OBJECTIVES: In April 1992, the American Academy of Pediatrics (AAP) recommended that healthy infants be positioned for sleep on their side or back to reduce the risk of Sudden Infant Death Syndrome (SIDS). The authors hypothesized three different forms of the intervention to examine the impact of the recommendation according to theory such as technology diffusion. Seasonality was included in the models to control its effect when testing. METHODS: Box and Tiao time-series intervention methodology was used to examine the effect of the AAP recommendation on SIDS rates. Sudden Infant Death Syndrome mortality data from Philadelphia and Chicago were examined separately for white and nonwhite populations over 32 quarters. RESULTS: Overall SIDS rates dropped significantly according to an abrupt effect from the intervention. However, the effect appeared to be gradually declining in Philadelphia but permanent in Chicago. In Philadelphia, a decline of 62.3% was estimated in whites in the first quarter after the intervention but decreased to only 5% in the last quarter of 1994. A decline of 35.8% was estimated in nonwhites in the first quarter after the intervention but decreased to only 9.4% in the last quarter of 1994. An abrupt and permanent decrease of 26.7% and 16.5% was found in Chicago for whites and nonwhites, respectively. CONCLUSIONS: Evidence of an abrupt adoption of the recommendation can be explained by the authority innovation decision made by the AAP. Some evidence was found that the effect is temporary, perhaps because physicians are reversing earlier decisions. The demonstrated methodology provides a powerful way to test naturally occurring interventions from quasiexperimental designs to test the impact of policy guidelines.
OBJECTIVES: In April 1992, the American Academy of Pediatrics (AAP) recommended that healthy infants be positioned for sleep on their side or back to reduce the risk of Sudden Infant Death Syndrome (SIDS). The authors hypothesized three different forms of the intervention to examine the impact of the recommendation according to theory such as technology diffusion. Seasonality was included in the models to control its effect when testing. METHODS: Box and Tiao time-series intervention methodology was used to examine the effect of the AAP recommendation on SIDS rates. Sudden Infant Death Syndrome mortality data from Philadelphia and Chicago were examined separately for white and nonwhite populations over 32 quarters. RESULTS: Overall SIDS rates dropped significantly according to an abrupt effect from the intervention. However, the effect appeared to be gradually declining in Philadelphia but permanent in Chicago. In Philadelphia, a decline of 62.3% was estimated in whites in the first quarter after the intervention but decreased to only 5% in the last quarter of 1994. A decline of 35.8% was estimated in nonwhites in the first quarter after the intervention but decreased to only 9.4% in the last quarter of 1994. An abrupt and permanent decrease of 26.7% and 16.5% was found in Chicago for whites and nonwhites, respectively. CONCLUSIONS: Evidence of an abrupt adoption of the recommendation can be explained by the authority innovation decision made by the AAP. Some evidence was found that the effect is temporary, perhaps because physicians are reversing earlier decisions. The demonstrated methodology provides a powerful way to test naturally occurring interventions from quasiexperimental designs to test the impact of policy guidelines.