Sedation early after return of spontaneous circulation and during pre-hospital transport after out-of-hospital...

Cardiac arrest is a major medical emergency with survival rates below 10%. Among care components after the return of spontaneous circulation (ROSC), sedation remains poorly standardized. Initial sedation may be withheld from the most severely ill patients in order to facilitate the early identification of neurological recovery.

We analyzed data from the prospective AfterROSC1 and AfterROSC2 cohort studies. We compared post-cardiac arrest patients in whom sedation was vs. was not started during pre-hospital care. Patients' characteristics, day-90 survival, and day-90 survival with a favorable neurological outcome were analyzed (modified Rankin scale 0-3). The day-90 neurological outcome was adjusted for disease severity (modified Cardiac Arrest Hospital Prognosis score [mCAHP]) and for center effect.

Of the 1596 included patients, 1324 (83%) were and 272 (17%) were not sedated before intensive-care-unit (ICU) admission. In both groups, males and cardiac arrest at home predominated. The mCAHP score was higher in the group without pre-ICU sedation (median: 90 [70-106] vs. 105 [90-116], P = 0.001). The pre-ICU sedated group had a higher proportion of patients with a favorable day-90 outcome (32.4% vs. 11.8%; P = 0.001). This difference remained significant after adjustment for the mCAHP score and center (adjusted odds ratio, 2.19; 95% confidence interval, 1.31-3.65; P = 0.003) but not after propensity score adjustment: 1.40 [95%CI: 0.70; 2.09] (P = 0.26).

The decision of sedation started after ROSC and before ICU admission was associated with a good functional recovery on day 90. However, the design of this study does not allow us to establish a causal link between pre-hospital sedation and improved neurological recovery. In particular, it is not possible to formally rule out residual indication bias. A randomized trial assessing the effect of early sedation on treatment strategies and patient outcomes is needed.