Evaluation of Intensive Care Unit and Ventilator-free Days as Alternative Outcomes in Out-of-Hospital Cardiac Arrest Trials

Objective: Compared with traditional out-of-hospital cardiac arrest (OHCA) trial outcomes such as hospital survival, alternative outcomes such as intensive care unit-free (IFD) and ventilator-free days (VFD) have potential advantages including requiring smaller sample sizes to detect significant differences. Few studies have evaluated these outcomes in OHCA. We sought to evaluate the utility and validity of IFD and VFD as candidate outcomes for OHCA trials. Methods: We analyzed data from the Pragmatic Airway Resuscitation Trial (PART), which tested laryngeal tube (LT) vs. endotracheal intubation (ETI) airway strategies in adult OHCA. We defined IFD as the number of days alive and permanently out of ICU during the first 30 days after the OHCA event. We examined IFD distribution and correlation with Modified Rankin Scale (MRS). To test associations with trial interventions, we applied a range of analytic strategies, including modeling IFD as continuous (Generalized Estimating Equations - GEE), non-parametric (Wilcoxon Rank-Sum test - WRS) and count (zero-inflated negative binomial regression - ZINB) data. We also modeled time-to-ICU discharge using a survival model with death as a competing risk. We repeated the analysis assessing VFD. Results: IFD was available for 2,898 of 3,004 patients. IFD was skewed and J-shaped; 91% IFD=0 and 5.3% IFD≥22. Mean (±SD) IFD was 2.0±6.6 days; 21.2±7.5 days for survivors. IFD varied by intervention (LT 2.4±7.2, ETI 1.6±5.8 days). IFD and VFD strongly correlated with MRS (ρ= -0.88, -0.89). LT was associated with increased IFD using GEE, WRS, and ZINB, but not competing risks model. (Table) LT was associated with increased VFD using GEE and WRS, but not ZINB or competing risks models. Conclusion: IFD and VFD differentiated OHCA interventions, suggesting their utility. IFD and VFD were highly correlated with MRS, suggesting validity. IFD and VFD have important features that may influence OHCA trial design.