G391 How useful is head centile shift as a screen for hydrocephalus?. (7th April 2014)

• Record Type: Journal Article
• Authors: Wright, CM; Emond, A
• Is Part Of: Archives of disease in childhood. Volume 99:Supplement 1(2014)
• Page Start: A162
• Page End: A162
• Publication Date: 2014-04-07
• Subjects: Children -- Diseases -- Periodicals; Infants -- Diseases -- Periodicals; 618.920005
• DOI: 10.1136/archdischild-2014-306237.373
• Languages: English
• ISSNs: 0003-9888

Abstract

Abstract : Background and aims: Detection of hydrocephalus via upward centile shifting is one of the main justifications for routine head measurements in infancy. Using a prospective population-based cohort study, we aimed to describe the incidence of head centile shifting, and relate this to the expected number of cases of intracranial expansion (ICE) in the first year. Methods: In the Avon Longitudinal Study of Parents and Children (ALSPAC), head circumference was measured by service health visitors at 2, 9 and 18–24 months. All measures were retrieved from the Avonchild health computer and converted into UK-WHO Z scores. Centile shift was defined as a difference between successive Z scores >1 SD. The expected numbers of cases was modelled using a Norwegian study 1 which identified all cases of ICE with head circumference crossing at least 2 centile lines upwards (approx 1 SD) and found an incidence of 7 per 10, 000 births. Results: There were 9, 269 childen with 3 head measurments, so that 6 cases of ICE would have been expected. In total 18% (967) children shifted >1 SD upwards between 8 weeks and 9 months, but 2/3 of these had a lower centile at 18–24 m, with 148 (15%) dropping >1 SD. Of 324 children who showed further upward shift, only 33 (10%) rose to above the 99.6 th . If all children with continuing upward shifts were further investigated to identify the 6 expected cases, this would give a maximum positive predictive value of only 2%. If investigation was restricted to the children rise crossing above the 99.6 th this might give a PPV of 22%, if all 6 cases were also >99.6 th . Conclusions: Upward centile shifts were very common, but over half were not sustained, suggesting high levels of measurement error. Most stayed within the normal range, suggesting that this represented regression to the mean rather than pathology. The rarity of ICE and high levels of measurement error challenges the value of routine head measurement as a screening procedure. Investigation could be targetted only at those infants with sustained upward centile shifts in head circumference crossing the 99.6 th centile Reference: Zahl SM, Wester K. Pediatrics 2008 Mar 1;121(3):e416-e420.