. The scale is better at saying "probably not ASD" than "definitely ASD."

Sappok et al. (2013) saw the same pattern with the ADOS in adults with ID: high sensitivity but lots of false positives. Both studies warn that a single score should never replace clinical judgment.

Bitsika et al. (2019) went deeper, showing that ADOS-2 severity changes which SRS-2 items separate boys from girls. This means specificity problems may shift depending on the child's sex and severity level.

Berument et al. (2005) fixed a similar issue by creating a new cut-off (15) for their adapted PL-ADOS. The SRS may also need a local cut-off rather than the manual one.

If you work intake for an ASD clinic, treat the SRS as a red-flag, not a verdict. Use it to decide who needs a full assessment, not to tell parents their child has autism. Pair it with observation tools like the ADOS and always adjust the cut-off for your own population.