Two articles in this issue of The Journal reach apparently conflicting conclusions regarding the safety of the short postpartum hospital stays that are now—and are likely to remain—standard for apparently well mothers and newborns. Both studies, 1 from Washington State1 and 1 from Wisconsin,2used case-control designs and large secondary databases to examine the association between early discharge and neonatal readmissions. Although this question has been explored before, few studies have been so large or have addressed limitations so skillfully. Interpreting the implications of these studies and their seemingly contradictory conclusions raises difficult issues about science and society; the issues are not unique to these studies.
Using linked birth certificate and hospital discharge abstracts from eligible vaginal singleton live births in Washington State from 1991 through 1994, Liu and colleagues2 investigated whether "early" (<30 hours after delivery) compared with "later" (30-78 hours after delivery) discharge increased the risk