Prospective audit of perforation rates following upper gastrointestinal endoscopy in two regions of England

Royal College of Surgeons, London, UK.
British Journal of Surgery (Impact Factor: 5.54). 04/1995; 82(4):530-3. DOI: 10.1002/bjs.1800820430
Source: PubMed


After cardiopulmonary complications, perforation is the second most important cause of complications following flexible upper gastrointestinal endoscopy. A recent audit of 14,149 procedures detected a perforation rate of 0.05 per cent (overall mortality rate 0.008 per cent) during diagnostic endoscopy, and a perforation rate of 2.6 per cent (overall mortality rate 1.0 per cent) following oesophageal intubation or dilatation. The incidence of perforation following both diagnostic and therapeutic upper gastrointestinal endoscopy has not changed over the past 10 years. The risk factors are numerous but this audit demonstrated that inexperience increases the likelihood of perforation.

7 Reads
  • Source
    • "The report acknowledged that enthusiastic general practitioner (GP) endoscopists were capable of offering some endoscopy services in primary care but questioned whether it could be cost effective. The death rate for diagnostic gastroscopy in hospital practice was found to be one in 2000 by Quine in 1995. 3 The BSG report suggested that gastroscopy was a procedure best performed in hospital because of this mortality risk. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Long waiting lists in district general hospitals and savings from fundholding led to the setting up of a number of endoscopy units in primary care. Concerns have been expressed over safety, supervision and cost effectiveness. Increasingly, general practitioners (GPs) are being encouraged to become specialists and offer intermediate care. Endoscopy is frequently cited as an example of intermediate care that could be offered by primary care specialists. This is the first survey of such a service. To examine whether endoscopy in primary care can be considered to be a safe procedure. A questionnaire-based survey. Twenty-eight general practice units performing endoscopy in primary care. Units performing endoscopy in primary care were identified using the Primary Care Society of Gastroenterology (PCSG) database and following an appeal in the GP press. A postal questionnaire was sent to each unit covering its history, throughput, and case-mix, experience of endoscopists, supervision, audit and CME, equipment, waiting times and complication rates. Of the 28 units identified, 27 (96%) replied to the questionnaire, 13 units provided both upper and lower bowel examination, six oesophago-gastro-duodenoscopy (OGD) only, and eight lower bowel only. Units had been openfor an average of five years (range = 2 to 18 years), and 41 doctors and 68 nurse assistants provided the service. The average experience of endoscopists was 16 years (range = 6 to 25 years), and 36,455 procedures had been performed by the time of the survey (24,195 OGD and 12,260 lower bowel examinations). Ninety-six per cent of the units undertook audit. Urgent waiting times were 1.2 weeks and routine 3.4 weeks (range = 1.0 to 6.0). The annual throughput of 22 units in the past year was 8,478 procedures (4506 OGD, 3,972 lower bowel examinations). Out of 24,195 OGDs there were three reported complications (one perforation of pharyngeal pouch, treated conservatively, one chest pain after over-insufflation, and one slow recovery after intravenous sedation); there was no mortality. Out of 12,260 lower bowel procedures there was one perforated caecal carcinoma after flexible sigmoidoscopy (died), three perforations at colonoscopy and seven other minor complications. Endoscopy in primary care appears to be a safe procedure. This good safety record is probably attributable to careful case selection and minimal use of intravenous sedation.
    British Journal of General Practice 08/2002; 52(480):536-8. · 2.29 Impact Factor

  • British Journal of Surgery 06/1996; 83(6):845-6. DOI:10.1002/bjs.1800830636 · 5.54 Impact Factor

  • Endoscopy 11/1996; 28(8):718-21. DOI:10.1055/s-2007-1005583 · 5.05 Impact Factor
Show more