Level 2: Mid-level evidence
For out-of-hospital cardiac arrest, a clinical prediction rule recommends termination of resuscitation when three criteria are met: (1) no return of spontaneous circulation before transport to the hospital, (2) no shock given before transport to the hospital, (3) arrest not witnessed by emergency medical services (EMS) personnel.
This rule was evaluated in a prospective study of 1,240 consecutive adults treated for an out-of-hospital arrest by one of 24 emergency medical systems in Ontario, Canada, between 2002 and 2004 (N Engl J Med. 2006;355:478-487).
The prediction rule recommended termination of resuscitation in 776 (62.5%) patients; there was no return of spontaneous circulation in 1,172 (94%), no shocks delivered in 868 (70%), and cardiac arrest was not witnessed by EMS personnel in 1,120 (90.3%). In all, 1,140 (91.9%) patients were pronounced dead in the emergency department, 59 (4.8%) died after admission to the hospital, two (0.2%) were still in the hospital at six-month follow-up, and 39 (3.1%) survived to discharge. Only four (0.5%) patients survived out of the 776 who met the prediction-rule criteria. The clinical prediction rule had 64.6% sensitivity, 90.2% specificity, 99.5% positive predictive value, and 8% negative predictive value for death before hospital admission.
Data forms were not completed for 380 of 1,620 eligible patients (24.5%), so there were insufficient data to determine survival among missing patients and how that might affect results.