P69 Insourcing colonoscopy – is there a difference?

Abstract

Introduction Gastrointestinal endoscopy services are under increasing pressure to meet demand. Subcontracting procedures to external providers - ‘insourcing’, is increasing and often used in addition to waiting list initiative (WLI) lists by local teams. Guidance suggests the quality of provision is assessed through core key performance indicators (KPIs). This study assesses the quality of weekend colonoscopy by endoscopists from a single insourcing provider compared to equivalent weekend WLI lists performed by the local hospital team. Methods A retrospective analysis of colonoscopies performed at a single site across weekends for 12 months from Jan 2019 was conducted. Pseudo-anonymised data was extracted from the endoscopy reporting system and correlated with pathology. Confirmation of caecal intubation was independently verified by reviewing photo-documentation (P-D). Outcomes were defined by core KPIs and reporting standards. Chi square test was used to assess for associations between dichotomous variables and independent t test for continuous outcome variables. Results 2157 procedures were performed, 55.9% were performed by the local team. Mean patient age was 54.9 years (SD 14.6) with no age difference between WLI and insourced procedures (p = 0.81). There were 28 local and 27 insourced endoscopists. Mean procedure count per endoscopist was 43.0 (SD 36.5) and 35.0 (SD 44.9) for local and insourced respectively. The local team performed more procedures for surveillance (26.3%) or inflammatory bowel disease (21.4%). Insourced procedures were more likely to be for anaemia (25.2%) or change in bowel habit (33.1%). Table 1 highlights the KPIs assessed: Conclusions The local team achieved significantly higher ADR, PDR and caecal P-D rates. Insourced procedures had lower sedation rates. In both cohorts, P-D did not support self-reported CIR. ADR was below the minimum standard for insourcing procedures, although this may reflect the difference in case mix observed and all other KPIs were achieved. More work is needed to establish whether findings are related to weekend working or effects of differing working environments.