Article Text

Download PDFPDF

Comparing cystic fibrosis outcomes across the pond
  1. David C Taylor-Robinson1,2,
  2. Michael S Schechter3,
  3. Rosalind L Smyth2
  1. 1Department of Public Health and Policy, University of Liverpool, Liverpool, UK
  2. 2UCL Institute of Child Health, London, UK
  3. 3Department of Pediatrics, Division of Pulmonary Medicine, Virginia Commonwealth University, Children's Hospital of Richmond at VCU, Richmond, Virginia, USA
  1. Correspondence to Dr David C Taylor-Robinson, Department of Public Health and Policy, Whelan Building, University of Liverpool, Liverpool L69 3GB, UK; dctr{at}

Statistics from

Analysis of registry data has provided key insights into the changing demographics, outcomes and treatments in cystic fibrosis (CF), but few studies have made use of the great potential for cross-country comparisons. The linked study by Goss and colleagues1 in Thorax does just this, using registry data collected in the UK and the USA to compare CF outcomes and use of treatments on opposite sides of the Atlantic. The study certainly makes for uncomfortable reading from a UK perspective, but raises more questions than it addresses. The analysis suggests that the USA does better in terms of lung function in children, and the authors conclude this is due to more intensive treatment in the early years. But is this interpretation correct?

One of the most important findings of the study is the striking gap in mean % predicted FEV1 between the UK and the USA at around 6 years, the age at which lung function can be consistently collected and measured reliably at all care centres. The gap subsequently narrows and disappears by age 30. Overall, children under 12 years of age in the UK had significantly lower lung function by 7.6 percentage points compared with children in the USA. This is a big difference in a study where the large sample sizes mean that the estimations of population level differences are quite precise, even for imprecisely measured outcomes such as % predicted FEV1.2

In these initial attempts to compare data collected in slightly different ways in different countries, it is important to ensure that the comparisons are valid. For example, we must ensure that the findings of this analysis are not biased by the possibility that the samples used in the study are not fully representative of the entire CF populations in their respective countries. Ascertainment …

View Full Text

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles