TY - JOUR T1 - The wisdom of investigating early developmental impairment JF - Archives of Disease in Childhood JO - Arch Dis Child DO - 10.1136/archdischild-2017-313415 SP - archdischild-2017-313415 AU - Arnab Kumar Seal Y1 - 2017/09/15 UR - http://adc.bmj.com/content/early/2017/09/15/archdischild-2017-313415.abstract N2 - Aetiological investigation of a child with early developmental impairment (EDI) is a common consideration for paediatricians, yet the variation in practice is staggering. It seems straightforward on the face of it but complex in practice. Attempts to standardise such assessments are not new1 2 and the efforts of three teams, Hart et al,3 Best et al4 and Ghosh et al5 explore various aspects of this challenge, and help us take another step forward in making sensible investigative decisions.EDI is commonly deemed to be a developmental trajectory greater than 2 SD below the norm in two or more developmental domains, usually in a child under 5 years of age. Should we investigate all children presenting with EDI for an underlying cause? Not all children are investigated in clinical practice. Hart et al found 8.7% with EDI were not investigated at all. First, let us consider what the utility of such diagnostic investigations would be. Investigations are requested for a variety of reasons. Some to identify a diagnostic cause or provide pointers towards a diagnostic pathway; others help identify comorbid conditions or help monitor the disease process. Hart et al and others previously2 have discussed the reasons why children with EDI should undergo investigations. Occasionally, the rationale for ordering investigations is unclear and attempts to bring an evidence-informed pragmatic approach to this is laudable. Hart and colleagues have done just that and found that the majority of ‘screening’ investigations for EDI do not contribute to a diagnosis. No doubt their efforts will be appreciated by clinicians and managers alike. There are significant cost savings to be made by restricting investigations to ‘useful’ tests with higher yield, only performing the others when specifically indicated.6An important question is when should we not investigate? If there are evident social, environmental … ER -