Introduction There has been a sustained increase in demand for gastrointestinal (GI) endoscopy. 1.4%–1.6% of the population undergo upper GI endoscopy per annum, 0.8 % flexible sigmoidoscopy (FS) and 0.6% colonoscopy.1 Complications occur due to the risk of the procedure or sedation. With the advent of the bowel cancer screening programme there has been increasing scrutiny of the safety of endoscopy and strict quality assurance. Both transient ischaemic attacks (TIAs) and strokes (cerebrovascular accidents (CVAs)) are recognised to occur both during and following endoscopic procedures,1 2 however data regarding prevalence are lacking. Our objective was to establish the frequency of stroke after endoscopy in our hospital.
Methods We performed a retrospective audit of stroke occurrence after endoscopy. Hospital episode statistics were cross referenced with endoscopy reporting system from November 2009 to November 2011. Patients admitted with a stroke within 28 days of an endoscopic procedure (OGD, colonoscopy or FS) were identified. The notes were then examined to ascertain further information about demographics, procedure type, comorbidities, complications, haemodynamic changes, time period between procedure and symptoms, length of stay and survival.
Results 8790 procedures were performed: colonoscopy 1953, OGD 4084, FS 2753. Seven strokes were identified; 5 OGD, 1 FS and 1 colonoscopy. 6 of 7 (86%) of the strokes occurred within 10 days, 4 (57%) within 4 days of procedure. Four patients died. Five strokes were cerebral infarcts, two intracerebral haemorrhages. There were no cardiovascular changes or hypoxia during any procedures. 86% of the patients were aged over 75 years. Data from 2 UK audits of OGD and colonoscopy have found the rate of stroke to be 0.04%.1 2 Our rates of stroke following endoscopy are similar for colonoscopy at 0.05% but are 3 times higher for OGD at 0.12%. This suggests post endoscopy stroke is a more common occurrence than is previously documented. Although the relatively small numbers make bias likely, an alternative reason could be the under reporting of strokes occurring in the 28 days following endoscopy.
Conclusion Endoscopy is a safe procedure but it does have risks, we are performing more procedures and have an aging population. Stroke is a serious event with high mortality and long hospital stay. Quality assurance of endoscopy is an important factor in all procedures and our data would suggest that stroke should be specifically looked for following endoscopy. We need to consider if there is any alternative ways of monitoring patients to be able to predict those who are at risk of stroke following endoscopy.
Competing interests None declared.
References 1. Bowles CJ, et al. A prospective study of colonoscopic practice in the UK today. Gut 2004;53:277–83.
2. Quine MA, et al. Prospective audit of upper gastrointestinal endoscopy in two regions of England. Gut 1995;36:462–7.