MCA
Of 22 variables included in MCA (Table-2), the top three dimensions (of
a possible 39) were chosen as they represented the largest number of
independent variations between each dimension (Table-2). The top 4
variables (total n=12) from each dimension (e.g. LOS, low birth weight,
preterm, weight-for-length z-score, supplemental oxygen requirement,
accessory
muscle use, RSV, HRV, any bacteria, any co-morbidity, bronchiectasis and
previous respiratory hospitalisation) were then selected for LCA,
including as many variables included in Dumas and colleagues’
study3.