[Validation of the cedars-sinai score in the prediction of rebleeding and mortality in non variceal upper gastrointestinal hemorrhage].

BACKGROUND

Upper gastrointestinal hemorrhage is a serious health problem with mortality rates that have remained unchanged for the last years in spite of improvements in management.

OBJECTIVE

Validating the Cedars Sinai score for predicting rebleeding and mortality at the Arzobispo Loayza National Hospital.

METHODS

Prospective longitudinal study in patients with upper gastrointestinal hemorrhage, collecting demographic, clinical and endoscopic data. The Cedars-Sinai score was applied in order to classify patients: low risk, medium risk and high risk. The final results were rebleeding and mortality. Calibration was done by X2, and the discrimination of the score was assessed by analysis of ROC curves. An area below the curve greater than 0.5 was considered good discrimination.

RESULTS

We evaluated 143 patients, predominantly male (54.5%), the median age was 57.8. Melena (48.95%) and hematemesis (16.08) were the most frequent symptoms. Out of the total number of patients, 36.3% was taking some type of medication; NSAID's were taken by 55.7% and ACEI's by 25%. Associated comorbidity was found in 56.6%. The following diagnoses were made: gastric ulcer (32.15%), duodenal ulcer (30.75%), erosive hemorrhagic gastritis (20.3%), advanced gastric neoplasia (6.3%) and vascular lesions (4.9%). Rebleeding and mortality were 8.4% and 4.2%, respectively. The low risk group did not show any adverse events; the medium risk group, 7.25% and the high risk group, 13.2%. The areas under the curve were for rebleeding (0.79+/-0.13; p=0.01) and mortality (0.73+/-0.12; p=0.01) and considering both (0.66+/-0.12; p=0,034).

CONCLUSIONS

This score is applicable in our environment and it is suitable for predicting rebleeding and mortality in patients with upper gastrointestinal hemorrhage, especially for mortality.