Poster Session E - Tuesday Afternoon
Category: GI Bleeding
Maryam Alahmad, MBBS
Sheikh Shakhbout Medical City, in partnership with Mayo Clinic College of Medicine
Abu Dhabi, Abu Dhabi, United Arab Emirates
| AUC-1 | AUC-2 | AUROC-3 | AUC-4 |
Predictor variable | Unadjusted | GBS-adjusted | Age-adjusted | GBS- and age-adjusted |
Glasgow-Blatchford score | 0.493 |
| 0.704 |
|
Age | 0.695 | 0.704 |
|
|
Gender | 0.590 | 0.605 | 0.724 | 0.731 |
Nationality | 0.630 | 0.634 | 0.737 | 0.745 |
Diabetes | 0.563 | 0.565 | 0.697 | 0.705 |
Hypertension | 0.609 | 0.627 | 0.696 | 0.706 |
Dyslipidemia | 0.523 | 0.530 | 0.698 | 0.707 |
Chronic kidney disease | 0.577 | 0.602 | 0.694 | 0.704 |
Ischemic heart disease | 0.578 | 0.566 | 0.695 | 0.715 |
Atrial fibrillation | 0.568 | 0.582 | 0.704 | 0.718 |
Congestive heart failure | 0.510 | 0.519 | 0.696 | 0.706 |
Chronic liver disease | 0.526 | 0.530 | 0.700 | 0.708 |
Cerebrovascular accident | 0.563 | 0.575 | 0.691 | 0.702 |
Malignancy | 0.508 | 0.522 | 0.696 | 0.705 |
Antiplatelets | 0.533 | 0.537 | 0.696 | 0.706 |
Anticoagulant | 0.527 | 0.548 | 0.699 | 0.709 |
Proton pump inhibitor | 0.524 | 0.525 | 0.695 | 0.703 |
Steroid | 0.506 | 0.514 | 0.695 | 0.702 |
NSAIDs | 0.520 | 0.537 | 0.704 | 0.712 |
Admission unit | 0.515 | 0.534 | 0.700 | 0.706 |
Admission time | 0.538 | 0.533 | 0.699 | 0.711 |
Prior GI bleeding | 0.517 | 0.513 | 0.693 | 0.705 |
Hematemesis | 0.578 | 0.572 | 0.711 | 0.717 |
Melena | 0.587 | 0.602 | 0.716 | 0.719 |
Drop in haemoglobin | 0.519 | 0.504 | 0.693 | 0.706 |
PR Bleeding | 0.528 | 0.541 | 0.698 | 0.705 |
Syncope | 0.502 | 0.511 | 0.694 | 0.705 |
Urea | 0.532 | 0.518 | 0.698 | 0.706 |
Initial Hgb | 0.498 | 0.516 | 0.693 | 0.701 |
SBP | 0.383 | 0.448 | 0.698 | 0.710 |
DBP | 0.599 | 0.648 | 0.738 | 0.738 |
Heart rate | 0.593 | 0.599 | 0.700 | 0.708 |
Unstable Pulse | 0.570 | 0.607 | 0.700 | 0.705 |
Blood Transfusion | 0.505 | 0.497 | 0.698 | 0.705 |
AUC=area under the receiver operating characteristic curve, GBS=Glasgow-Blatchford score. AUC-1 estimates were obtained from single variable logistic regression models separately for each predictor variable. AUC-2 estimates were obtained from logistic regression models separately for each predictor variable with GBS included as an additional predictor variable. AUC-3 estimates were obtained from logistic regression models separately for each predictor variable with age included as an additional predictor variable. AUC-4 estimates were obtained from logistic regression models separately for each predictor variable with GBS and age included as an additional predictor variables. |