Objective: To determine whether population density is an independent predictor of survival from out-of-hospital cardiac arrest managed by basic life support (BLS) services using automated external defibrillators (AEDs).
Methods: A retrospective, observational study in Kentucky of 34 BLS services covering 22 counties during the years 1992 to 1994 who used AEDs to treat patients who had out-of-hospital cardiac arrests.
Results: Of 311 patients who had out-of-hospital cardiac arrests, 110 (35%) were defibrillated, 46 (15%) were resuscitated to hospital admission, and 19 (6%) survived to hospital discharge. Univariate predictors for survival to hospital discharge were emergency medical services response interval (from call receipt to ambulance arrival) < 8 minutes, defibrillation by the AED, initial rhythm of ventricular fibrillation or ventricular tachycardia (VF/VT), and population density > 100/square mile (sq mi) for the BLS service area (p < 0.001). A forced logistic regression model of survival to hospital discharge, using these 4 factors plus the presence of a witnessed arrest or bystander CPR, demonstrated that population density > 100/sq mi was highly significant (OR 9.4, 95% CI: 1.7 to 51.4, p < 0.01). Stepwise logistic regression models with combinations of these 6 factors found that survival to hospital discharge was best predicted by an initial rhythm of VF/VT (p = 0.004) and population density > 100/sq mi (p = 0.011).
Conclusions: Population density is strongly associated with survival from out-of-hospital cardiac arrest. BLS services within areas with population densities < or = 100/sq mi sustained little benefit from the addition of AEDs to their treatment of patients who had out-of-hospital cardiac arrests.