BackgroundIntranasal midazolam can produce procedural sedation in frail older patients with dementia who are unable to tolerate necessary medical or dental procedures during domiciliary medical care. Little is known about the pharmacokinetics and pharmacodynamics of intranasal midazolam in older (>65 yr old) people. The aim of this study was to understand the pharmacokinetic/pharmacodynamic properties of intranasal midazolam in older people with the primary goal of developing a pharmacokinetic/pharmacodynamic model to facilitate safer domiciliary sedation care. MethodsWe recruited 12 volunteers: ASA physical status 1–2, aged 65–80 yr, and received midazolam 5 mg intravenously and 5 mg intranasally on two study days separated by a 6 day washout period. Concentrations of venous midazolam and 1ʹ–OH–midazolam, Modified Observer's Assessment of Alertness/Sedation (MOAA/S) score, bispectral index (BIS), arterial pressure, ECG, and respiratory parameters were measured for 10 h. ResultsTime to peak effect of intranasal midazolam for BIS, MAP, and SpO2 were 31.9 (6.2), 41.0 (7.6), and 23.1 (3.0) min, respectively. Intranasal bioavailability was lower compared with intravenous administration (Fabs 95%; 95% confidence interval: 89–100%). A three-compartment model best described midazolam pharmacokinetics following intranasal administration. A separate effect compartment linked to the dose compartment best described an observed time-varying drug-effect difference between intranasal and intravenous midazolam, suggesting direct nose-to-brain transport. ConclusionsIntranasal bioavailability was high and sedation onset was rapid, with maximum sedative effects after 32 min. We developed a pharmacokinetic/pharmacodynamic model for intranasal midazolam for older persons and an online tool to simulate changes in MOAA/S, BIS, MAP, and SpO2 after single and additional intranasal boluses. Clinical trial registrationEudraCT (2019-004806-90).