Article Text
Abstract
Introduction Prompt endoscopy for acute upper gastrointestinal bleeding (AUGIB) is associated with improved outcome. NICE recommends early endoscopy (<24 hour from admission) for all patients with AUGIB whereas the JAG Global Rating Scale stipulates early endoscopy in ≥75% as a minimum standard for service accreditation. We aimed to audit these outcomes and identify predictors of delayed endoscopy (>24 hour from admission).
Methods A prospective, pan-Midlands, multi-centre study was jointly undertaken by GARNet and WMRIG trainee networks. Adults admitted with AUGIB and had inpatient endoscopy between Nov-Dec 2017 were enrolled over 30d. Admission, endoscopy referral and procedure times were collected, along with clinical, laboratory, endoscopic and post-endoscopic variables. Heterogeneity between sites was assessed using Mann-Whitney and chi2. Multivariate binary logistic regression analysis was used to study factors associated with delayed endoscopy.
Results 337 patients met inclusion criteria (median age 69.5, SD 18.8). The median time from admission to endoscopy (figure 1) was 20.9 hour (IQR 11.5–31.8). The time from admission to endoscopy referral were comparable between East and West Midlands (median 8.1 hour, IQR 3.6–18.1; p=0.242], as was the time from referral to endoscopy (median 6.6 hour, IQR 3.0–22; p=0.219). 61.1% of patients received endoscopy within 24 hour of admission (p=0.025 across sites) and 79.3% within 24 hour of referral (p=0.012). 4/20 sites (20%) met the minimum JAG standard. On multivariate analysis (table 1), 7 pm-7 am admission, rectal examination <1 hour, higher Glasgow-Blatchford score (GBS) were associated with early endoscopy. Each 1 hour increment in referral time led to a 4% added risk of delayed endoscopy. Weekend admission, region, melaena or suspected varices did not affect this outcome. Early endoscopy did not affect rates of endoscopic therapy (p=0.536), 30d readmission or death (p=0.985), but reduced length of stay (median difference 1d; p=0.039).
Conclusions Time to endoscopy for AUGIB generally fell below national standards during the period of Nov-Dec. Early endoscopy can reduce length of stay, but is dependent on prompt recognition, assessment and referral. As such, ongoing audit and strategic initiatives involving acute care services may be required to improve this outcome.