G483(P) Predicting healthcare outcomes in prematurely born infants using cluster analysis

Aims Infants born prematurely are at increased risk of RSV lower respiratory tract infection (LRTI) and its consequences. We aimed to determine which prematurely born infants were at risk of rhinovirus (RV) LRTI and whether their outcomes differed according to risk factors. Methods Birth weight, gestational age, duration of mechanical ventilation and supplemental oxygen and functional residual capacity and respiratory system compliance and resistance measured at 36 weeks postmenstrual age in 168 prematurely born infants were used to classify them into clusters using hierarchical agglomerative clustering with the Euclidean distance metric. All LRTIs in the first year after birth (in hospital or the community) were recorded and their NPAs were tested for 13 respiratory viruses. Fifty-six children were followed up at a median (IQR) age of 7.0 (6.4–7.3) years and healthcare utilisation costs from age one to the point of follow-up were determined. Results The infants could be classified with 100% accuracy into three clusters. Cluster three compared with Cluster two and Cluster two compared with Cluster one were of significantly lower gestational age, birthweight, had a longer duration of mechanical ventilation and supplemental oxygen and had worse lung function at 36 weeks PMA At one year of age, Cluster three had had a higher incidence of RV LRTI (p At school age, healthcare utilisation costs for respiratory (p=0.008) and non-respiratory (p=0.003) causes were higher in Clusters two and three than in Cluster one (see Table). Conclusion Extremely prematurely born infants who require prolonged ventilation and have poorer premorbid lung function are at increased risk of RV LRTIs and have higher healthcare utilisation costs even at school age.