JUSTIFICATION: Automatic crash notification (ACN) systems using crash sensor and communications technology are being developed, but their potential effect is uncertain.

OBJECTIVE: To estimate reduction in mortality expected from an ideal ACN system.

METHODS: 1997 FARS data from 30875 cases of incapacitating or fatal injury with complete information on EMS notification and arrival times were analyzed considering cases in four compartments: 1) Alive prior to notification, 2) Alive after notification, 3) Alive after arrival, and 4) Dead.  For each minute after crash, transition probabilities were calculated for each possible change of status.  These data were used to construct a compartmental model with 1) Incapacitating injuries equal to 1997 GES estimates, 2) Transitions to death from other compartments equal to FARS totals and proportional to FARS rates, and 3) Other transition rates equal to FARS rates. The outcomes from this model were compared to those from an otherwise identical model in which all EMS notification times were set to 1 minute.

RESULTS: The model conformed to FARS estimates of 12833 deaths prior to notification, 1805 after notification, and 14000 between EMS arrival and 6 hours.  After notification times were set to 1 minute, the model predicted 9577 deaths prior to notification, 2267 after notification, and 15118 after arrival.  The overall mortality at 6 hours was reduced from 28638 to 26962 (1676 per year, or 6%).

SIGNIFICANCE: Modest but important reduction in mortality should be expected from a perfect national ACN system.  Imperfect systems would be less effective.