MMBGR protocol - diagnostic accuracy of clinical examination in preschoolers.

There is no validated reference standard for orofacial myofunctional disorder, so the comparison was against the consensus judgement of three specialists rather than a confirmed diagnosis, and the authors acknowledge that the absence of validated severity instruments may affect the accuracy estimates. Both the index test and the reference standard were made from stored photographs and videos rather than from the children themselves. The authors say the image method does not always allow the best view of a domain, and that judging tonus without direct palpation is a particular problem, so accuracy for tonus in younger children is likely underestimated. The children in the database had no speech-language complaints and no neurological or developmental diagnoses, so the study does not tell you how the protocol performs in the referred, symptomatic children a clinician usually sees. Sampling was by convenience rather than random, and the authors recommend future studies with probability sampling. Fourteen speech-language pathologists had to be used because two dropped out and had to be replaced partway through; as a result, the cases were not reassessed by the same rater and intra-rater agreement could not be checked for the reference standard. Accuracy was only fair in most domains, and the authors recommend cautious clinical use until multicentre studies in other populations are done.

The authors declare no financial support and nothing to declare.

Do not read the high sensitivity as meaning the protocol can clear a child. Specificity was non-ideal in almost every domain, and at 24 to 35 months it was 25.0% for orofacial structures and 0% for orofacial myofunctional disorder, meaning most children the specialist panel judged unaffected still scored above the cutoff. Use a score below the cutoff as one piece of information, not as proof there is nothing wrong.