Early dexmedetomidine and postoperative RASS-defined examinability after microsurgical clipping for aneurysmal...

After microsurgical clipping for aneurysmal subarachnoid hemorrhage (aSAH), sedation should preserve neurological examinability without compromising hemodynamics or nimodipine delivery. We examined whether early postoperative dexmedetomidine was associated with a greater post-landmark RASS-defined examinability.

We retrospectively reviewed adults with aSAH treated with clipping at one center from October 2021 to October 2025. Early dexmedetomidine was initiation within 24 h after clipping and continuation for at least 2 h. Exposure was classified during 0-24 h. The primary endpoint was the time-weighted proportion of observed 24-72 h time in RASS-defined examinability defined without reference to the contemporaneous ordered target (Richmond Agitation-Sedation Scale [RASS] -2 to 0). Ordered-target attainment was the key secondary endpoint. Hour-level states were modeled in consecutive 6-h bins using generalized estimating equations (GEE); target-conditional and shared examinable-target-band analyses tested whether easier prescription targets explained the association. The endpoint hierarchy was finalized after data analysis during peer review; the study was not preregistered, no multiplicity adjustment was applied, and all estimates are exploratory.

Among 123 eligible patients, 120 were alive and classifiable at 24 h; 41 received early dexmedetomidine and 79 did not. Examinable time during 24-72 h was 68% versus 55% (Hodges-Lehmann location shift, 12 percentage points; 95% CI, 5-19), and the GEE-adjusted marginal difference was 11.8 points (95% CI, 4.1-19.5). Early users had less post-landmark deep-sedation time; post-landmark co-sedative exposure, dose, and duration differences were directionally lower but imprecise. Ordered-target attainment was directionally concordant; the primary association remained after adding ordered-target stratum and in the shared-target restriction. Window-specific differences were 15.2 points (95% CI, 6.0-24.4) and 6.3 points (95% CI, -2.4 to 15.0); the exposure-by-window interaction was imprecise (difference-in-differences, 8.9 points; 95% CI, -3.6 to 21.4; p = 0.164). Bradycardia was more frequent (22.0% vs. 7.6%; risk difference, 14.4 points; 95% CI, 1.6-29.7), whereas other safety estimates were imprecise.

The principal finding was greater post-landmark RASS-defined examinability; related care-process and co-sedative measures were not independent validation. Its descriptive concentration during 24-48 h was not confirmed by the formal interaction test and remains hypothesis-generating, with bradycardia as the principal trade-off. The findings remain vulnerable to confounding by indication and do not establish a drug-specific causal effect or downstream clinical benefit.