The articles by Grote et al,1 and Fayter et al,2 in the March and April issues of Archives of Disease in Childhood, respectively, are important contributions to the thorny issue of growth monitoring as a community-wide screening tool. However the leading article on this theme in the April issue by Tam Fry, which criticises current UK policy developed at the “Coventry Consensus” on growth monitoring, and the “Atom” presented by the Editor, are misleading.

That Consensus meeting, contrary to what Fry implies, did indeed produce a consensus – as far as we know, just two out of over thirty members dissented. We have compared our recommendations with the evidence presented by Grote et al and Fayter et al. Grote et al found that length/height monitoring in children under the age of three performed poorly in identifying growth disorders; this is in keeping with the report of Mei et al3 regarding the high incidence of centile crossing in very young children. They also observed that centile crossing (the “height deflection” rule in their paper) made little contribution to case finding. Fayter et al in their systematic review noted that virtually all the published studies were based on single measurements and were unable to identify evidence that serial measurements were significantly more productive. All these points support the views of the consensus group.

We do agree, however, that the issue of adjusting the school entry height measurement for parental height deserves to be re-visited, although the practical difficulties are downplayed by Grote et al. Estimated heights are unreliable,4 so actual measured heights are vital, but the parents are often unavailable. Even in the well-documented series of girls with Turner’s syndrome,5 included in Grote et al’s analysis, parent heights were missing in over half the cases. The correction process, which looks easy to paediatricians, may be challenging to many of the staff who undertake screening unless we can automate it. Nevertheless, such an adjustment might improve the sensitivity of the screen with negligible reduction in specificity.6

The Editor’s Atoms comments that the yield of growth monitoring would be greater than reported in these two papers for two reasons. First, children with idiopathic short stature would be identified and referred for growth hormone treatment. Treating this group is controversial – the gains in height are modest, the health-related quality of life may not be improved7 and the cost is estimated at 20 000 Euro per cm of height gained.8 Second, height monitoring for celiac disease was shown by Grote et al to be a poor tool for identifying cases – if we do want to screen for the “sub-clinical” cases that are more common in older children than in infancy, there are better ways of doing it.9

Growth monitoring of all children is difficult in our healthcare system and becoming more so as primary care fragments. In the UK, coverage (percentage of eligible children measured) is often low and measuring is often badly done (though we hope not as badly as depicted on the cover of the April Archives of Disease in Childhood). Quality assurance and clear referral pathways are vital, as in any screening programme. We have stated previously that it is better to do a few things well than many things badly.

Does the challenge of childhood obesity call for a change in the Coventry Consensus recommendations? Country-wide anonymised BMI data collection is a Government-led initiative that is expected to facilitate monitoring of the obesity “epidemic” and of public health interventions; but there is no evidence that the proposed feedback of individual BMI results to parents will have any effect on the obesity epidemic.10 The label of overweight or obese may indeed be harmful and stigmatising, particularly as interventions are unproven and treatment resources are inadequate. Nevertheless, if political pressures demand that such a programme is to be introduced, the opportunity should be seized to implement the straightforward recommendation of the Coventry Consensus, to undertake an accurate school entry height measurement and act on it where appropriate.