Aim: The aim of this study was to describe the relationship between prehospital clinical variables, interventions and survival time after cardiac arrest and to determine whether various factors affect the risk of death differently at early and late time periods.
Methods: Time-to-death (days) after collapse was identified using the paramedic record, the social security death index or obituaries for adult (> or =18 years) out-of-hospital cardiac arrest (OOHCA) cases from Pittsburgh, Pennsylvania between 1998 and 2002. Clinical prehospital variables included age of patient, sex, witnessed collapse, bystander CPR, use of an automated external defibrillator (AED), initial ECG rhythm, medications and response time intervals. We used Cox regression with time varying coefficients to describe the effects of each covariate separately upon short-term (on day 1) and long-term (after day 1) survival.
Results: Of 1496 adult patients, overall mortality was 89.5%. The majority (75%) of deaths occurred on day 1. Of the 1339 deaths, 1213 (90.6%) occurred by day 3, 1272 (95.0%) occurred by day 7 and 1299 (97.0%) occurred by day 14. Witnessed collapse (hazard ratio 0.85; 95% CI: 0.75-0.97) or the use of epinephrine (adrenaline) (1.57; 1.20-2.07), lidocaine (0.78; 0.66-0.91) or dopamine (0.75; 0.56-1.00) were independently associated with risk of death on day 1 (day of collapse). Epinephrine use (1.84; 1.23-2.78) and age (1.02; 1.01-1.03) were independently associated with risk of death after day 1. The proportional hazards assumption was satisfied.
Conclusions: Survival after out-of-hospital cardiac was characterized by a large number of deaths on day 1. Most subsequent deaths were identified within 14 days after collapse. Prehospital factors have markedly different relationships with short- and long-term survival. Linkage between prehospital intervention and short- and long-term outcomes must consider the survival time characteristics of this population.