PTH-061 Missed oesophagogastric cancer correlates with higher list intensity but not rates of sedation

Abstract

Introduction Oesophagogastric (OG) cancers diagnosed within three years of an unremarkable oesophagogastroduodenoscopy (OGD) can be considered a failure to earlier diagnose the OG cancer, or post-OGD upper GI cancer (POUGIC). Retrospective studies suggest that they comprise up to 11% of OG cancer diagnosis [Menon, Endosc Int Open 2014] and auditing rates of POUGIC is a recent quality standard for endoscopy units [Beg, Gut 2017]. We examined whether patient sedation or procedural burden affects the rate of POUGIC. Methods Cases of OG cancer diagnosed at OGD between Jan 2013 and Dec 2016 at Sheffield Teaching Hospitals were identified from our upper GI cancer database. OGD performed up to three years prior to diagnostic OGD were reviewed to identify cases of POUGIC. Rates of sedation and number of procedural points (one for OGD; two for colonoscopy; plus one for therapeutics) on lists were compared in three groups: a) the index procedures, b) the diagnostic procedures and c) age, sex and endoscopist matched patient control procedures in which focal (mucosal or vascular) lesions (FL) were identified. FL were approximated in size and location (oesophageal or gastric) as a surrogate for an early neoplastic lesion. Results A total of 553 patients (64.2% male, mean age 72±1, 50.4% gastric) were diagnosed with OG cancer. Forty (7.2%, mean age 74±2, 55% male, 55% gastric,) patients had 47 non-diagnostic procedures up to three years prior to diagnosis. Mean time from index to diagnostic OGD was 486±55 days. In 42.5% OGD was performed within one year of diagnosis. There was no difference in the age, gender and rates of sedation (25 vs 28.5%) between patients at index and diagnostic procedures. At index OGD the sedation rates were higher (44.7 vs 26.3% p=0.049) than at diagnostic OGD but there was a greater number of procedural points on the list (7.9 vs 9.3 p=0.007). Control patients (n=38, mean age 72±2 p=0.64 compared to POUGIC patients) with FL had OGD done a median of 33 days (−357 to 728 days) from the index OGD. No suitable controls were identified in 2 patients. There was no difference in the sedation rates (25.0 vs 26.3% p=0.89) but there was a trend towards a higher number of procedural points (9.3 vs 8.7 p=0.057) between the index OGD for POUGIC patients and their controls. Conclusions The local POUGIC rate is 7.2%. No differences in sedation rate between index, diagnostic or control procedures with representative FL suggest use of sedation may not help detection of early neoplastic lesions. However, endoscopy lists with OGDs which miss OG cancer seem to have a heavier burden than ones that diagnosed cancer and other FL suggesting that reduced list intensity may reduce the likelihood of missed pathology.