Abstract
It is often difficult to know how far to investigate children with delayed development, and the paediatrician may feel frustrated by the low yield of the usual investigations. Yet the benefits of diagnosis are sufficiently great that few would quibble with the need for some investigations. It would be best if clinical signs directed the testing, but in practice most retarded children provide no clues to the direction in which investigation should go. This may be true, for example, of the young child with X-linked mental retardation with fragile X chromosome--Sanfilippo syndrome--one of the rare inborn errors of metabolism or a brain malformation like agenesis of the corpus callosum. Testing is, therefore, most often performed in a screening fashion with few guidelines as to when to stop. Most would agree with a karyotype and urine amino acid screen but are urine mucopolysaccharides, cerebral ultrasound or CT scan, CPK in boys and thyroid function tests necessary? The decision will, in general, depend on the individual paediatrician's training, the clinical environment in which he or she works and past diagnostic successes and failures. The final element is the ease and cost of testing. Laboratories may encourage screening when the tests can be done in batches in an automated fashion, while demanding clear clinical indications when technologies are costly or labour intensive. Recent recognition of the extent of the peroxisomopathies, mitochondrial cytopathies and some inborn errors of neurotransmission has increased the range of investigations that needs to be considered. At present the disorders are still being defined by their more severe variants where clinical features can direct investigation.