This report examines the magnitude and impact of changes in sleep position and sudden infant death syndrome (SIDS) before and after the `Back to Sleep'(BTS) campaign. 751 SIDS cases in Philadelphia from 1987-1996 were compared to 857 controls from the same period matched for important demographic characteristics, including race, gestational age, sex, and maternal age. Groups were compared before and after the July, 1994 initiation of BTS, using student's t-test and chi-square, with significance at p < 0.05.

Currently, 70% of infants usually sleep non-prone (back or side), an increase from 31% prior to BTS (p<0.05). SIDS infants found non-prone have also increased, from 20.6% to 33% (p<0.05). During the same period, SIDS infants co-sleeping at the time of death remained unchanged (46% to 50%, p=ns). This is consistent with the 46% co-sleeping rate in the control group during both periods. Before and after BTS, >90% of all SIDS occurred at home.

Though 70% of infants in Philadelphia are put to sleep non-prone, only 20% of SIDS cases are found that way, suggesting an important causal role for prone position in SIDS. As non-prone sleep increases, more infants have died on their backs or sides.This suggests the existence of other important factors contributing to SIDS risk. In addition, the data do not implicate co-sleeping as a risk factor for SIDS.