Background: Blood urea nitrogen and creatinine ratio (BCR) may suggest the source of bleeding is from the upper gastrointestinal tract. The objective of this study is to ascertain the diagnostic value of BCR for non-variceal upper gastrointestinal bleeding in high prevalence chronic kidney disease (CKD) patients.
Methods: All adult patients admitted for suspected upper gastrointestinal bleeding (UGIB) and had upper endoscopic examination from November 2017 until November 2018 were reviewed. Apart from demographic data, other analyzed parameters were BCR at presentation, degree of renal failure, endoscopic findings and its intervention.
Results: A total of 104 patients was admitted within the study period. Eleven patients were excluded due to bleeding varices. Seventy-nine (76%) of them had non-variceal causes of bleeding and other remaining patients were either normal or had mild gastro-duodenitis (n=14, 15%). Prevalence of renal failure was much higher in the former group but not statistically significant, 54.4% vs 35.7%, p=0.197, respectively. The mean BCR also was significantly higher in the former group, 37.1 ± 16.79 vs 22.32 ± 10.99, p=0.002, respectively. Stepwise forward and backward multivariable logistic regression analysis showed BCR (adjusted OR 1.10, 95% CI: 1.041 1.178, p=0.001) and renal failure (adjusted OR 6.37, 95% CI: 1.470 27.606, p=0.013) were independently associated with non-variceal upper gastrointestinal bleeding with good ROC value (ROC=0.835, 95% CI 0.741 0.929, p=0.0001). Sensitivity, specificity, positive and negative predictive value for BUN creatinine ratio of > 15 and > 30 were 94.9%, 35.7%, 89.3%, 55.6% and 67.1%, 64.3%, 91.4%, 25.7%, respectively in predicting upper gastrointestinal bleeding.
Conclusions: Blood urea nitrogen creatinine ratio remains a significant predictor for UGIB even in high prevalence CKD patients.
Key words: blood urea nitrogen; creatinine; gastrointestinal bleeding; melena