Are We So Worried About Overdiagnosing CAS That We're Missing It?
Why clinical judgement matters more than ticking diagnostic boxes.
I keep hearing that Childhood Apraxia of Speech (CAS) is being overdiagnosed.
And I keep wondering: where are all these overdiagnosed children?
Because in my clinical practice, I'm frequently seeing the opposite.
Children who have spent months, sometimes years, receiving phonological intervention with limited progress, yet continue to demonstrate difficulties far more consistent with CAS.
Sometimes their therapists recognise the possibility of CAS but feel unable to diagnose it because the child doesn't meet a particular checklist threshold or assessment cutoff.
Children who struggle to combine even two sounds into a syllable. Children whose speech breaks down during transitions between sounds and syllables. Children whose difficulties aren't adequately explained by phonological errors alone.
Yet CAS has been ruled out.
Are we becoming so concerned about overdiagnosing CAS that we're creating unnecessary barriers to recognising it?

The consequences of getting it wrong aren't necessarily equal
Of course, CAS can be misdiagnosed.
Phonological and articulation disorders, dysarthria, speech motor delay, structural differences and prosodic differences associated with autism can produce overlapping characteristics.
Differential diagnosis matters enormously.
There are legitimate concerns about unnecessary diagnoses, inappropriate treatment intensity and the ethical responsibility to recommend intervention based on a child's actual needs.
But what alarms me most is the opposite problem: children with genuine CAS missing out on appropriate motor speech intervention while clinicians remain uncertain or reluctant to diagnose it.
A child incorrectly diagnosed with CAS may receive unnecessary therapy. Their progress and response to intervention should prompt us to review the diagnosis and treatment plan.
But a child whose CAS goes unidentified may spend years receiving therapy that doesn't adequately address their speech motor planning difficulties.
Those are years of potential progress that cannot simply be given back.
In my CAS-only clinic, I encounter concerns about missed CAS almost weekly, through second-opinion assessments and supervision of other speech pathologists.
I regularly hear about children making limited progress with phonological intervention, while clinicians struggle to reconcile compelling clinical evidence of CAS with diagnostic rules they believe they must follow.
I cannot establish from my clinical experience whether CAS is more frequently underdiagnosed than overdiagnosed in the wider population.
But I know which problem is keeping me awake at night.
When did research criteria become clinical diagnostic rules?
ASHA (2007) identified three consensus characteristics associated with CAS:
Inconsistent errors on consonants and vowels across repeated productions.
Lengthened or disrupted coarticulatory transitions between sounds and syllables.
Inappropriate prosody, particularly lexical or phrasal stress.
These characteristics are important.
But ASHA explicitly states:
"Importantly, these features are not proposed to be the necessary and sufficient signs of CAS."
In other words, ASHA did not establish that every child must demonstrate all three characteristics to receive a CAS diagnosis.
Similar concerns apply to numerical cutoffs, such as requiring four Mayo-10 characteristics across three assessment tasks. (Shriberg et al., 2012).
Researchers need reproducible criteria to select participants. That's understandable.
But research classification criteria are not automatically validated clinical diagnostic rules.
A threshold designed to identify a relatively homogeneous research sample may exclude children who genuinely have CAS.
What evidence demonstrates that a particular cutoff is sufficiently sensitive to identify all children with CAS?
If a child demonstrates compelling evidence of impaired speech motor planning but doesn't satisfy a research threshold, does that mean they don't have CAS?
Or simply that they wouldn't qualify for that particular study?
Those aren't necessarily the same thing.
What Murray and McCabe's research can teach us
Some of the most interesting evidence supporting a nuanced approach to CAS diagnosis comes from Murray, McCabe and colleagues.
In their 2015 study, Murray et al. assessed 47 children referred by community speech pathologists with suspected CAS.
Of those children, 32 were classified as having CAS, including four with additional speech difficulties. The remaining 15 were classified as having phonological disorders, dysarthria or structural differences.
These findings are sometimes discussed in relation to CAS overdiagnosis.
But there's an important distinction.
Suspecting CAS and referring for further assessment is not the same as making an incorrect, definitive diagnosis.
The study was not designed to determine how frequently CAS is overdiagnosed in the general population.
Nor could it establish how many children with genuine CAS might have been overlooked because they were never referred with suspected CAS.
We need to remember these limitations when interpreting the findings.
Is inconsistency as diagnostically important as we think?
One particularly interesting finding concerned inconsistency.
Despite being one of ASHA's three consensus characteristics, inconsistency measured using the DEAP Inconsistency subtest did not significantly contribute to the researchers' final diagnostic models.
When examined independently, this measure classified children with only 30% accuracy in their discriminant analysis.
Instead, the final model included syllable segregation, lexical stress, polysyllabic word accuracy and articulatory accuracy during repetition of /pətəkə/.
This doesn't mean inconsistency is irrelevant to CAS. The CAS group demonstrated greater average inconsistency than the non-CAS group.
But it does challenge the weight we sometimes place on this individual characteristic.
How many children might have CAS dismissed because they aren't considered inconsistent enough?
What about dysarthria?
Another important consideration is that CAS and dysarthria aren't necessarily mutually exclusive.
Murray et al. identified children with CAS alongside additional speech difficulties, including dysarthria.
McCabe, Korkalainen and Thomas (2024) also discuss the complexity of overlapping motor speech disorders.
Identifying dysarthria doesn't automatically rule out co-occurring CAS.
A child may demonstrate impaired speech motor execution alongside difficulties with speech motor planning and programming.
We must investigate potential neuromuscular and structural abnormalities and seek appropriate medical assessment when indicated.
But identifying one condition shouldn't prevent us from investigating another.
Sometimes the question isn't which single diagnosis fits, but which combination of impairments best explains the child's speech.
Are we counting characteristics instead of understanding the problem?
Consider a child diagnosed with inconsistent phonological disorder.
They've received phonological intervention for an extended period with limited progress.
During a motor speech assessment, however, we observe unusual difficulty establishing movement sequences, disrupted transitions between sounds and syllables, and difficulty combining even relatively simple sounds.
These difficulties aren't adequately explained by a phonological disorder alone.
Yet CAS may have been dismissed because the child wasn't sufficiently inconsistent or didn't demonstrate enough checklist characteristics.
Clinical judgement isn't simply deciding whether a child has four, five or six features.
It's understanding what those features mean.
We need to consider developmental history, hearing, articulation/phonological abilities, speech motor performance, oral structures, possible neuromuscular involvement and alternative explanations.
Ruling out other conditions isn't enough to establish CAS. We still need positive evidence of impaired speech motor planning and programming.
But if the overall presentation strongly supports CAS, should we withhold the diagnosis simply because the child doesn't demonstrate enough characteristics across a prescribed number of tasks?
I don't believe so.
This is especially relevant for young children unable to complete formal assessment tasks. A recent survey of 298 speech-language pathologists found that 40% believed CAS could not be diagnosed before age three, while 52% reported that they would not disclose a CAS diagnosis to caregivers of children in this age group (Valentine et al., 2026). The authors emphasised that chronological age is not itself a diagnostic criterion and that some children can be reliably diagnosed before their third birthday.
This raises another important question: are we sometimes allowing rigid assumptions about when CAS can be diagnosed to delay its identification?
For young children who cannot yet complete formal assessment tasks, carefully documented diagnostic therapy can help us understand their movement sequences, responses to cueing and the nature of their speech breakdowns.
And for older children, we must consider their history.
A child who was highly inconsistent at three may be considerably more consistent at eight, particularly following effective motor speech intervention.
Their current presentation doesn't erase their earlier difficulties.
For more on eliciting diagnostic evidence, see Are We Taxing the Speech Motor System Enough to Identify CAS?.
Could the future of CAS diagnosis involve looking beyond checklists?
McCabe, Korkalainen and Thomas (2024) discuss emerging approaches that may help address diagnostic uncertainty.
Their review highlights overlapping characteristics, co-occurring disorders and limitations in existing assessment approaches.
They also discuss work by Littlejohn and Maas (2023), proposing a shift towards identifying the underlying processes within speech production that are impaired. Rather than relying primarily on matching observable characteristics to diagnostic labels, this approach asks where the breakdown occurs.
Comprehensive validated tools aren't yet available, but I find the direction encouraging.
It reflects something I believe is central to good clinical assessment:
Our goal should be to understand why a child is struggling to produce speech, not simply whether they demonstrate enough characteristics to satisfy a checklist.
Final thoughts
I understand the concern about CAS overdiagnosis, particularly where a diagnosis might lead to unnecessary or excessively intensive intervention. But...
My greater concern is the child with genuine CAS who spends years receiving intervention that doesn't address their underlying difficulties.
Early childhood offers valuable opportunities for speech and communication development.
While meaningful progress remains possible later,
We should not be comfortable allowing avoidable diagnostic uncertainty to delay appropriate intervention.
Research criteria, checklists and standardised assessments are valuable. They should inform clinical reasoning, not replace it.
Clinical judgement isn't guesswork. It involves understanding speech development, considering alternative explanations, recognising co-occurring difficulties and integrating evidence over time.
When we're uncertain, we can acknowledge that uncertainty, seek specialist input, undertake diagnostic therapy and continue gathering evidence.
We don't need to abandon diagnostic rigour.
We need to ensure that our pursuit of diagnostic certainty doesn't become an unnecessary barrier to helping children with CAS.
Are we trying to establish whether a child has CAS—or simply whether they can satisfy the diagnostic rules we've chosen to apply?
Because those aren't always the same question.
Does this resonate with you?
Have you ever suspected a child had CAS but felt you couldn't confidently make the diagnosis?
If you're looking for more of this kind of nuanced clinical reasoning, opportunities to work through tricky cases, and ongoing guidance and mentorship, the OKAT CAS Academy may be just what you're looking for.
Join the OKAT CAS Academy Waitlist — launching February 2027.
References
American Speech-Language-Hearing Association. (2007). Childhood apraxia of speech [Technical report]. https://www.asha.org/policy/TR2007-00278/
Littlejohn, M., & Maas, E. (2023). How to cut the pie is no piece of cake: Toward a process-oriented approach to assessment and diagnosis of speech sound disorders. International Journal of Language & Communication Disorders. https://doi.org/10.1111/1460-6984.12934
McCabe, P., Korkalainen, J., & Thomas, D. (2024). Diagnostic uncertainty in childhood motor speech disorders: A review of recent tools and approaches. Current Developmental Disorders Reports, 11, 105–112. https://doi.org/10.1007/s40474-024-00295-x
Murray, E., McCabe, P., Heard, R., & Ballard, K. J. (2015). Differential diagnosis of children with suspected childhood apraxia of speech. Journal of Speech, Language, and Hearing Research, 58(1), 43–60. https://doi.org/10.1044/2014_JSLHR-S-12-0358
Shriberg, L. D., Lohmeier, H. L., Strand, E. A., & Jakielski, K. J. (2012). Encoding, memory, and transcoding deficits in childhood apraxia of speech. Clinical Linguistics & Phonetics, 26(5), 445–482. https://doi.org/10.3109/02699206.2012.655841
Valentine, H. C., Keller, S., Harel, D., & Grigos, M. I. (2026). Clinicians’ perspectives on diagnsing and treating childhood apraxia of speech in children under 3 years old. American Journal of Speech-Language Pathology, 35(1), 266–283. https://doi.org/10.1044/2025_AJSLP-25-00215

.png)


Comments