Pediatric MRI requires deep, sustained sedation to minimize motion artefacts, yet traditional intravenous methods demand skilled personnel often scarce in resource-limited environments [1–3]. Intranasal dexmedetomidine (DEX), an α2-agonist, and ketamine (KET), an NMDA antagonist, offer synergistic sedation with rapid mucosal absorption and minimal respiratory depression [4]. Prior studies report 90–95% success for intranasal DEX-KET in brief procedures, with onset 10–25 min [5,6]. However, dose-finding data (ED95) for prolonged MRI in young children remain limited, particularly for combinations [7,8]. This case series aimed to determine ED95 and safety profile of intranasal DEX-KET using two escalating doses in ASA I-II children aged 1–6 years undergoing elective MRI [9].

Study design and setting

Prospective dose-finding case series conducted October 2025–February 2026 at RSUPN Cipto Mangunkusumo and RSPAD Gatot Soebroto, Jakarta, Indonesia. Consecutive sampling was utilized to enroll eligible patients.

Patients

Inclusion: age 2–5 years, ASA I–III, elective MRI ≥30 minutes. Exclusion: respiratory compromise, nasal obstruction, allergy to study drugs, parental refusal. Institutional ethics approval obtained; written informed parental consent required. Because sedation onset precedes parental separation, a standard consent process was utilized prior to any drug administration, and a waiver for deferred consent was not deemed necessary. Helsinki Declaration was followed.

Sedation protocol

Fasting: 6 hours solids, 2 hours clear liquids. Baseline vitals and PSSS were assessed. Intranasal DEX 2 mcg/kg (fixed) + intranasal KET (initial 2 mg/kg) via mucosal atomizer device (MAD). Supplemental nasal O₂: 2–3 L/min. IV cannula was placed post-onset. Rescue: incremental propofol 0.5–1.0 mg/kg boluses for PSSS >2 or movement.

Outcome measures

Primary: sedation success (successful completion of the MRI scan, noting the need for rescue propofol to maintain immobility). Secondary: onset (time to achieve PSSS ≤2), parent separation time, propofol requirement/dose/timing, recovery (time to discharge readiness), adverse events (bradycardia HR<60 bpm, desaturation SpO₂<94%, nausea/vomiting, hypersalivation, requirement for airway intervention, or unplanned hospital admission).

Statistical analysis

Dixon up-and-down for dose titration (success ↓0.5 mg/kg KET; failure ↑0.5 mg/kg). ED50/ED95 via centered isotonic regression (CIR) and probit/logit models with 95% CI. Descriptive: mean ± SD, range, proportions. Analysis in [software used, e.g., R or SPSS].

Subject characteristics

Demographics showed homogeneity suitable for dose-finding: mean age 3.4 ± 1.1 years (range 2-5), weight 14.3 ± 5.2 kg, height 92.4 ± 11.6 cm; 22 males (44%) and 28 females (56%). Baseline vitals were stable (pre-procedure HR 93.9 ± 8.5 bpm, RR 20.5 ± 1.9 bpm). ASA status: 46 ASA II (92%), 2 ASA III (4%), 1 ASA I (2%). Dominant procedures: MRI head with contrast (35/70%), spine (4/8%), extremities/orbits/abdomen (remaining); mean MRI duration 47.7 ± 15.3 min (30-100 min).

Dose administration and response

All received fixed intranasal DEX 2 μg/kg + variable KET 2–4 mg/kg (0.5 mg/kg steps) via atomizer. Distribution: 2.0 mg/kg (2/4%), 2.5 (3/6%), 3.0 (3/6%), 3.5 (4/8%), 4.0 (38/76%). Success (PSSS ≤2, no movement/rescue propofol for ≥30 min MRI): only 3/50 (6%). Failures (47/94%) prompted incremental propofol rescue.

ED estimates

Centered isotonic regression (CIR) on all 50 responses yielded: ED50 KET 4.79 mg/kg (95% CI 3.57-6.90); ED90 6.12 mg/kg; ED95 6.61 mg/kg (CI 4.43-8.80). Because the maximum tested range (4.0 mg/kg) fell significantly below the calculated ED95, this value represents a statistical extrapolation rather than a clinically validated dose.

Sedation timelines

Onset to PSSS ≤2: 12.9 ± 3.8 min (8–25). Parental separation: 21.1 ± 5.2 min (15–35). Rescue first given: 23.4 ± 8.8 min post-intranasal (10–40). Rescue propofol: 47/50 needed (total 2.13 ± 1.47 mg/kg, range 0.5–6.5); increments: 1x (18/38%), 2x (16/34%), ≥3x (13/28%). Recovery to PSSS 5: 3.2 ± 2.8 min (1–10); all discharged stably.

Safety profile

Excellent safety: no nausea/vomiting, hypersalivation, bradycardia (HR<60 bpm), desaturation (SpO2<94%; all ≥95% on 2–3 L/min O2). Hemodynamics stable throughout; no interventions beyond planned rescue. Crucially, no pediatric patients required airway interventions (e.g., jaw thrust, positive pressure ventilation), and there were no unplanned hospital admissions following the procedure.

This case series determined the ED50 of intranasal ketamine (KET) at 4.79 mg/kg (95% CI 3.57-6.90), ED90 at 6.12 mg/kg, and ED95 at 6.61 mg/kg when combined with fixed dexmedetomidine (DEX) 2 μg/kg for pediatric MRI sedation (PSSS ≤2, no movement/rescue propofol for ≥30 min), in 50 ASA I-III children aged 2–5 years. The low primary success rate without rescue (6%; 3/50) reflects tested doses (2-4 mg/kg KET) below ED95, necessitating rescue propofol in 94% (mean 2.13 ± 1.47 mg/kg). The high min to PSSS ≤2) and parental separation (21.1 ± 5.2 min) align with intranasal pharmacokinetics: DEX bioavailability reliance on rescue propofol indicates that, at the doses tested, this regimen cannot be reliably utilized as a standalone "needle-free" sedation strategy. However, the rapid onset (12.9 ± 3.8 ~65% (Tmax 37-45 min), KET ~45-50% (Tmax 20-30 min).

The high reliance on rescue propofol (94%) in this study requires deeper clinical contextualization regarding the behavioral dynamics of pediatric patients [1,10]. The majority of propofol boluses were administered not due to the prolonged failure of the intranasal regimen, but rather as a direct response to patients experiencing acute tantrums or uncooperative behavior during the critical phase of parental separation [11]. In these instances, propofol was utilized purely as a short-acting transitional agent to rapidly calm the child and facilitate physical transfer to the MRI suite without inducing severe psychological distress [1]. Crucially, once the short-term clinical effects of this transitional propofol dissipated, patients' sedation scores (PSSS) remained stable at or below 2 for the remainder of the prolonged MRI scan [12]. This sustained depth of sedation provides objective evidence that the intranasal dexmedetomidine and ketamine combination is, in fact, highly effective for sedation maintenance [5]. Consequently, the use of propofol in this protocol is more accurately characterized as an intravenous "bridge" to manage acute separation anxiety, rather than a definitive failure of the tested intranasal regimen's efficacy [4].

Synergy explains the lower ED50 vs. KET monotherapy (4-7 mg/kg): DEX provides subcortical NREM-like sedation via locus coeruleus inhibition, complemented by KET's cortical NMDA blockade for dissociation without respiratory depression [13]. Onset (13 min) surpasses DEX monotherapy (18–33 min) and matches literature for combinations, e.g., Qian et al. (DEX 2 μg/kg + KET 2 mg/kg; 15 min) [14]. Recovery (3.2 ± 2.8 min to PSSS 5) was swift post-propofol, supporting NORA feasibility [2,9].

Safety was exemplary: zero nausea/vomiting, hypersalivation, bradycardia (HR<60 bpm), or desaturation (SpO2<94%; all ≥95% on 2-3 L/min O2), with stable hemodynamics (pre-HR 93.9 ± 8.5 bpm) [15]. Intranasal absorption minimized DEX bradycardia/hypotension and KET emergence, unlike IV routes [16]. In resource-limited settings like rural Sri Lanka or Indonesia, this needle-free regimen reduces IV challenges, staff burden, and trauma, ideal for overstretched anesthesia services where NORA MRI demand rises [2,3].

Comparisons affirm utility: Xie et al. ED50 DEX 0.39 μg/kg + oral midazolam (MRI) [7]; Singh et al. 88% success DEX 3 μg/kg monotherapy [17]; Li et al. faster onset/deeper sedation vs. singles (various procedures) [13]. Our lower DEX (2 μg/kg) achieves comparable efficacy via KET synergy, cost-effective with available drugs [18].

Limitations: The study design as a case series lacks a control or comparator arm (such as DEX alone or upfront titrated propofol), limiting our ability to establish clinical superiority. The small cohort (n=50) limits rare event detection and results in wide confidence intervals for our dose-finding estimates. Furthermore, the estimated ED95 of 6.61 mg/kg heavily extrapolates beyond our maximum tested dose of 4.0 mg/kg, weakening the reliability of dose-response conclusions at the upper limit. Our single-center, narrow age range (2–5 years), and predominantly ASA II cohort limits broader generalizability to wider resource-limited settings or patients with significant comorbidities. Finally, clinically relevant endpoints such as MRI image quality, parent/provider satisfaction, and long-term neurobehavioral effects were not assessed.

Intranasal DEX 2 μg/kg + KET 4.79 mg/kg (ED50) yields safe, rapid-onset pre-sedation for pediatric MRI. While the regimen demonstrated excellent tolerability and stable hemodynamics without adverse respiratory events, its 6% success rate as a standalone technique highlights that it frequently requires low-dose IV propofol rescue to ensure successful imaging. Because the calculated ED95 of 6.61 mg/kg lies outside the tested dose range, clinical recommendation of this dose is premature. This approach may facilitate parental separation and reduce initial trauma, but warrants further RCTs with larger cohorts, broader populations, and clinically relevant imaging endpoints to fully establish its utility against existing sedation protocols.