Leaving the
higher - risk births in the MANA sample when they are virtually non-existent in the birth
center sample creates tremendous confounders that are not necessarily due to place of birth, but rather to the risks
inherent in the risk factors themselves.