Background Newborn screening (NBS) for cystic fibrosis (CF) improves nutritional outcomes. Despite early dietetic intervention some children fail to grow optimally. We report growth from birth to 2 years in a cohort of children diagnosed with CF by NBS and identify the variables that influence future growth. Methods One hundred forty-four children were diagnosed with CF by the West Midlands Regional NBS laboratory between November 2007 and October 2014. All anthropometric measurements and microbiology results from the first 2 years were collated as was demographic and CF screening data. Classification modelling was used to identify the key variables in determining future growth. Results Complete data were available on 129 children. 113 (88%) were pancreatic insufficient (PI) and 16 (12%) pancreatic sufficient (PS). Mean birth weight (z score) was 3.17 kg (− 0.32). There was no significant difference in birth weight (z score) between PI and PS babies: 3.15 kg (− 0.36) vs 3.28 kg (− 0.05); p = 0.33. By the first clinic visit the difference was significant: 3.42 kg (− 1.39) vs 4.60 kg (− 0.48); p < 0.0001. Weight and height remained lower in PI infants in the first year of life. In the first 2 years of life, 18 (14%) infants failed to regain their birth weight z score. The median time to achieve a weight z score of − 2, − 1 and 0 was 18, 33 and 65 weeks respectively. The median times to reach the same z scores for height were 30, 51 and 90 weeks. Birth weight z score, change in weight z score from birth to first clinic, faecal elastase, isolation of Pseudomonas aeruginosa, isolation of Staphylococcus aureus and sweat chloride were the variables identified by the classification models to predict weight and height in the first and second year of life. Conclusions Babies with CF have a lower birth weight than the healthy population. For those diagnosed with CF by NBS, the weight difference between PI and PS babies was not significantly different at birth but became so by the first clinic visit. The presence of certain factors, most already identifiable at the first clinic visit can be used to identify infant at increased risk of poor growth.

IntroductionCystic Fibrosis (CF) newborn screening (NBS) was implemented across the UK in 2007. It has been associated with improved clinical outcomes particularly related to nutrition. We reviewed the nutritional progress of infants diagnosed with CF by NBS in the West Midlands. Our aim was to develop a model for predicting height and weight in the first 2 years of life based on information available at the first clinic visit.MethodsAnthropometric data is recorded at each outpatient visit for children with CF. This data was reviewed in conjunction with the CF NBS data for all children diagnosed with CF in the West Midlands between November 2007 and October 2014. Cluster analysis, classification and polynomial regression modelling were used to analyse these data. Models were validated using the 5-fold cross validation method.Results144 children were identified with CF at a mean age of 22 days. There was no difference in birth weight z scores between pancreatic insufficient and pancreatic sufficient children (−0.05 vs −0.36, p = 0.29) however a significant difference was observed in rate of weight gain from birth to first clinic visit (−0.1 vs −0.33, p = 0.007). Time taken for children to reach a z score of 0 for weight was 65 weeks and length was 90 weeks. Cluster analysis identified two distinct groups of children. Faecal elastase (FE) being the main determinant of class, with a cut off of 212 μg/g. Our models can predict weight z score at 1 and 2 years with a mean absolute error of 0.51 and 0.67 and length z scores at 1 and 2 years with an accuracy of 0.7 and 0.85. The most important factor when predicting future nutritional parameters was birth weight z score.ConclusionsWe have developed and validated models that can provide a good estimate of weight and height z scores in the first 2 years of life for children diagnosed with CF by NBS. These models only require data available at the first clinic visit. They can potentially be used by clinicians to identify children at risk of poor nutritional outcomes thus, encouraging closer monitoring and earlier intervention.Abstract P89 Figure 1Schematic representation of the accuracy of the model predicting weight z score at age 1

No abstract