The dilemma in the management of suspected upper GI bleeding in patients with COVID-19 pneumonia

To the Editor:

We read with interest the article entitled “Management of upper GI bleeding in patients with COVID-19” by Cavaliere et al,¹ published recently in Gastrointestinal Endoscopy. The authors reported a case series of 6 patients who presented with coronavirus disease 2019 (COVID-19) and upper GI bleeding (UGIB). Cavaliere et al¹ concluded that COVID-19 patients with UGIB could be treated conservatively without endoscopy because they responded to conservative treatment in 24 hours. This study is very interesting and important; however, we would like to put forth some suggestions.

First, endoscopy, which could identify the cause of UGIB,² was not performed in this study, and the exact cause of hematemesis or melena is still unclear. Indeed, not all patients with hematemesis or melena had “true” UGIB episodes. Some coffee-ground hematemesis episodes are due to pulmonary embolism, myocardial infarction, and renal failure.³ Also, the intake of substances, including iron supplements, bismuth subsalicylate, and foods such as blood soup, can lead to black stools similar to melena.⁴ It is unknown whether the patients had consumed the above-mentioned medications or foods before admission. In addition, the nature of hematemesis (bloody, fresh coffee-ground, or old) is not reported.

Second, the presence of coexisting diseases such as liver cirrhosis, renal disease, Helicobacter pylori infection, and history of variceal bleeding, peptic ulcers, GI tumor, severe vomiting, and use of nonsteroidal anti-inflammatory drugs or antiplatelet medications was unclear. Many studies^{5 , 6} have demonstrated that patients with severe COVID-19 have subclinical or obvious coagulation abnormalities with increased risk of thromboembolic disease. However, prothrombin time, activated partial thromboplastin time, and international normalized ratio were not reported in Table 1 of the article. It should be noted that detailed medical history taking and elaborate laboratory tests are very important to diagnose the cause of UGIB when endoscopy is not performed.

Third, we suggest that the authors use the Horibe GI bleeding prediction score (HARBINGER),⁷ which is simple and accurate for triage in patients with suspected upper GI bleeding. In a study of 1486 patients with suspected UGIB, the HARBINGER score was more accurate (area under the curve [AUC] 0.76) than both the Glasgow-Blatchford Score (GBS) (AUC 0.68) and AIMS65 (AUC 0.54). Moreover, the HARBINGER is the first score used when neither the presence of UGIB nor the variceal or nonvariceal cause of the UGIB could be identified. In a previous study involving 3012 patients, a GBS of ≥7 was shown to predict with greater accuracy whether or not a patient requires endoscopy.⁸ However, all of the GBS were >7 in this study, which suggests that the GBS was not accurate in this case series.

Fourth, if available, magnetically assisted capsule endoscopy (MACE) could be safely used to examine the cause of acute UGIB. A recent study⁹ showed that MACE had a higher diagnostic performance for focal lesions and could correctly identify the cause of UGIB, with a better tolerance than EGD, and a lower risk of aerosol transmission¹⁰ than EGD in theory.

Finally, this case series of UGIB in COVID-19 patients can reflect the “real-world” findings; however, the medical history and blood tests warrant further study. Because the sample size is small, a large-scale study should be conducted in the future for assessing the risk for COVID-19 patients with suspected UGIB. Innovative noninvasive methods for triage before endoscopy during the COVID-19 pandemic needs further exploration.

Disclosure

All authors disclosed no financial relationships.