Rescue propofol at emergence is associated with reduced severe emergence agitation in children undergoing dental general anesthesia: a retrospective cohort study

Background Emergence agitation occurs in up to 80% of children following sevoflurane anesthesia; universal prophylaxis exposes non-susceptible patients to unnecessary sedation. A selective rescue approach—delivering intervention only when dysregulation signs first manifest—may offer an alternative. Methods In this single-center retrospective cohort study (May 2024–February 2026), children aged 4–12 years (ASA I–II) with a Watcha Agitation Scale score of 3 or 4 at PACU arrival received rescue propofol (1 mg/kg IV; n = 122) or no immediate pharmacological intervention ( n = 123) at clinician discretion; children with scores of 0–2 ( n = 205) formed a reference cohort. Because treatment was not randomized and the groups differed substantially at baseline, confounding by indication was expected. The primary outcome was severe emergence agitation (peak PAED ≥ 12 at 5, 15, or 25 min). Modified Poisson regression with robust variance provided the primary adjusted estimate, with complementary propensity-score sensitivity analyses including matching, inverse-probability weighting, and restriction to empirical common support. Results Severe emergence agitation occurred in 30% of the propofol group versus 77% in controls (RR 0.39, 95% CI 0.30–0.52; risk difference − 46.9% points; p < 0.001). The primary adjusted estimate, from propensity score–adjusted modified Poisson regression, was RR 0.33 (95% CI 0.24–0.45; p < 0.001); the propensity score–adjusted odds ratio (OR 0.07, 95% CI 0.03–0.18; p < 0.001) is retained as a sensitivity comparison. PAED scores were significantly lower at all time points (all p < 0.001). The Watcha Agitation Scale demonstrated sensitivity 84.8% (95% CI 78.2–91.5%) and specificity 87.0% (82.6–91.5%). No serious adverse events occurred. Conclusions Receipt of clinician-initiated rescue propofol was associated with a lower observed risk of severe emergence agitation, but the observational design, marked baseline imbalance, single unblinded assessor, residual sedation, and limited propensity-score overlap preclude causal interpretation. The time-dependent PAED domain pattern was compatible with an early sedative contribution. The Watcha Agitation Scale findings are exploratory internal classification characteristics and require prospective external validation. Trial registration Not applicable. This was a retrospective observational study and was not a clinical trial.