care and for whom NEC was not suspected in this period. This reduction enables us to use summaries of the feeding regimen in this period as background variables in a potential outcomes framework. Large size of the cohort is a distinct advantage of our study. Its results inform the design of a randomised clinical trial for preventing NEC, and the choice of its active treatment(s) in particular.
Nicholas T. Longford
Statistics in Transition New Series , ISSUE 1, 87–117