Benefits in adults allowed to drink clear liquids before anaesthesia until called to the operating room

Citation: Haas TE, Kranke P, Stegemann MN, Helmer P, Schmid B, Diehl FM, et al. Benefits in adults allowed to drink clear liquids before anaesthesia until called to the operating room: A randomised pilot study. Eur J Anaesthesiol. 2025;43:207–216. doi:10.1097/EJA.0000000000002309.

Study at a glance:

- Design and setting: Prospective, three-arm, randomised controlled pilot trial at a single German academic centre (University Hospital Würzburg) in 174 adult ASA I–III patients (mainly gynaecology, trauma and urology) undergoing elective or less-urgent surgery with low aspiration risk; web-based 1:1:1 randomisation, single-blind (blinded outcome assessors), intention-to-treat analysis with 0% dropout; overall RoB 2: “some concerns” mainly due to unblinded patient-reported outcomes.

- Interventions:

• Control (usual care) – standard national guideline advice to stop clear fluids 2 h before anticipated induction, no extra support.

• Conservative – same 2 h rule plus “instructed adherence”: yellow fasting card, information on schedule and extra ward visits when delayed to maintain the 2 h interval.

• Liberal – same support as conservative but actively encouraged clear fluids up to 200 ml between 2 h before induction and call to the OR (≈100 ml h⁻¹ of delay), i.e. small volumes allowed right up to transfer.

- Primary outcome (preoperative thirst before induction): Thirst (none / moderate / severe) via a modified Bauer item. Compared with usual care, both strategies reduced thirst severity: conservative vs control OR 0.41 (95% CI 0.20–0.82; p=0.013), liberal vs control OR 0.21 (95% CI 0.10–0.43; p<0.001). Symptom benefit was greatest with the liberal “drink-until-call” protocol (moderate-certainty evidence, limited by unblinded PROs).

- Key secondary symptoms:

• Postoperative thirst (~2 h post-op): conservative vs control OR 0.80 (95% CI 0.31–2.03; NS); liberal vs control OR 0.32 (95% CI 0.13–0.75; p=0.009), indicating less thirst post-op with liberal intake.

• Preoperative headache: conservative vs control OR 0.52 (95% CI 0.21–1.25; NS); liberal vs control OR 0.24 (95% CI 0.07–0.66; p=0.009), suggesting reduced headache with liberal fluids.

• Overall preoperative satisfaction and other PROs favoured the liberal strategy but were less completely reported and remain lower certainty.

- Fasting behaviour, safety and feasibility:

• Liquid fasting time (median [IQR], hours): control 5.38 [3.67–9.53], conservative 3.0 [2.23–4.25], liberal 1.97 [1.20–3.02] – both interventions shortened actual fasting, with liberal clustered near the 2 h boundary.

• Pre-op oral liquid volume (median [IQR], ml): control 200 [20–300], conservative 400 [200–600], liberal 400 [212–500].

• No signal of haemodynamic or metabolic harm: similar blood glucose (≈95–98 mg/dl; out-of-range values 6 vs 4 vs 4), similar HR and BP changes around induction (RRsys fall ~54–58 mmHg; RRdia ~29–34 mmHg), and low vasopressor use (Akrinor median 0 ml in all arms).

• Safety endpoints: no bronchopulmonary aspiration, no unplanned ICU/intermediate-care admission for respiratory insufficiency, no deaths, and no postoperative delirium (CAM-ICU) in any group; PONV rates were numerically lower with interventions (19% control, 12.1% conservative, 10.3% liberal) but underpowered for firm conclusions. Overall, in low-risk ASA I–III adults, structured support and especially liberal clear-fluid intake until call to the OR substantially improve thirst and headache without evident short-term harm, though rare aspiration-related events remain too infrequent in this pilot to definitively exclude risk.