Endoscopy involves using a flexible camera to examine the stomach and bowel. It is performed almost 2 million times a year in England to diagnose early cancer, stop bleeding and treat symptoms. In most patients the procedures are safe, however there are uncommon but important risks following the test. Previous studies have not provided a complete picture of these risks, as they have focussed only on events related to the stomach and bowel, had short follow up times, or not been large enough to capture less common events. The routinely recorded data from primary care and hospitals in the Clinical Practice Research Datalink provides an opportunity to measure these risks in a larger population. In addition, the records provide the background information on patients' pre-existing disease that puts any measured risks in context. The data is fully anonymised so that individual patients cannot be identified, and this allows this study to be conducted on a scale that would not be possible direct patient recruitment.
We will measure excess 30 day absolute risks of emergency hospital admission or death following gastrointestinal endoscopy. These risks will be stratified by type of event, age, sex, socioeconomic status, whether the procedure was planned or an emergency, whether a therapy was performed, and by pre-existing co-morbidity using the Charlson index and bespoke co-morbidity score derived from the CPRD. Events due to diagnoses that might have been related to the indication for the endoscopy will be censored, and will be included in the models as a competing risk using cumulative incidence functions. In addition, further procedures will also be censored. The excess risk associated with each procedure will be calculated by the difference between the risks from the endoscopy cohort with those in an age and sex frequency matched population comparison cohort. We will also perform two secondary analyses to assess the excess risk over an extended 60-days post procedure to identify the time period when any excess risk occurs and whether this returns to the baseline in the comparison cohort by 30 days. Finally, we will assess changes in risks with a test for trend over the calendar years of the study.
Emergency Hospital admission within 30 days; All-cause mortality within 30 days.
Colin Crooks - Chief Investigator - University of Nottingham
Colin Crooks - Corresponding Applicant - University of Nottingham
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation