1) Overall descriptive data
Our surgical and GI divisions were consulted to manage bleeding episodes in 122 COVID-19 patients. There were 67 non-GI bleeds (55%) and 55 GI bleeds (45%). Overall mortality was 59% (n= 72), 34 (28%) and 38 (31%) deaths followed non-GI and GI bleeds respectively. World health organization grade 2 and 3 were the most common bleeding grades in the series. For non-GI bleeds the distribution of WHO bleeding grades were; WHO 0 (n=0), WHO 1 (n= 12), WHO 2 (n= 36), WHO 3 (n= 18), WHO 4 (n=1). For GI bleeds; WHO 0 (n=6), WHO 1 (n=7). WHO 2 (n=20), WHO 3 (n=22), WHO 4 (n=0).
Of all 67 non-GI bleeds 5 patients (7.5%) required invasive interventions to control the bleeding with 100% success rate to resolve the bleeding event (Table 1). Thirty-four patients died within 30 days; 7 in patients who had retroperitoneal and abdominal bleeds, 2 in patients who had hematuria, 15 in patients who had NPA bleeds, 5 in patients who had brain bleeds and 5 were unclear.
At initial assessment, of all 55 GI bleeds, the source of bleeding was deemed to be upper GI in 12 patients, lower GI in 4 patients and 39 were unknown. Eighteen patients had esophagogastroduodenoscopy (EGD) (32.7%), and 5 had colonoscopy (9.1%). Of the 18 EGD, 13 (72.2%) were performed at time of consultation and 5 (27.3%) after failure of conservative treatment. Of the 5 colonoscopies, 4 (90%) were performed at time of consultation and 1 (10%) after failure of conservative treatment. The yield rate (having a positive finding) for EGD was 88.8% (only two patients had normal EGDs) and 40% for colonoscopies (three patients had normal colonoscopes and or were diagnosed with piles). Overall, the prevalence of invasive interventions for GI bleeds was 16.3% (8 upper and one lower GI bleeds) with 100% success rate (Table 1).
DOSE OF ANTICOAGULATION
Overall, 79 patients (67.7%) were on therapeutic anticoagulation, 15 patients (12.3%) were on intermediate dose, 22 patients (18%) were on prophylactic dose and 6 patients (5%) were on none. In the non-GI bleeding group, 45 (67%) patients were on therapeutic anticoagulation, 9 (13.4%) were on intermediate dose, 11 (16.4%) patients were on prophylactic dose and 2 (3%) were on none. In the GI bleeding group, 34 (62%) were on therapeutic dose, 6 (11%) were on intermediate dose, 11 (20%) were on prophylactic dose and 4 (7.2%) were on none.
NON-GI BLEEDING
a) Retroperitoneal, intraperitoneal and abdominal wall hematoma
We identified 9 cases of retroperitoneal bleeding (7.3%). Three patients were on prophylactic anticoagulation, five were on therapeutic dose and one was on none. Only one patient was on antiplatelet (aspirin) as well as anticoagulation. Two patients (both were on prophylactic dose) required angioembolization to control the bleeding and both were resolved. The rest resolved with no intervention. Of the 7 patients who did not require interventions 3 died within 30 days of consultation from sepsis.
We identified 2 intraperitoneal bleeding. Both patients were on therapeutic dose anticoagulation and both required angioembolization to successfully control the bleeding. However, one of the two patients died later within 30 days from sepsis.
Five patients with abdominal wall hematomas were encountered. One patient was on intermediate dose and 4 on therapeutic dose. All bleeding resolved with noninvasive interventions except one which requited angioembolization to control the bleeding. Of all patients with abdominal wall bleeding, one died within 30 days after the consultation from COVID-19 multiorgan failure.
b) Hematuria
Twelve patients had hematuria. Four patients were on prophylactic dose, five on full dose and three on intermediate dose anticoagulation. All bleeding resolved without the need for invasive intervention. Two patients died within 30 days.
c) Brain
Seven patients had cerebral bleeding events. Two patients were not on anticoagulation; one patient developed subarachnoid hemorrhage which had stabilized without intervention, however, patient later died from sepsis and the other patient survived the bleeding event. Five patients were on therapeutic anticoagulation; 4 patients expired; one as a direct consequence from intracerebral bleeding (deemed inoperable) other following combined spontaneous subdural and epidural hematomas (deemed inoperable), and 2 had hemorrhagic infarcts (died from other comorbidities within 30 days).
d) Nasopharyngeal bleeding
We identified 25 patients who had NPA bleeding. Diagnoses ranged from epistaxis, oral bleeding and tracheal site bleeding. Interventions, which were all noninvasive, included holding the anticoagulation till bleeding resolved, nasal packing and administering vitamin K. All bleeding resolved with these techniques. Fifteen patients died within 30 days from sepsis. Of all NPA consultations, four patients were on prophylactic dose, 3 on intermediate dose, and 18 on therapeutic dose anticoagulation. Of patients who died, all were on therapeutic dose anticoagulation except two, one was on intermediate dose, and the other was on none.
GI BLEEDING
Out of 55 patients who had signs of GI bleed at initial presentation 47 patients were managed with full dose proton pump inhibitors (PPI) and active observation, of whom 6 failed this approach and required endoscopic intervention, five EGD and 1 colonoscopy.
Of the 19 patients who had endoscopic procedures 18 EGDs and 5 colonoscopies were performed. Nine patients (16.3%) required an endoscopic hemostatic intervention to control the bleeding, all were successful. Eight patients required upper GI hemostatic intervention and 1 lower GI hemostatic intervention. The upper GI interventions included gold probe, epinephrine injection, APC and hemoclips applications for bleeding ulcers. Bleeding resolved for all cases. Etiology for bleeding events were duodenal & gastric ulcers in 10 patients (55.5%), esophagitis and Roux-en-Y anastomosis ulcer in 1 patient, and gastritis in 5 patients (27.7%). Three patients died within 30 days of consultation. The only lower GI endoscopic intervention was for large colonic ulcer in the cecum which was treated with APC. Bleeding resolved but patient later died from sepsis. Other colonoscopies done without interventions were for; small rectal ulcer with piles for 1 patient, ischemic colitis for 1 patient, piles for 1 patient, and one colonoscopy was essentially normal.
Overall, of 55 patients with signs of GI bleed, 38 (69.1%) expired within 30 days of consultation. Two patients had unclear source of bleeding and died while still bleeding, both were on Plavix and therapeutic anticoagulation at the time of consultation. The rest died after resolution of the bleeding episode from COVID-19 complications.
2) Demographics and baseline characteristics by primary outcomes, Univariate analysis (1) – Table 2
Mortality outcome
Male patients (80.56%) were significantly more likely to die than female patients. The mean age for patients who died was 60 years old. They were also significantly more likely to be admitted to intensive care unit (ICU) and had shorter hospital stay compared to patients who survived. Medical comorbidities at the time of admission did not have significant implication on the risk of mortality. These included having a history of hypertension requiring medications, asthma or chronic obstructive pulmonary disease (COPD), having complicated diabetes, remote history of myocardial infarction, or being on antiplatelets (single or dual) at the time of admission. Moreover, admission baseline hemoglobin level, before the onset of bleeding episode, did not affect the risk of mortality.
Resolution of bleeding event & having GI bleed outcomes
Gender appears to have a significant association with bleeding resolution outcome, but it did not appear to influence the risk of having a GI bleeding compared to a non-GI bleed. ICU patients were more likely to have a GI bleed than non-GI bleed.
3) Inpatient therapies and complications developed during patient’s admission by primary outcomes, Univariate analysis (2) – Table 2
Mortality, resolution of bleeding event, & having GI bleed outcomes
The use of PPI, vasopressors, and inotropes appears to be significantly associated with the risk of death. We also found GI bleeds to be significantly associated with the need to use PPI and inotropes but not vasopressors. Furthermore, the dose but not the type of anticoagulation at the time of admission was significantly associated with risk of death following the bleeding episode. The dose of anticoagulation however did not influence bleeding event resolution or the type of bleeding.
Being on systemic steroids appears to be significantly associated with increased risk of death but not the bleeding resolution or the type of bleeding. Furthermore, being on invasive ventilation was associated with increased risk of mortality and having a GI bleed rather than a non-GI bleed. Moreover, severe ARDS was associated with higher risk of mortality, but it did not affect the bleeding resolution or the type of bleeding event. With regard to mortality outcome, having a cardiac injury, liver injury, acute kidney injury, Glasgow coma scale (GCS) less than 15, systolic blood pressure less than 100 and sepsis and high qSOFA score were associated with higher risk of death. However, the CCI score and the WHO bleeding grade did not affect the risk of death. Furthermore, having cardiac injury, liver injury, being on renal replacement therapy, having a respiratory rate over 22, systolic blood pressure less that 100, and sepsis appears to significantly influence the type of bleeding event. Also, the qSOFA and WHO bleeding grade but not CCI score significantly affected the type of bleeding event. With regard to resolution of bleeding event outcome, there was a significant association with the WHO bleeding grade. None of the complications the patients developed during admission affected bleeding resolution chance.
4) At consultation symptoms, laboratory values and interventions by primary outcomes; Univariate analysis (3) – Table 2
Mortality
Having an occult source of bleeding rather than a specific symptoms and signs indicative of a source of bleeding was significantly associated with the risk of death. Furthermore, hemoglobin level, white blood cell (WBC) count, platelet, international normalized ratio (INR), D-dimer level, e-glomerular filtration rate (eGFR) level, urea, creatinine, and C reactive protein (CRP) were significantly associated with the risk of death. There was no relationship between the type of intervention (invasive, noninvasive or hemostatic) and the risk of death within 30 days.
Resolution of bleeding event
We did not identify a significant relationship between any coagulation profile derangements and the ability to control bleeding episode. However, CRP and Procalcitonin (PCT) were significantly associated with bleeding resolution.
Having GI bleed outcome
Having an occult bleed appeared to be significantly associated with having a GI bleed. Also, having high urea but not deranged coagulation profile was significantly associated with having a GI bleed. Furthermore, there was a significant relationship between the type of intervention (invasive, noninvasive or hemostatic) and the type of bleeding event.
5) Relationships between major outcomes – Univariate analysis (4) – Table 3
We found having a GI bleeding event was significantly associated with the risk of death (P = 0.04). The prevalence of death following bleeding event was higher following a GI bleed compared to a non-GI bleed, 52.7% vs. 48.3% respectively.
6) Risk factors predictors of primary outcomes
Mortality - Multivariate logistic regression (1) – Table 4
Patient who had longer hospital stay appeared to be less likely to die, odds ratio (OR) 0.95 (95% CI, 0.92-0.98, p= 0.003). We also found that patients who were on therapeutic dose of anticoagulation were more likely to die compared to patients who were on none, on prophylactic or intermediate anticoagulation doses. This risk appears to be significant when therapeutic dose was compared to prophylactic dose, OR 0.07 (95%CI 0.02-0.028, p=0.03) and no anticoagulation, OR 0.1 (95%CI0.97-0.99, p < 0.00) but not significant when compared to intermediate dose, OR 0.36 (95%CI 1.02-1.15, p=0.13). Furthermore, having an occult bleeding appeared to be a significant predictor of risk of death, OR 15 (95% CI 1.97-29.1, p= 0.013). Also, WBC and platelet levels appeared to independently affect risk of death.
Resolution of bleeding event - multivariate logistic regression (2) – Table 5
Patients who were on PPI were more likely to have resolution of bleeding event compared to patients who were not. Out of all GI symptoms and signs melena appeared to be significantly associated with lower odds of bleeding resolution, OR 0.03 (95%CI 0.01-0.18, p < 0.00). C-reactive protein appeared as well to be significantly associated with lower odd of bleeding resolution, OR 0.98 (95% CI 0.97-0.99, p < 0.00).
Type of bleeding event - multivariate logistic regression (3)- Table 6
The risk of GI bleeding increased when patient was on inotropes (OR 7.33, 95%CI 1.03-55.28, p= 0.005), had cardiac injury (OR 6.73, 95%CI 0.92-49.43, p= 0.06), had liver injury (OR 74.08, 95%CI 4.18-132.08, p= 0.03), had qSOFA score of 3 (OR 23.43, 95%CI 4.94-374.73, p= 0.02), had hematemesis (OR 19.79, 95%CI 2.23-175.74 p= 0.00), and had occult bleed (OR 32.24, 95%CI 3.34-311.08, p= 0.00). The mortality variable had poor correlation with the type of bleeding event on multivariate analysis model and so it was removed from the model.