Background

Gastrointestinal endoscopy has rapidly evolved over the last decade. The volume of diagnostic endoscopy coupled with advances in technology and therapeutic interventions has resulted in an expanded remit for endoscopy. This is in the context of an ageing population with associated complex comorbidity. In order to rise to this challenge and deliver an effective service, a considered analysis of endoscopy safety is required. The achievements made in the UK by the Joint Advisory Group for gastrointestinal endoscopy (JAG) through initiatives such as the Global Rating Scale (GRS) and the National Endoscopy Training programme, are significant and have set the endoscopy safety and quality agenda both nationally and internationally.

In 2000, endoscopy patients in England experienced variable and often substandard services with reported waiting times of up to a year.1 In 2004, Bowles' study of colonoscopy practice in the UK outlined significant safety issues.2 Suboptimal caecal intubation rates, inadequate patient consent processes and wide variation in safe sedation practices were highlighted as issues to be addressed. These findings were corroborated by the National Confidential Enquiry into Patient Outcomes and Death3 following gastrointestinal endoscopy, illustrating further examples of avoidable error.

In response to this, there have been significant improvements in endoscopy4 in the UK underpinned by addressing endoscopy training issues. This was primarily driven by the requirement for a national bowel cancer screening programme5 necessitating consistent high-quality practice with rigorous safety standards required for population-based screening.

There is no doubt that endoscopy across the UK has improved significantly and improving quality is an international aspiration. However, the goals have also changed (getting to the caecum vs getting to the submucosa). Similarly, endoscopy is increasingly therapeutic and may be considered a surgical specialty. In light of this, we must recognise that the safety profile will change as technology …