Feasibility of using routinely collected inpatient data to monitor quality and inform choice: a case study using the UK inflammatory bowel disease audit