Objective To identify the methods employed within the UK practice prior to diagnostic gastroscopy and compare with published guidelines for patients undergoing general anaesthesia.

Design National Health Service (NHS) endoscopy units were invited to take part in a structured telephone survey to determine the length of time patients are kept nil-by-mouth (NBM) for food and fluids prior to gastroscopy, and whether a preprocedure mucolytic drink was used.

Methods 212 NHS Trusts providing endoscopy services were identified from the Joint Advisory Group on GI Endoscopy. Trusts were excluded if they were children's hospitals (n=5).

Results 207 NHS Trusts were telephoned. 193 completed the survey (93%), 11 Trusts declined and there was no response from 3 Trusts. 13 separate policies regarding NBM timings were identified. 51 Trusts (21%) used the timings ratified by Surgical and Anaesthetic Societies (6 h NBM for food, 2 h for clear fluid). 135 Trusts (70%) used a policy which starved patients in excess of the standard surgical guidelines. No Trust used a mucolytic drink prior to gastroscopy.

Conclusions The survey revealed large variation in NHS Trust's policies regarding the times patients were starved prior to gastroscopy. Results of surgical studies demonstrate increased risk of significant pulmonary aspiration with increased fluid-starvation periods, 68% of NHS endoscopy policy would be deemed excessive by surgical practice. There is no routine use of a mucolytic drink to improve mucosal visualisation in the UK practice.