Diagnostic and assessment issues related to pharmacotherapy for children and adolescents with autism.
No lab test can track autism medication effects, so stick with standardized behavior checklists.
01Research in Context
What this study did
Davison et al. (1991) wrote a narrative review. They talked about how hard it is to measure autism before giving kids pills.
They said no blood test or brain scan can spot autism. Teams must watch behavior and use rating scales.
What they found
The paper did not run an experiment. It simply warned that without good checklists, doctors cannot tell if a pill helps or harms.
They urged everyone to pick the same rating tools so results can be compared.
How this fits with other research
Hudson et al. (2012) later checked 33 drug trials and found only a few pills have solid proof. Their big review swallows the 1991 warning and adds real numbers.
Bolte et al. (2013) counted 289 different tools across autism studies. That huge mess proves the 1991 plea for standard scales is still ignored.
Wang et al. (2025) hunted for early biological markers. They conclude markers are still too weak for clinic use, so behavior ratings remain the gold standard, just like M et al. said.
Why it matters
When a pediatrician wants to start risperidone, you can hand them the ABC or SRS instead of a vague parent note. Picking the same tool each visit turns “he seems calmer” into real data you can graph and defend to insurance.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Pick one published scale (e.g., ABC) and plot baseline scores before the next med change.
02At a glance
03Original abstract
Autism involves not only developmental delays but also aberrant behavior, both of which change in nature over time. Rating instruments may be useful to assess maladaptive and adaptive behaviors of autistic children in a standardized way and, perhaps, to measure change due to treatment. With the expansion of basic science, knowledge, and technology, there is increasing evidence that autism is etiologically heterogeneous. Currently, there is no biological marker specific to autism, although hyperserotonemia is a consistent finding in one third of autistic children. An aim of basic science research has been to develop a rational pharmacotherapy based upon the underlying neurochemistry. However, at the present time, this approach has not always been successful. It is expected that the development and use of more restrictive criteria, delineation of subtypes of autism, and interaction of descriptive, behavioral, clinical, and basic research will lead to more effective planning for treatment. The relationship of assessment to treatment response is presented and discussed.
Behavior modification, 1991 · doi:10.1177/01454455910153004