Monitoring during endoscopy

Attention to sedation techniques may reduce mortality

EDITOR,--J E Charlton draws attention to the estimated overall death rate of 1 in 2000 from upper gastrointestinal endoscopy, which is usually performed under sedation or local anaesthesia, or both.1 When compared with the overall mortality solely attributable to anaesthesia, which is 1 in 185000,2 this mortality seems unacceptably high.

Recommendations for standards of sedation and for monitoring patients during gastrointestinal endoscopy were published by the British Society of Gastroenterology in 1991,3 and guidelines for sedation by non-anaesthetists were published by a working party of the Royal College of Surgeons in 1993.4 Despite this, practice across Britain seems to vary widely.

We were surprised to find that in some regions the preferred benzodiazepine for sedation during endoscopy is diazepam,5 which has an active principal metabolite (desmethyldiazepam) with a half life of 48-96 hours. Midazolam has been recommended as the drug of choice for sedation on the basis of pharmacokinetic and pharmacodynamic properties4 and lends itself to a careful titration technique. Combinations of opioids and benzodiazepines are best avoided owing to the additive and syngergistic effects on the cardiovascular and respiratory systems. It must be remembered that the antagonist drug flumazenil has a shorter half life than the benzodiazepines in common clinical use.

Local anaesthetic sprays have been implicated in the development of pneumonia after gastroscopy.5 Their use may be unnecessary in most diagnostic endoscopic procedures. Formal training of junior endoscopists in sedation techniques together with adequate supervision should ensure the rational use of the available drugs.

When single operator sedationists practise endoscopy the British Society of Gastroenterology recommends at a minimum that a nurse trained in endoscopy techniques and dedicated to patient care is present throughout. Supplemental oxygen and monitoring should ideally be continued throughout the period of recovery, and premature discharge to maximise turnover must be resisted. The employment of qualified staff who are able to monitor patients, interpret changes in the monitored variables, and start cardiopulmonary resuscitation when necessary may have implications for general practice endoscopy services, where it may prove difficult to ensure both safety and cost effectiveness.

We may be able to influence these worrying statistics by centralising services, ensuring proficiency of endoscopists, providing skilled help, and adopting the excellent guidelines already available.