Intranasal sedative lifted post-op sleep in 15 trials

Sleep after an operation is often badly disrupted, and poor postoperative sleep is thought to slow recovery. That has made short-term pharmacological support for sleep in surgical patients an active question. Dexmedetomidine is of particular interest because its sedative effect is described as sleep-like rather than simply anaesthetic, and the intranasal route is convenient and non-invasive compared with an infusion. Before it could be considered a reasonable option, though, someone had to pool the scattered randomised trials and ask whether the sleep benefit is real, how large it is, and whether it comes with a safety cost.

The authors ran a systematic review and meta-analysis of randomised controlled trials, registered in advance on PROSPERO. They searched PubMed, the Cochrane Library and Embase up to 20 May 2026 for trials comparing intranasal dexmedetomidine with placebo or other controls in adult surgical patients. Fifteen trials were included. Outcomes covered subjective sleep scores on various scales, objective measures including sleep efficiency, total sleep time, sleep latency and individual sleep stages, the risk of sleep disturbance, and adverse events. Subgroup analyses examined route, timing and duration of administration, patient characteristics, surgical and anaesthesia type, and which rating scale was used.

On subjective scales where a lower score means better sleep, dexmedetomidine improved postoperative sleep (SMD −1.29, 95% CI −1.80 to −0.77), and on scales where higher is better it also came out ahead (SMD 1.75, 95% CI 0.94 to 2.57). Objectively, sleep efficiency and total sleep time increased and sleep latency fell. The risk of sleep disturbance dropped substantially (RR 0.53, 95% CI 0.39 to 0.73). Notably, REM sleep, N1 and N3 sleep did not change, so the architecture of sleep looked largely unaltered. Adverse events including delirium, nausea and vomiting, and anxiety were less common with dexmedetomidine.

The direction of effect is consistent with the idea that dexmedetomidine produces something closer to natural sleep than conventional sedation, and it extends that idea to a simple intranasal route. But the pooled estimates carry very high heterogeneity, and the subgroup analyses point to timing of administration, sex distribution and surgical type as sources of that inconsistency, meaning the average effect may not describe any individual setting well. The unchanged REM and N3 findings complicate a simple 'better sleep' reading. The authors conclude the evidence is limited and call for more high-quality trials.

This is hospital medicine, not a sleep strategy for everyday life: the drug was given around surgery and the outcomes were measured over the following nights. What it does illustrate is that measured sleep has several dimensions — how quickly you fall asleep, how much of your time in bed is actually spent asleep, and which stages you spend it in — and that an intervention can shift some of those while leaving others untouched. It is also a reminder that when pooled trial results disagree wildly with one another, the headline average deserves scepticism until better trials arrive.