Enrollment and Randomization
The first site commenced the run-in phase in June 2007. All the sites stopped enrollment in November 2009, when the data and safety monitoring board recommended that the trial be stopped early because continuing recruitment was unlikely to change the outcome of the study. Of 13,460 patients screened, 10,365 were enrolled, and 10,153 underwent randomization. Of these, 195 were excluded from the data analysis when their cardiac arrest was confirmed to be due to drowning, strangulation, or electrocution, and 25 were excluded because the outcome with respect to the primary end point was unknown. Thus, 9933 patients were included in the primary data analysis.
Characteristics of the Two Study Groups
The early-analysis group comprised more patients than the later-analysis group (5290 vs. 4643) owing to early termination of the trial. The two study groups were evenly balanced with respect to baseline characteristics except that there were small group imbalances in the distribution of patients across sites; however, these would not have any appreciable effect on the results because of the cluster-crossover design, which yields treatment comparisons within clusters. Not all the scheduled cluster crossovers had occurred at the time of termination, although each cluster had crossed over at least once.
The median time to the analysis of cardiac rhythm was 42 seconds (interquartile range, 27 to 80) in the early-analysis group and 180 seconds (interquartile range, 151 to 190) in the later-analysis group. A majority of patients in each group received rhythm analysis within the targeted range for that group: 68% of patients in the early-analysis group received analysis of cardiac rhythm within the targeted range of 0 to 60 seconds and 60% of patients in the later-analysis group received analysis of cardiac rhythm within the targeted range of 150 to 210 seconds.
Primary and Secondary Outcomes
A total of 310 patients in the early-analysis group (5.9%) and 273 patients in the later-analysis group (5.9%) survived to hospital discharge with a modified Rankin score of 3 or less, with a cluster-adjusted difference between later cardiac analysis and early cardiac analysis of −0.2 percentage points. There was also no significant difference between the study groups with respect to any of the secondary outcomes. An analysis adjusted for potential confounders evaluated the effect of study group on survival and showed a difference of −0.3 percentage points (95% CI, −1.3 to 0.7) between later cardiac analysis and early cardiac analysis (P=0.61).
When the outcomes were analyzed on an as-treated basis, the rates of survival with satisfactory functional status were 6.0% among the 3982 patients in whom the analysis of cardiac rhythm was performed between 0 and 60 seconds and 5.9% among the 3115 patients in whom the analysis of cardiac rhythm was performed between 150 and 210 seconds (P=0.97). In an additional exploratory analysis, we evaluated the rate of survival as a function of the actual time to the first rhythm analysis, regardless of the study group. The chance of survival with satisfactory functional status did not improve with increasing time to the first analysis of cardiac rhythm, and among patients with an initial rhythm of ventricular tachycardia or ventricular fibrillation who received CPR from a bystander, the rate of survival tended to decline with increasing time to the first rhythm analysis.