The Danish Out‐of‐Hospital Cardiac Arrest Trial: A Statistical Analysis Plan
Simon Mølstrøm, Jacob E. Møller, Henrik Schmidt, Anders M. Grejs, Steffen Christensen, Jesper Kjærgaard, Christian Hassager, Martin A. S. Meyer, Lars P. K. Andersen, Bodil S. Rasmussen, Jo B. Andreasen, Sören MöllerABSTRACT
Background
Post‐cardiac arrest care for patients resuscitated from out‐of‐hospital cardiac arrest (OHCA) includes multiple pharmacological and physiological interventions, yet optimal strategies to reduce post‐cardiac arrest syndrome–related morbidity and mortality remain uncertain. The DANOHCA (Danish Out‐of‐Hospital Cardiac Arrest) trial evaluates four such interventions—high‐dose dexamethasone, prophylactic olanzapine, elevated backrest position, and early wakeup—in a factorial trial framework. This manuscript presents the complete statistical analysis plan for analyses common to all four DANOHCA interventions.
Methods
DANOHCA is an investigator‐initiated, multicenter, randomized, 2 × 2 × 2 × 2 factorial clinical trial enrolling comatose adult OHCA patients with a presumed cardiac etiology who achieve sustained return of spontaneous circulation, with randomization within 180 min of return of spontaneous circulation. Participants are co‐randomized to Dexamethasone versus placebo, olanzapine versus placebo, backrest elevation at 35° versus 5°, and early (≤ 6 h) versus late (28–36 h) wakeup, with pharmacological interventions blinded and physiological interventions open‐label. This statistical analysis plan specifies the primary and secondary outcomes, covariate adjustment, handling of missing data, assessments of interactions between interventions, and assumption checks, with all primary analyses conducted according to the intention‐to‐treat principle. The analysis plan is finalized prior to completion of randomization.
Results
The primary outcomes are 90‐day all‐cause mortality for the dexamethasone and backrest interventions, and days alive outside hospital within 30 days for the olanzapine and early wakeup interventions, with common secondary and tertiary endpoints including mortality at later time points, neurological function, health‐related quality of life, organ function, and safety outcomes. Mixed‐effects regression models, Cox regression, stratified nonparametric tests, and prespecified sensitivity analyses will be applied to minimize bias and quantify intervention effects, including their associated uncertainties.
Conclusion
This predefined statistical analysis plan provides a detailed, transparent framework for the primary analyses and reporting of the DANOHCA trial.
Trial Registration: Clinical Trials Information System (CTIS): 2024‐515997‐28‐00