Can pre-endoscopic assessment predict active upper gastrointestinal bleeding? A retrospective study in patients with symptoms of upper gastrointestinal bleeding outside regular working hours.

BACKGROUND/AIMS: To evaluate the usefulness of pre-endoscopic assessment for predicting active up-per gastrointestinal bleeding (UGI-B) at emergency esophagogastroduodenoscopy (E-EGD, within 6 hours).

METHODOLOGY: We retrospectively analysed the medical records of patients that had an E-EGD performed outside working hours and considered 15 pre-endoscopic variables in a univariate analysis. Active UGI-Bat E-EGD was taken as end-point.

RESULTS

Of 228 E-EGD performed during 75 months, 195 were motivated by the suspicion of UGI-B. We excluded 83 cases as they were hospitalised at the time of first symptoms of bleeding. Thus, 112 cases were included. The following clinical signs triggered E-EGD: hematemesis (56/50%),melena (55/49.1%), hematochezia (20/17.8%), anae- mia (7/6.2%). Patients' age was 65.5+14.2 years. Sixty nine (61.6%) cases were male. The relative risk and p-value of the variables for the presence of active bleeding at E-EGD were as follows: hematemesis: 1.54/0.3; malignancy and cirrhosis: 1.73/0.07; haemoglobin <8g/dL: 1.38/0.3; white blood count >12,000/tL: 1.18/0.6;systolic blood pressure (SBP) <100 mmHg: 0.53/0.03;pulse >100/min: 1.42/0.2; platelets <14000/nL:1.5/0.2; INR >1.17: 1.89/0.049. In the multivariate analysis none of these variables independently predicted UGI-B.

CONCLUSIONS

No relevant pre-endoscopic variables for the prediction of active UGI-B at E-EGD could be found. Our data suggest that pre-endoscopic evaluation cannot replace rapid endoscopy.