Discrepancies between pediatric laboratories in pulmonary function results from healthy children.

Background: Multi-center research studies that include pulmonary function as an objective outcome are increasingly important in pediatric respiratory medicine. The need for local controls rather than depending on published normative data for lung function remains debatable.

Aim: To compare pulmonary function in childhood controls with no respiratory symptoms from three centers in the United Kingdom and ascertain the extent to which current reference equations are appropriate for this population.

Methods: Spirometry, plethysmographic lung volumes, and specific airways resistance (sRaw) were measured within specialized pediatric laboratories in children from three geographical locations throughout the UK (London, Leicester, and Glasgow), using identical equipment, software and standard operating procedures. Results were compared between centers and in relation to recent or commonly used UK pediatric reference values.

Results: Pulmonary function was assessed in 94 healthy children (mean (SD) age: 7.7 (0.6) years; 88% white Caucasians; ∼30 from each center). There were no significant differences in background demographics or spirometric outcomes when compared between centers. By contrast, statistically significant differences in plethysmographic lung volumes and sRaw were observed between-centers. Significant differences in relation to published reference data for white subjects were noted for FEV(1) in all three centers and occasionally for other lung function measures but the differences from predicted values were small (within ± 0.5 z-score) and not clinically significant.

Conclusion: After appropriate inter-laboratory standardization, spirometric measurements in children can be measured in different centers without evidence of systematic differences. However, even after extensive standardization procedures, plethysmographic measures appear more prone to inter-center differences and cannot, at present, be reliably compared across centers without incorporating controls at each location.