Dreaming Under Anesthesia: A Century of Evidence, Finally Reviewed
Patients have reported dreaming during surgery for over a century, but the phenomenon has long been treated as a clinical curiosity or, worse, confused with intraoperative awareness. A new scoping review from Stanford University, published in the British Journal of Anaesthesia, synthesizes 157 studies spanning 1921 to 2024 and involving nearly 88,000 participants to provide the most comprehensive picture yet of what anesthesia dreaming actually is; and what it is not.
The review, led by Pilleriin Sikka of Stanford’s Department of Anesthesiology, Perioperative and Pain Medicine, finds that anesthesia dreams are surprisingly common, predominantly pleasant, pharmacologically specific, and fundamentally distinct from the terrifying experience of waking up during surgery.
What they found
Frequency varies wildly by context. Dream recall during anesthesia ranged from 40-80% in controlled experimental settings to just 3-8% in large-scale clinical trials. The authors attribute this gap primarily to methodological differences; experimental studies typically interview patients immediately upon emergence, while clinical trials rely on delayed, less systematic recall.
Not all anesthetics are equal. Ketamine was the drug most consistently associated with dreaming (44-100% in experimental settings), followed by propofol (0-77%). The pharmacological specificity suggests that anesthesia dreaming is not a generic sign of light anesthesia but depends on the specific neurochemical action of the drug involved.
Dreams are mostly pleasant. Contrary to the common assumption that anesthesia dreams might signal distress, the review found that dreams were predominantly pleasant, featuring ordinary autobiographical content; daily activities, conversations, familiar places. However, the minority of patients who had unpleasant dreams were less willing to undergo the same anesthetic again, suggesting that dream content matters for patient satisfaction.
Dreaming is not awareness. This is the review’s most clinically important finding. Dream recall was not associated with intraoperative awareness; the rare but traumatic experience of being conscious and paralyzed during surgery. The authors argue that anesthesia dreaming and intraoperative awareness represent different states of consciousness, with distinct neural signatures. Emerging evidence from neuroimaging and electrophysiology suggests that dreaming exhibits brain activity patterns separable from both unconsciousness and connected consciousness.
What predicts dreaming. Three factors consistently predicted higher dream recall: higher home dream recall frequency (patients who dream more at home also dream more under anesthesia), preoperative suggestion or priming (patients told they might dream were more likely to report it), and immediate post-emergence assessment (the sooner patients were asked, the more they recalled).
Major knowledge gaps. The review identified significant gaps in the literature: inconsistent definitions of anesthesia dreaming, inadequate assessment tools, limited mechanistic understanding, sparse pediatric research, and virtually no data on long-term outcomes.
Why it matters
This review reframes anesthesia dreaming from a clinical footnote to a potentially meaningful phenomenon. The finding that dreams are predominantly positive and can be influenced by preoperative suggestion opens the door to therapeutic applications; could pleasant anesthesia dreams be harnessed to improve postoperative mood, reduce anxiety, or enhance recovery? The authors explicitly call for investigation of this possibility.
At the same time, the clear separation between dreaming and awareness should reassure clinicians and patients alike. A patient who reports dreaming after surgery is not signalling a near-miss awareness event; they are describing a distinct, pharmacologically driven state of disconnected consciousness.
Limits
As a scoping review, the study maps the literature rather than performing a formal meta-analysis, so no pooled effect sizes are provided. The included studies span a century of changing anesthetic practices, definitions, and assessment methods, making cross-study comparisons inherently limited. Evidence on pediatric populations and long-term outcomes is particularly thin.
Bottom line
Anesthesia dreaming is common, mostly pleasant, distinct from intraoperative awareness, and pharmacologically specific to drugs like ketamine and propofol. Standardised assessment tools and further research into its therapeutic potential are overdue.
Source
Sikka P, Hu S, Heifets BD. Dreaming during anesthesia: a scoping review. Br J Anaesth. 2026 Jun 11. DOI: 10.1016/j.bja.2026.05.002