ABSTRACT
Background
Upper gastrointestinal bleeding (UGIB) is one of the most common medical emergencies. Currently, there is a paucity of data on the clinical profile and outcome of UGIB in resource‐limited settings.
Objective
We aim to describe the etiology and outcomes of the patients who presented to the Emergency Department (ED) of a tertiary hospital in Ghana with UGIB.
Methods
This was a single‐center prospective cohort study involving 195 adults who presented with symptoms of UGIB from May 2022 to April 2023. Relevant baseline demographic and clinical characteristics were obtained. The cause of UGIB was determined as per findings at endoscopy. Patients were followed up for 6 weeks from admission, looking out for rebleeding, need for transfusion, length of hospital stay, and mortality.
Results
There were 145 (74.4%) males and 50 (25.6%) females, and the mean age ± SD was 51.4 ± 17.4 years. The main clinical presentations included melena (87.2%), hematemesis (69.7%) and postural dizziness (73.8%). The commonest findings at endoscopy were esophageal and gastric varices (33.3%), erosive gastritis and duodenitis (27.7%) and peptic ulcers (21.5%). The median length of hospital stay (IQR) was 7 days (5 days). 70.8% required whole blood transfusion with a median (IQR) of 2 units (2 units). The 6‐week mortality and rebleeding rates were 17.4% and 7.2%, respectively.
Conclusion
Variceal bleeding was the most common cause of UGIB at the emergency. One out of every fourteen patients that recover from acute UGIB may rebleed within the first six weeks one out of every six patients who present with acute UGIB may die within the succeeding six weeks.
Keywords: Ghana, mortality, upper gastrointestinal bleeding, varices
1. Introduction
Upper gastrointestinal bleeding (UGIB) is a potentially life‐threatening emergency that remains a common cause of hospitalization. It is defined as bleeding from a gastrointestinal (GI) source proximal to the ligament of Treitz [1]. The annual incidence ranges from 50 to 150 per 100 000 population, and it shows an overall mortality rate of 6%–10%, which can be up to 30% in sub‐Saharan Africa (SSA) [2, 3, 4]. The annual worldwide incidence of hospitalization for UGIB is approximately one per thousand adults, and this generally increases with age and is more common in men [5, 6]. UGIB is associated with significant morbidity, mortality, and resource utilization [7]. This burden is compounded in poor‐resource settings, where patients often pay out‐of‐pocket for care, thereby hampering appropriate early diagnosis to determine the location and severity of bleeding [8].
The causes and outcomes of UGIB vary in different geographic regions depending on the local demographics and socioeconomic characteristics [9]. Worldwide, peptic ulcer disease (PUD) is the most common cause of UGIB, accounting for approximately 50% of cases, followed by esophageal varices [10, 11]. However, in parts of SSA, esophageal varices are the most common cause of UGIB [3, 12]. Other causes include esophagitis, Mallory‐Weiss syndrome, Dieulafoy's lesion, gastric and duodenal erosions, upper gastrointestinal malignancies, and aorto‐enteric fistula. It usually presents as hematemesis and/or melena, hematochezia, postural dizziness, or shock depending on the severity. Clinical history and examination are essential in determining the risk factors and possible complications associated with UGIB. Initial management involves fluid resuscitation, hemodynamic monitoring, blood transfusion if indicated, and timely esophagogastroduodenoscopy (EGD) [13].
Central to the current management of UGIB is timely EGD, which provides both diagnostic and therapeutic opportunities. It is essential for anatomic and pathological diagnosis as well as hemostatic interventions and risk stratification. EGD is recommended within 24 h after initial resuscitation [14]. This recommendation may not be followed in some patients with UGIB, especially in SSA, where there are inadequate health resources and trained medical personnel.
International guidelines do not take into consideration these peculiarities and challenges; hence the need for more data on the etiology, endoscopic profile, and clinical outcome of patients of UGIB from this region.
2. Methods
2.1. Study Design
The study was a prospective cohort study involving 206 participants over a 1‐year period who presented with symptoms of UGIB at the Emergency Department (ED) between May 2022 and April 2023.
2.2. Study Setting
This study was conducted at the Emergency Department (ED) of Komfo Anokye Teaching Hospital (KATH). It is a 1200‐bed tertiary facility located in Kumasi, the capital of the Ashanti region in Ghana. It is the second largest teaching hospital in Ghana, and due to its central location, it serves as a referral center for 12 out of the 16 administrative regions. The study was approved by the Komfo Anokye Teaching Hospital Institutional Review Board (KATH IRB) with approval number KATH IRB/AP/022/22.
2.3. Study Participants
All adults, 18 years and above who presented with overt symptoms of UGIB characterized by hematemesis, melena, and/or hematochezia and consented were included in this study. Patients who presented with signs of generalized bleeding, trauma‐related UGIB, a history of ingestion of substances causing black discoloration of stool, and those who failed to give consent were excluded from this study.
2.4. Study Procedure
Based on the inclusion criteria, informed consent was obtained from all participants. Data on demographics, medical history including presenting symptoms, drug history including the use of antiplatelets, non‐steroidal anti‐inflammatory drugs (NSAIDs), anticoagulants like warfarin and direct‐acting oral anticoagulants, and comorbidities were obtained from participants or legal proxies in case they were unconscious using a structured questionnaire. Admission vital signs from the ED (blood pressure, pulse rate, respiratory rate and oxygen saturation) were recorded, and a physical examination was performed, looking out for signs like stigmata of chronic liver diseases that is, finger clubbing, gynecomastia, ascites, spider naevi, and jaundice that may point to the cause of UGIB and signs of severity like pallor and shock.
Laboratory investigations obtained included complete blood count, blood grouping and crossmatching, prothrombin time, international normalized ratio (INR), liver biochemical tests (serum transaminases, serum total bilirubin, serum protein, serum albumin), blood urea, electrolytes, and creatinine.
Initial resuscitation included insertion of two wide bore cannulae, Intravenous (IV) fluids; IV antibiotics and octreotide or IV proton pump inhibitor (PPI) based on clinical suspicion of variceal or non‐variceal bleeding respectively and when required; whole blood transfusion. Antibiotic prophylaxis was only given in cases of variceal bleeding. The indications for whole blood transfusion were hemoglobin levels less than 7 g/dL, hemoglobin levels less than 8 g/dL if the participant had an underlying cardiac condition or persistent hypotension even with adequate IV fluid resuscitation. Shock was defined as systolic BP below 90 mmHg.
Clinical severity was assessed using the pre‐endoscopic Rockall score and Glasgow Blatchford score. EGD was performed for participants when they were stable. Based on the findings at EGD, the full Rockall Score was calculated.
All participants were followed up daily while on admission, then biweekly after discharge for up to 6 weeks post‐admission to assess their clinical outcomes including rebleeding, requirement of transfusion, length of admission, and mortality.
2.5. Statistical Analysis
Data was collated, checked, and analyzed using the Statistical Package for Social Sciences (SPSS version 26.0; Chicago, IL, USA). Results were enumerated in means and proportions at a 95% confidence interval (CI) and presented in tables, charts, or graphs. Simple descriptive statistics were used. Depending on the distribution of the data after assessing normality by the Shapiro–Wilk test, continuous variables were presented as mean (standard deviation, SD) or median (Interquartile range, IQR). Proportions and frequency tables were used to summarize categorical variables. Categorical variables and outcomes were analyzed using the Pearson χ 2 test and Fisher's exact test. The differences in median values between continuous variables and outcome groups were analyzed using the Mann–Whitney U test.
In the multivariate analysis, Cox regression analysis was computed on variables to determine the independent risk of 6‐week mortality. A confidence interval of 95% was used, and p < 0.05 was considered statistically significant.
3. Results
3.1. Demographic and Clinical Characteristics of Participants
During the study, there were 23,231 admissions at the ED, of which 215 (0.9%) presented with UGIB. Out of the 215, 206 (95.8%) met the inclusion criteria. The participants/legal proxies failed to give consent for the study in 9 (4.2%) patients; hence, they did not meet the inclusion criteria. Endoscopy could not be performed in 11 (5.3%) of the cases because participants died during the resuscitation process and were excluded because the cause of UGIB could not be determined objectively.
Based on the endoscopy findings of 195 participants, 68 (34.9%) participants had variceal bleeding and 127 (65.1%) had non‐variceal bleeding. Males, 145 (74.4%) were predominant with a male‐to‐female ratio of 2.9:1. There was a statistically significant difference in the gender distribution between the two groups of variceal and non‐variceal bleeding (p = 0.038). The mean age (±SD) was 51.4 years (±17.4) and showed a statistically significant difference between the variceal bleeding group and the non‐variceal bleeding group 9 (p < 0.001).
Patients with variceal bleeding were more likely to present with malena 64/68 (94.1%) and hematochezia 106/127 (83.5%) compared to patients with non‐variceal bleeding (p = 0.042 and 0.044 respectively). There were no statistically significant differences among the distribution of symptoms of collapse 20 (10.3%), hematemesis 136 (69.7%) and postural dizziness 144 (73.8%) among the groups with variceal and non‐variceal bleeding (p = 0.330, 1.000 and 0.610 respectively).
Patients with variceal bleeding were more likely to have a past medical history of chronic liver disease 59/68 (86.8%) than patients with non‐variceal bleeding 11/127 (8.7%) with a significant p < 0.001. History of Jaundice was more common in patients with variceal bleeding 52/68 (76.5%) than patients with non‐variceal bleeding 12/127 (9.4%) with a significant p < 0.001. Peptic ulcer disease was significantly (p < 0.001) found in patients with non‐variceal bleeding 62/127 (48.8%) than in variceal bleeding 9/68 (13.2).
Aspirin/antiplatelets/NSAID use was 69 (35.4%) and their distribution among participants with variceal and non‐variceal bleeding showed statistically significant differences (p ≤ 0.001).
At examination, 124 (63.6%) had tachycardia, 32 (16.4%) came in with hypotension, 177 (90.8%) had conjunctival pallor, 123 (63.1%) had epigastric tenderness, and 60 (30.8%) had scleral jaundice. Almost half, 79 (40.5%), had ascites; 43 (22.1%) had splenomegaly; 24 (12.3%) had flapping tremors; and 69 (35.56%) tested positive for the rapid urease test for Helicobacter pylori infection. Scleral jaundice, the presence of ascites, splenomegaly, flapping tremors, and Helicobacter pylori infection showed significant differences between the groups with variceal and non‐variceal bleeding (all p < 0.001) (Table 1).
TABLE 1.
Clinico‐demographic characteristics of study participants.
| Variable | Variceal N = 68 | Non‐variceal N = 127 | Overall N = 195 | p |
|---|---|---|---|---|
| Age (years), mean ± SD | 44.5 (11.1) | 55.2 (19.0) | 51.4 (17.4) | < 0.001 |
| Gender | ||||
| Male | 57 (83.8) | 88 (69.3) | 145 (74.4) | 0.038 |
| Female | 11 (16.2) | 39 (30.7) | 50 (25.6) | |
| Marital status | ||||
| Single | 20 (29.4) | 22 (17.3) | 42 (21.5) | |
| Married | 41 (60.3) | 87 (68.5) | 128 (65.6) | 0.217 |
| Divorced | 3 (4.4) | 5 (3.9) | 8 (4.1) | |
| Widow/Widower | 4 (5.3) | 13 (10.2) | 17 (8.7) | |
| Hospital referral | 60 (91.2) | 106 (83.5) | 168 (86.2) | 0.191 |
| National health insurance | 58 (85.3) | 115 (90.6) | 173 (88.7) | 0.342 |
| Symptoms | ||||
| Collapse | 9 (13.2) | 11 (8.7) | 20 (10.3) | 0.330 |
| Hematemesis | 47 (69.1) | 89 (70.1) | 136 (69.7) | 1.000 |
| Melena | 64 (94.1) | 106 (83.5) | 170 (87.2) | 0.042 |
| Hematochezia | 17 (25.0) | 16 (12.6) | 33 (16.9) | 0.044 |
| Postural dizziness | 52 (76.5) | 92 (72.4) | 144 (73.8) | 0.610 |
| Past medical history | ||||
| Peptic ulcer disease | 9 (13.2) | 62 (48.8) | 71 (36.4) | < 0.001 |
| Jaundice | 52 (76.5) | 12 (9.4) | 64 (32.8) | < 0.001 |
| Chronic liver disease | 59 (86.8) | 11 (8.7) | 70 (35.9) | < 0.001 |
| Renal failure | 5 (7.4) | 12 (9.4) | 17 (8.7) | 0.792 |
| Cardiac disease | 4 (5.9) | 19 (15.0) | 23 (11.8) | 0.066 |
| Malignancy | 5 (7.4) | 5 (3.9) | 10 (5.1) | 0.322 |
| Stroke | 0 (0.0) | 6 (4.7) | 6 (3.1) | |
| Schistosoma infection | 3 (4.4) | 0 (0.0) | 3 (1.5) | |
| Drug history | ||||
| Aspirin/Antiplatelet/NSAID use | 5 (7.4) | 64 (50.4) | 69 (35.4) | < 0.001 |
| Anticoagulants | 0 (0.0) | 4 (3.1) | 4 (2.1) | 0.300 |
| Steroid Use | 0 (0.0) | 5 (3.9) | 5 (2.6) | 0.165 |
| Herbal preparation use | 21 (30.9) | 33 (26.0) | 54 (27.7) | 0.504 |
| Social history | ||||
| Cigarette smoking | 2 (2.9) | 9 (7.1) | 11 (5.6) | 0.335 |
| Alcohol use | 27 (39.7) | 40 (31.5) | 67 (34.4) | 0.271 |
| Physical examination | ||||
| Tachycardia | 46 (67.6) | 78 (61.4) | 124 (63.6) | 0.437 |
| Hypotension (blood pressure < 90 mmHg) | 15 (22.1) | 17 (13.4) | 32 (16.4) | 0.155 |
| Conjunctiva pallor | 66 (97.1) | 111 (87.4) | 177 (90.8) | 0.035 |
| Sclera jaundice | 47 (69.1) | 13 (10.2) | 60 (30.8) | < 0.001 |
| Epigastric tenderness | 39 (57.4) | 84 (66.1) | 123 (63.1) | 0.276 |
| Ascites | 62 (91.2) | 17 (13.4) | 79 (40.5) | < 0.001 |
| Splenomegaly | 37 (54.4) | 6 (4.7) | 43 (22.1) | < 0.001 |
| Flapping Tremor | 21 (30.9) | 3 (2.4) | 24 (12.3) | < 0.001 |
| Helicobacter pylori infection | 12 (17.9) | 57 (44.9) | 69 (35.6) | < 0.001 |
Note: Bold values indicate statistically significant p‐values.
Abbreviation: NSAID = non‐steroidal anti‐inflammatory drugs.
3.2. Biochemical, Hematological and Risk Profile of Participants
The median hemoglobin level (IQR) was 6.8 g/dL (3.7 g/dL), platelet count (×109/L) (IQR) of 178 (147), INR(IQR) of 1.0 (0.4), alanine aminotransferase (IQR) of 44.5 U/L (39.5 U/L), aspartate aminotransferase (IQR) of 42.4 U/L (69.5 U/L), serum bilirubin (IQR) of 16.0 mmol/L (24.2 mmol/L), serum albumin (IQR) of 27.2 g/L (11.1 g/L), Glasgow Blatchford score (IQR) of 11 (4) and full Rockall score (IQR) of 5 (4). They all showed significant differences between the groups with variceal and non‐variceal bleeding, with p‐values of 0.016, < 0.001, < 0.001, < 0.001, < 0.001, < 0.001, < 0.001, < 0.001, < 0.001, and < 0.001 respectively (Table 2).
TABLE 2.
Biochemical, hematological, and risk profile of participants.
| Biochemical test/risk score | Variceal | Non‐variceal | Overall | p |
|---|---|---|---|---|
| Hemoglobin (g/dL), median (IQR) | 6.1 (3.6) | 7.0 (3.5) | 6.8 (3.7) | 0.016 |
| Platelet (×109/L), median (IQR) | 102.5 (105.0) | 210.0 (122.0) | 178 (147) | < 0.001 |
| WBC (×109/L), median (IQR) | 7.7 (6.8) | 8.0 (5.4) | 7.8 (5.8) | 0.080 |
| INR, median (IQR) | 1.4 (0.7) | 0.9 (0.3) | 1.0 (0.4) | < 0.001 |
| Serum urea (mmol/L), median (IQR) | 7.9 (3.9) | 7.4 (5.8) | 7.5 (4.9) | 0.696 |
| Serum creatinine (μmol/L), median (IQR) | 93.5 (55.3) | 94.0 (39.0) | 94.0 (45.0) | 0.659 |
| Alanine aminotransferase (U/L), median (IQR) | 63.3 (69.5) | 24.3 (23.7) | 44.5 (39.5) | < 0.001 |
| Aspartate aminotransferase (U/L), median (IQR) | 94.2 (272.6) | 32.8 (33.4) | 42.4 (69.5) | < 0.001 |
| Serum bilirubin (μmol/L), median (IQR) | 28.7 (43.3) | 12.1 (13.0) | 16.0 (24.2) | < 0.001 |
| Serum albumin (g/L), median (IQR) | 22.1 (7.3) | 29.4 (10.5) | 27.2 (11.1) | < 0.001 |
| Full rockall score, median (IQR) | 7.0 (2.0) | 4.0 (2.0) | 5.0 (4.0) | < 0.001 |
| Glasgow blatchford score, median (IQR) | 12.5 (4.0) | 10.0 (4.0) | 11.0 (4.0) | < 0.001 |
Note: Bold values indicate statistically significant p‐values.
Abbreviations: INR = international normalized ratio, IQR = interquartile range, PCV = packed cell volume, WBC = white blood cells.
3.3. Endoscopic Findings
Out of 206 participants, 195 (94.7%) underwent upper gastrointestinal endoscopy. The remaining 11 (5.3%) did not undergo upper gastrointestinal endoscopy because they were not stable and subsequently died before the procedure could be done. Most of the participants had their endoscopy from 24 to 48 h into the admission 73 (35.4%) (Figure 1).
FIGURE 1.
Time to endoscopy of participants.
These findings at endoscopy were esophageal and gastric varices 65 (33.3%), gastritis 51 (26.2%), gastric ulcers 27 (13.8%), duodenal ulcers 15 (7.7%), gastric neoplasm 12 (6.2%), esophagitis seven (3.6%). Others were 3 (1.5%) each for esophageal neoplasm and duodenitis/erosions, portal hypertensive gastropathy 2 (1.0%), Mallory‐Weiss tear and duodenal neoplasm 1 (0.5%) each. 8 (4.1%) of the participants had normal findings (Table 3).
TABLE 3.
Endoscopic findings of participants.
| Endoscopic findings of participants | Frequency N = 195 (percent) |
|---|---|
| Esophageal and gastric varices | 65 (33.3) |
| Gastric ulcers | 27 (13.8) |
| Duodenal ulcers | 15 (7.7) |
| Gastritis | 51 (26.2) |
| Duodenitis | 3 (1.5) |
| Gastric neoplasm | 12 (6.2) |
| Duodenal neoplasm | 1 (0.5) |
| Esophagitis | 7 (3.6) |
| Mallory Weiss tear | 1 (0.5) |
| Normal findings | 8 (4.1) |
| Esophageal neoplasm | 3 (1.5) |
| Portal Hypertensive gastropathy | 2 (1.0) |
3.4. Outcomes
The overall median length of hospital stay (IQR) was 7 days (5 days). This was shorter than the group with variceal bleeding, 9 days (4 days). Most of the participants, 138 (70.8%) received whole blood transfusion with a median (IQR) of 2 units (2 units). In the group with variceal bleeding, 55 (80.9%) received whole blood transfusion compared to 83 (64.5%) in the group with non‐variceal bleeding (p‐value 0.031). A few participants, 14 (7.2%) had rebleeds, with the majority occurring in the group with variceal bleeding 10 (14.7%) compared to 4 (3.1%) in the group with non‐variceal bleeding (p = 0.004). The overall 6‐week mortality rate was 34 (17.4%); the 6‐week mortality rate was 25 (36.8%) in the group with variceal bleeding compared with the group with non‐variceal bleeding, 4 (3.1%) (p < 0.001) Table 4.
TABLE 4.
Outcomes.
| Outcome | Variceal N = 68 | Non‐variceal N = 127 | Overall N = 195 | p |
|---|---|---|---|---|
| 6‐week mortality | 25 (36.8) | 9 (7.1) | 34 (17.4) | < 0.001 |
| Blood transfusion | 55 (80.9) | 83 (65.4) | 138 (70.8) | 0.031 |
| Rebleeding | 10 (14.7) | 4 (3.1%) | 14 (7.2) | 0.004 |
| Duration of hospital stay, median (IQR) | 9 (4.0) | 7 (5.0) | 7 (5) | 0.121 |
Note: Bold values indicate statistically significant p‐values.
3.5. Association Between Clinical Parameters and 6‐Week Mortality
Factors that had a significant relationship with mortality in the univariate analysis were presenting symptom of collapse (p = 0.033), history of PUD (p < 0.001), jaundice(p < 0.001), liver failure (p < 0.001) and malignancy (p < 0.001); the presence of scleral jaundice (p < 0.001), ascites (p < 0.001) and splenomegaly (p = 0.007); raised WBC (p = 0.001), INR (p < 0.001), serum creatinine (p = 0.039), alanine aminotransferase (p < 0.001), aspartate aminotransferase (p < 0.001), total serum bilirubin (p < 0.001), low serum albumin levels (p < 0.001), Helicobacter pylori infection, (p = 0.026), variceal bleeding (p < 0.001), high Glasgow Blatchford (p < 0.001) and Rockall scores (p < 0.001) (Tables 5 and 6).
TABLE 5.
Association between clinical characteristics and mortality (categorical data).
| Variable | Response | Survived (%) N = 161 | Died (%) N = 34 | p |
|---|---|---|---|---|
| Gender | Male | 119 (73.9) | 26 (76.5) | 0.832 |
| Female | 42 (26.1) | 8 (23.5) | ||
| Collapse | Yes | 14 (8.7) | 9 (39.1) | 0.033 |
| No | 147 (80.3) | 36 (19.7) | ||
| Hematemesis | Yes | 114 (79.2) | 30 (20.8) | 0.592 |
| No | 47 (75.8) | 15 (24.2) | ||
| Melena | Yes | 138 (76.7) | 42 (23.3) | 0.174 |
| No | 23 (88.5) | 3 (11.5) | ||
| Hematochezia | Yes | 26 (76.5) | 8 (23.5) | 0.795 |
| No | 135 (78.5) | 37 (21.5) | ||
| History of peptic ulcer disease | Yes | 68 (94.4) | 4 (5.6) | |
| No | 93 (69.4) | 41 (30.6) | < 0.001 | |
| History of Jaundice | Yes | 42 (56.8) | 32 (43.2) | < 0.001 |
| No | 119 (90.2) | 13 (9.8) | ||
| History of liver failure | Yes | 45 (56.3) | 35 (43.8) | < 0.001 |
| No | 116 (92.1) | 10 (7.9) | ||
| History of kidney failure | Yes | 11 (61.1) | 7 (38.9) | 0.067 |
| No | 150 (79.8) | 38 (20.2) | ||
| History of heart disease | Yes | 20 (83.3) | 4 (16.7) | 0.414 |
| No | 141 (77.5) | 41 (22.5) | ||
| History of malignancy | Yes | 2 (18.2) | 9 (81.8) | < 0.001 |
| No | 159 (81.5) | 36 (18.5) | ||
| Sclera jaundice | Present | 37 (53.6) | 13 (9.5) | < 0.001 |
| Absent | 124 (90.5) | 32 (46.4) | ||
| Ascites | Present | 53 (59.6) | 36 (40.4) | < 0.001 |
| Absent | 108 (92.3) | 9 (7.7) | ||
| Splenomegaly | Present | 30 (63.8) | 17 (36.2) | 0.007 |
| Absent | 131 (82.4) | 28 (17.6) | ||
| Endoscopy within 24 h | Yes | 24 (14.9) | 6 (17.6) | 0.793 |
| No | 137 (85.1) | 28 (82.4) | ||
| Helicobacter pylori infection | Yes | 63 (39.1) | 6 (18.2) | 0.027 |
| No | 98 (60.9) | 27 (81.8) | ||
| Etiology of bleeding | Variceal | 43 (26.7) | 25 (73.5) | < 0.001 |
| Nonvariceal | 118 (73.3) | 9 (26.5) |
Note: Bold values indicate statistically significant p‐values.
TABLE 6.
Association between clinical and laboratory characteristics and mortality (Continuous data).
| Variable | Survived N = 161 median (IQR) | Died N = 34 median (IQR) | p |
|---|---|---|---|
| Age (years) | 52 (24.5) | 48.0 (23.0) | 0.246 |
| PCV (%) | 19.4 (9.9) | 19.1 (10.0) | 0.345 |
| Hemoglobin (g/dL) | 6.8 (3.5) | 6.4 (3.6) | 0.545 |
| Platelet (×109/L) | 184 (145.5) | 148.0 (224.0) | 0.756 |
| WBC (×109/L) | 7.4 (5.7) | 10.6 (8.5) | 0.001 |
| INR | 1.0 (0.3) | 1.7 (1.0) | < 0.001 |
| Serum urea (mmol/L) | 7.4 (3.6) | 9.0 (13.9) | 0.249 |
| Serum creatinine (μmol/L) | 94.0 (36.0) | 102.0 (128.0) | 0.153 |
| Alanine aminotransferase (U/L) | 28.0 (29.7) | 76.3 (152.4) | < 0.001 |
| Aspartate aminotransferase (U/L) | 38.3 (54.8) | 193.0 (362.2) | < 0.001 |
| Total bilirubin (μmol/L) | 15.0 (15.1) | 51.0 (78.1) | < 0.001 |
| Serum albumin (g/L) | 28.5 (12.2) | 22.4 (7.5) | < 0.001 |
| Glasgow Blatchford Score, median (IQR) | 11.0 (5.0) | 13.5 (6.0) | < 0.001 |
| Full Rockall Score, median (IQR) | 4.0 (3.0) | 7 (2.0) | < 0.001 |
Note: Bold values indicate statistically significant p‐values.
Abbreviations: INR = international normalized ratio, IQR = interquartile range, PCV = packed cell volume, WBC = white blood cells.
In multivariate regression analysis, history of disseminated cancer cos hazard ratio of 7.7 (CI 2.2–26.6, p = 0.001), presence of sclera jaundice cos hazard ratio of 4.8 (CI 1.2–18.3, p = 0.022), flapping tremor cos hazard ratio of 3.3 (CI 1.2–8.7, p = 0.017), increasing white blood cell count cos hazard ratio of 1.1 (CI 1.1–1.2, p = 0.002), and Rockall Score cos hazard ratio of 1.5 (CI 1.1–1.9, p = 0.007) were independent predictors of mortality (Tables 6 and 7).
TABLE 7.
Multivariate regression of predictors of mortality.
| Variable | p | HR | 95.0% CI for HR | |
|---|---|---|---|---|
| Lower | Upper | |||
| Collapse | 0.405 | 0.6 | 0.2 | 1.9 |
| History of PUD | 0.361 | 0.5 | 0.1 | 2.1 |
| History of jaundice | 0.230 | 0.5 | 0.2 | 1.6 |
| History of liver failure | 0.486 | 1.7 | 0.4 | 7.3 |
| Disseminated cancer | 0.001 | 7.7 | 2.2 | 26.6 |
| Sclera jaundice | 0.022 | 4.8 | 1.2 | 18.3 |
| Ascites | 0.662 | 0.6 | 0.1 | 6.3 |
| Splenomegaly | 0.039 | 0.4 | 0.1 | 1.0 |
| Flapping tremor | 0.017 | 3.3 | 1.2 | 8.7 |
| White blood Cell | 0.002 | 1.1 | 1.1 | 1.2 |
| INR | 0.289 | 1.5 | 0.7 | 2.9 |
| Serum creatinine | 0.207 | 1.0 | 0.9 | 1.0 |
| Alanine aminotransferase | 0.244 | 1.0 | 1.0 | 1.0 |
| Aspartate aminotransferase | 0.037 | 1.0 | 1.0 | 1.0 |
| Total bilirubin | 0.493 | 1.0 | 0.9 | 1.0 |
| Serum albumin | 0.026 | 0.9 | 0.9 | 1.0 |
| Etiology of UGIB | 0.833 | 1.2 | 0.2 | 8.4 |
| Helicobacter pylori infection | 0.568 | 1.4 | 0.4 | 4.2 |
| Rockall Score | 0.007 | 1.5 | 1.1 | 1.9 |
| Glasgow Blatchford Score | 0.101 | 1.2 | 1.0 | 1.4 |
Note: Bold values indicate statistically significant p‐values.
Abbreviations: HR = Hazard ratio, INR = international normalized ratio, PUD = peptic ulcer disease, UGIB = upper gastrointestinal bleed.
4. Discussion
In this study, we found an overall 6‐week rebleeding rate of 7.2% and a mortality rate of 17.4%. Rebleeding was more common in patients with variceal bleeding compared to those with non‐variceal bleeding. The 6‐week mortality rate was also more common in patients with variceal bleeding compared with the group with non‐variceal bleeding. The mean age of presentation was 51.4 ± 17.4 years. UGIB was more common in males than females, with a male‐to‐female ratio of 3:1. The most common clinical presentations were melena, hematemesis, and postural dizziness. The median hemoglobin concentration was 6.8 g/dL with a median platelet count of 178 (×109/L). We found esophageal and gastric varices and erosive gastritis as the commonest causes of UGIB.
The risk of UGIB has been shown to increase with increasing age in most developed countries. Van Leerdam et al. showed that as many as 70% of acute upper GI bleed episodes occur in patients older than 60 years [15]. In France, Thiebaud et al. found that the median age at presentation was 66 years [16]. However, most studies in West Africa have found the median age to be below 60 years. In Ghana, Nkansah et al. found the median age at presentation to be 54 years [17]. Other studies in Nigeria by Ajayi et al., Rukewe et al., Alatise et al., Manko et al., and Fatusi et al. had ages at presentation below 60 years [8, 12, 18, 20]. The difference in predominant age groups among developed and developing nations may be due to the different risk factors associated with UGIB. Whereas risk factors of UGIB in developed countries are associated with increased NSAID and antiplatelet use among the aged population, chronic liver disease from hepatitis B, which mainly occurs at a younger age as a result of vertical and horizontal transmission, may be the driving force in developing countries.
Gender and its related risk behaviors may influence the incidence of UGIB. Most studies have shown male sex is a risk factor for upper GI bleeding [21, 22, 23]. Nkansah et al., in their study at a district hospital, Ghana, had 60.8% of respondents being male [17]. In Zaria, Nigeria, Manko et al. also mirrored similar male preponderance with 72.9% in a previous study at Ahmadu Bello University Teaching Hospital [12]. In this study also, there was a male preponderance, with a male to female ratio of 3:1. Plausible explanation for this observed finding could be due to increased risky behaviors such as alcohol misuse, smoking, and intravenous drug use and its associated transmission of hepatitis B and C viruses among males. Also, these risk behaviors may put males at high cardiovascular risk and the need for antiplatelet therapy.
UGIB commonly presents as hematemesis (either red or coffee‐ground vomiting) and/or melena stools (black, tarry stool) signifying that the bleeding is proximal to the ligament of Treitz. Laine et al. have demonstrated that the severity of bleeding and mortality rates are similar whether the patient presents with bloody emesis or coffee‐ground emesis [24]. Melena may be seen with variable degrees of blood loss, being observed with as low as 50 mL of blood, and the color change is due to the long transit time observed [25]. Some patients may present with hematochezia from massive upper GI bleeding [26]. Acute blood loss may lead to hemodynamic instability and consequent orthostatic hypotension. Thus, anemia is a common clinical finding that may require a blood transfusion. The median hemoglobin concentration observed in our study was 6.8 g/dL, with 70.8% of the participants requiring blood transfusion. In Nigeria, Aletise et al. showed that 83.6% of patients presenting with UGIB required blood transfusion in the course of their hospital admission [19]. In Tanzania, Rajan et al., in their cohort study, found that 48.8% of the participants with UGIB had blood transfusion [4]. In considering hemotransfusion following acute upper GI bleeding, two strategies are employed: the liberal (transfusion when hemoglobin concentration fell below 10.0 g/dL) and the restrictive (transfusion when hemoglobin concentration fell below 8.0 g/dL) transfusion strategy [27]. However, various clinical trials have shown a better outcome with the restrictive strategy [27, 28].
EGD remains the most effective modality in localizing and treating the etiology of UGIB, especially when it is done within 24 h. Of the 206 participants, 5.3% did not have endoscopy, 14.6% had endoscopy done within 24 h, 35.4% from 24 to 48 h, 21.4% from 48 to 72 h, and 23.3% after 72 h. This is lower than the study by Jaka et al. in Tanzania, which had 53.3% of endoscopies being done within 24 h, and Hearnshaw et al. in the United Kingdom found 50.0% of endoscopies being done within 24 h [29, 30]. The source of bleeding could not be identified in 4.1% of OGD done, which is relatively higher than in the study by Jaka et al. in Tanzania (3.1%) but significantly lower than Ajayi et al. (10.4%) and Mustapha et al. (11.9%) in Nigeria [18, 29, 31]. With erosive mucosal lesions healing very fast, the yield of OGD is reduced significantly if it is not done within 24 h of presentation, and this may explain why the cause of UGIB could not be identified in some cases.
The commonest finding at EGD was esophageal and gastric varices, followed by erosive gastritis, gastric ulcers, and duodenal ulcers. Archampong et al. in their retrospective study at Korle Bu Teaching Hospital in Ghana also found esophageal and gastric varices as the commonest findings at EGD [32]. However, Nkansah et al. in a study at a district hospital in Ghana showed that PUD is the commonest cause of UGIB [17]. A possible explanation for this difference could be that this study, together with the one by Archampong et al., was carried out in tertiary centers that have facilities to manage the most severe forms of CLD and as such received referrals from all over the country, compared to the one by Nkansah et al. The results were also consistent with most studies in sub–Saharan Africa, India, and Nepal [3, 12, 29, 31, 33, 34, 35, 36, 37, 38].
Majority of the studies in the developed world had PUD as the commonest cause of UGIB [39, 40, 41]. These differences may probably be due to differences in predominant risk factors. While NSAIDS and aspirin use are prevalent in Western countries due to the aging population and are the main risk factors of UGIB in developed countries, CLD from hepatitis B and schistosomiasis with their complications of portal hypertension are prevalent in developing countries and are the main drivers of UGIB.
The study also found a lower prevalence of malignancy as the cause of UGIB: gastric malignancies, 12 (6.2%); esophageal malignancy, three (1.5%) and duodenal malignancy, one (0.5%). This may be explained by the lower prevalence of risk factors like smoking (5.3%) in the study population.
The primary clinical outcomes in this study were the need for blood transfusion, duration of admission, 6‐week rebleeding, and mortality rates. The clinical outcomes varied from one study to another, and factors responsible for this variation may include the age of the patient, the severity of underlying conditions like CLD, malignancies, and other comorbidities, the volume of blood loss necessitating the need for multiple blood transfusions, and availability of expertise and facilities for the management of UGIB.
The median length of admission was 7 days and ranged from one to 30 days. This was longer than Alatise et al. (4.5 days) in Nigeria but similar to Jaka et al. (8 days) in Tanzania. Factors like the severity of underlying medical conditions like CLD, availability of expertise and facilities for the management of UGIB, and the need for transfusion of blood products, which may not be readily available, can explain these variations among the different studies. The 6‐week rebleeding rate in this study was 7.2%, similar to Moledina et al. (7.1%) in Tanzania [34]. Lee et al. also found the similar rebleeding rate (7.3%) in South Korea [40]. Full Rockall score ≥ 5 was an independent predictor of rebleeding (p = 0.033) and was 2.4 times more likely to lead to rebleeding.
The 6‐week mortality rate was 17.4%. Most of the studies in sub‐Saharan Africa had mortality rates ranging between 4.0% and 33.5% [3, 8, 18, 19, 29, 31, 34, 42]. The studies in developed nations had mortality rates below 4% [40, 43]. The high mortality rate in this study and in other sub‐Saharan African countries could be explained by delayed or more severe presentation to a referral facility, compounded by inadequate urgent endoscopic intervention and other interventional measures like blood transfusion. About a quarter of the mortalities in this study happened within the first 24 h of presentation, suggesting how critical early intervention is in managing UGIB.
4.1. Study Strengths and Limitations
This study was a prospective one where many variables that could potentially predict adverse outcomes were assessed. All the participants were followed up in this study.
The main drawback of this study is that it was conducted in a single tertiary facility, making it difficult to generalize its findings. The patient population may be different in smaller, lower‐capacity facilities. Interventions were limited by the absence of cyanoacrylate glue, hemostatic powders or gels, over‐thethe‐scope clips, and argon plasma coagulation in the hospital.
5. Conclusion
Esophageal varices, gastric varices, and erosive gastritis are the commonest causes of acute UGIB in Ghana. One out of every fourteen patients that recover from acute UGIB may rebleed within the first 6 weeks. Approximately one out of every six patients who present with acute UGIB may die within the succeeding 6 weeks; therefore, there is a need to increase public and physician awareness of associated outcomes and risk factors of common causes of UGIB to help with prevention and early presentation. Patients with background CLD should be screened for varices endoscopically, and prophylactic treatment should be offered for high‐risk varices. There should be a protocol for the management of UGIB with clear guidelines to improve outcomes and make the best use of scarce medical resources peculiar to this region. Efforts should be made to improve and make diagnostic and therapeutic endoscopy services accessible and affordable by training more endoscopists and other team members and providing appropriate and sufficient materials to work with.
Ethics Statement
The study was approved by the Komfo Anokye Teaching Hospital Institutional Review Board (KATH IRB) with approval number KATH IRB/AP/022/22.
Consent
Written informed consent was obtained from the patients.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgments
We would like to thank the staff of the emergency medicine directorate for their support toward this study.
Funding: The authors received no financial support for the research, authorship, and publication of this study.
Contributor Information
Nyantakyi Adu‐Darko, Email: nyantache@gmail.com.
Solomon Gyabaah, Email: solomongyabaah53@gmail.com.
Data Availability Statement
Data is available upon reasonable request to the corresponding author.
References
- 1. Ghosh S., Watts D., and Kinnear M., “Management of Gastrointestinal Haemorrhage,” Postgraduate Medical Journal 78 (2002): 4–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Mathew P., Kanni P. Y., Gowda M., Uppalapati S., Garg A., and Ansari J., “Retrospective Study of Clinical Profile, Endoscopic Profile and in Hospital Mortality in Acute Upper Gastrointestinal Bleeding in Tertiary Care Centre in South India,” International Journal of Contemporary Medical Research 6, no. 8 (2019): H1–H5. [Google Scholar]
- 3. Mulima G., Qureshi J. S., Shores C., Tamimi S., Klackenberg H., and Andrén‐Sandberg Å., “Upper Gastrointestinal Bleeding at a Public Referral Hospital in Malawi,” Surgical Science 05, no. 11 (2014): 501–507. [Google Scholar]
- 4. Rajan S. S., Sawe H. R., Iyullu A. J., et al., “Profile and Outcome of Patients With Upper Gastrointestinal Bleeding Presenting to Urban Emergency Departments of Tertiary Hospitals in Tanzania,” BMC Gastroenterology 19, no. 1 (2019): 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Sostres C. and Lanas A., “Epidemiology and Demographics of Upper Gastrointestinal Bleeding: Prevalence, Incidence, and Mortality,” Gastrointestinal Endoscopy Clinics of North America 21, no. 4 (2011): 567–581. [DOI] [PubMed] [Google Scholar]
- 6. Longstreth G. F., “Epidemiology of Hospitalization for Acute Upper Gastrointestinal Hemorrhage: A Population–Based Study,” American Journal of Gastroenterology 90, no. 2 (1995): 206–210. [PubMed] [Google Scholar]
- 7. Esrailian E. and Gralnek I. M., “Nonvariceal Upper Gastrointestinal Bleeding: Epidemiology and Diagnosis,” Gastroenterology Clinics of North America 34 (2005): 589–605. [DOI] [PubMed] [Google Scholar]
- 8. Rukewe A., Otegbayo J. A., and Fatiregun A., “Clinical Characeristics and Outcome of Patients With Upper Gastrointestinal Bleeding at the Emergency Department of a Tertiary Hospital in Nigeria,” Annals of the Institute of Postgraduate Medicine 13, no. 2 (2015): 89–93. [PMC free article] [PubMed] [Google Scholar]
- 9. Theocharis G. J., Thomopoulos K. C., Sakellaropoulos G., Katsakoulis E., and Nikolopoulou V., “Changing Trends in the Epidemiology and Clinical Outcome of Acute Upper Gastrointestinal Bleeding in a Defined Geographical Area in Greece,” Journal of Clinical Gastroenterology 42, no. 2 (2008): 128–133. [DOI] [PubMed] [Google Scholar]
- 10. Dinesen L. and Benson M., “Managing Acute Upper Gastrointestinal Bleeding in the Acute Assessment Unit,” Clinical Medicine 12, no. 6 (2012): 589–593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Saleem S. and Thomas A. L., “Management of Upper Gastrointestinal Bleeding by an Internist,” Cureus 10 (2018): e2878. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Manko M., Bello A. K., Mohammed M. F., et al., “Demographic Profile and Endoscopic Findings Among Patients With Upper Gastrointestinal Bleeding in Ahmadu Bello University Teaching Hospital, Zaria, North‐Western Nigeria,” Nigerian Journal of Clinical Practice 23, no. 8 (2020): 1163–1166. [DOI] [PubMed] [Google Scholar]
- 13. Kumar N. L., Travis A. C., and Saltzman J. R., “Initial Management and Timing of Endoscopy in Nonvariceal Upper GI Bleeding,” Gastrointestinal Endoscopy 84, no. 1 (2016): 10–17. [DOI] [PubMed] [Google Scholar]
- 14. Laine L., Barkun A. N., Saltzman J. R., Martel M., and Leontiadis G. I., “ACG Clinical Guideline: Upper Gastrointestinal and Ulcer Bleeding,” American Journal of Gastroenterology 116, no. 5 (2021): 899–917. [DOI] [PubMed] [Google Scholar]
- 15. van Leerdam M. E., Vreeburg E. M., Rauws E. A. J., et al., “Acute Upper Gi Bleeding: Did Anything Change?,” American Journal of Gastroenterology 98, no. 7 (2003): 1494–1499. [DOI] [PubMed] [Google Scholar]
- 16. Thiebaud P. C., Yordanov Y., Galimard J. E., et al., “Management of Upper Gastrointestinal Bleeding in Emergency Departments, From Bleeding Symptoms to Diagnosis: A Prospective, Multicenter, Observational Study,” Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine 25, no. 1 (2017): 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Nkansah A. A., Duah A., and Alfonso Romero M., “Aetiology of Upper Gastrointestinal Bleeding in Patients Presenting at St. Dominic Hospital, Akwatia, Ghana,” International Journal of Medicaland Health Sciences Journal Home 8, no. 2 (2019): 22–27. [Google Scholar]
- 18. Ajayi A. O., Adegun P. T., Ajayi E. A., Solomon O. A., Adeoti A. O., and Akolawole M. A., “Aetiology and Management Outcome of Upper Gastrointestinal Bleeding in Adult Patients Presenting at Ekiti State University Teaching Hospital, Ado‐Ekiti, Nigeria,” Greener Journal of Medical Sciences 3, no. 20 (2013): 93–97. [Google Scholar]
- 19. Alatise O. I., Aderibigbe A. S., Adisa A. O., Adekanle O., Agbakwuru A. E., and Arigbabu A. O., “Management of Overt Upper Gastrointestinal Bleeding in a Low Resource Setting: A Real World Report From Nigeria,” BMC Gastroenterology 14, no. 1 (2014): 210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Fatusi A. O., “Pattern and Validity of Clinical Diagnosis of Upper Gastrointestinal Diseases in South‐West Nigeria,” African Health Sciences 6 (2006): 98–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Van Leerdam M. E., Vreeburg E. M., Rauws E. A. J., et al., “Acute Upper GI Bleeding: Did Anything Change? Time Trend Analysis of Incidence and Outcome of Acute Upper GI Bleeding Between 1993/1994 and 2000,” American Journal of Gastroenterology 98, no. 7 (2003): 1494–1499. [DOI] [PubMed] [Google Scholar]
- 22. Vora P., Pietila A., Peltonen M., Brobert G., and Salomaa V., “Thirty‐Year Incidence and Mortality Trends in Upper and Lower Gastrointestinal Bleeding in Finland,” JAMA Network Open 3, no. 10 (2020): e2020172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Sadiku E., Cuko L., Pasho I., et al., “Epidemiologic Data on Acute Gastrointestinal Bleeding in Albania: An Overview of Upper and Lower GI Bleeding,” Egyptian Liver Journal 14, no. 1 (2024): 1–7. [Google Scholar]
- 24. Laine L., Laursen S. B., Zakko L., et al., “Severity and Outcomes of Upper Gastrointestinal Bleeding With Bloody vs. Coffee‐Grounds Hematemesis,” American Journal of Gastroenterology 113, no. 3 (2018): 358–366. [DOI] [PubMed] [Google Scholar]
- 25. Doherty Gerard , “Current Diagnosis and Treatment Surgery,” accessed November 5, 2024, https://books.google.com/books/about/CURRENT_Diagnosis_and_Treatment_Surgery.html?id=a14yDQKv4vMC.
- 26. Laine L. and Shah A., “Randomized Trial of Urgent vs. Elective Colonoscopy in Patients Hospitalized With Lower GI Bleeding,” American Journal of Gastroenterology 105, no. 12 (2010): 2636–2641. [DOI] [PubMed] [Google Scholar]
- 27. Jairath V., Kahan B. C., Gray A., et al., “Restrictive Versus Liberal Blood Transfusion for Acute Upper Gastrointestinal Bleeding (TRIGGER): A Pragmatic, Open‐Label, Cluster Randomised Feasibility Trial,” Lancet 386, no. 9989 (2015): 137–144. [DOI] [PubMed] [Google Scholar]
- 28. Villanueva C., Colomo A., Bosch A., et al., “Transfusion Strategies for Acute Upper Gastrointestinal Bleeding,” New England Journal of Medicine 368, no. 1 (2013): 11–21. [DOI] [PubMed] [Google Scholar]
- 29. Jaka H., Koy M., Liwa A., et al., “A Fibreoptic Endoscopic Study of Upper Gastrointestinal Bleeding at Bugando Medical Centre in Northwestern Tanzania: A Retrospective Review of 240 Cases,” BMC Research Notes 5 (2012): 1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Hearnshaw S. A., Logan R. F. A., Lowe D., Travis S. P. L., Murphy M. F., and Palmer K. R., “Use of Endoscopy for Management of Acute Upper Gastrointestinal Bleeding in the UK: Results of a Nationwide Audit,” Gut 59, no. 8 (2010): 1022–1029. [DOI] [PubMed] [Google Scholar]
- 31. Mustapha S. K., Mustapha S., Ajayi N., and Shehu A., “Aetiology of Upper Gastrointestinal Bleeding in North‐Eastern Nigeria: A Retrospective Endoscopic Study,” Internet Journal of Third World Medicine 8 (2008): M74–82. [Google Scholar]
- 32. Archampong E. Q., “Management of Upper Gastrointestinal Bleeding,” 2013, http://ugspace.ug.edu.gh/handle/123456789/26489.
- 33. Alema O. N., Martin D. O., and Okello T. R., “Endoscopic Findings in Upper Gastrointestinal Bleeding Patients at Lacor Hospital, Northern Uganda,” African Health Sciences 12, no. 4 (2012): 518–521. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Moledina S. M. and Komba E., “Risk Factors for Mortality Among Patients Admitted With Upper Gastrointestinal Bleeding at a Tertiary Hospital: A Prospective Cohort Study,” BMC Gastroenterology 17, no. 1 (2017): 165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Banerjee A., Bishnu S., and Dhali G. K., “Acute Upper Gastrointestinal Bleed: An Audit of the Causes and Outcomes From a Tertiary Care Center in Eastern India,” Indian Journal of Gastroenterology 38, no. 3 (2019): 190–202. [DOI] [PubMed] [Google Scholar]
- 36. Dhande S. and Pichaimuthu A., “Endoscopic Profile of Acute Upper Gastrointestinal Bleed in Adults: A Tertiary Care Center‐Based Study in South India,” Journal of Datta Meghe Institute of Medical Sciences University 16, no. 2 (2021): 325–328. [Google Scholar]
- 37. Surendran M. and Kumar K. S., “Clinical and Endoscopic Profile of Upper Gastrointestinal Bleed: A Cross‐Sectional Study from a Tertiary Care Hospital in Southern India,” Journal of Clinical and Diagnostic Research 15, no. 3 (2021): OC14–OC17. [Google Scholar]
- 38. Adhikari S., Shah J., Rijal S., and House D. R., “Clinical Profile of Patients Presenting to Emergency With Upper Gastrointestinal Bleeding in a Tertiary Hospital of Nepal,” Journal of Patan Academy of Health Sciences 8, no. 2 (2021): 105–111. [Google Scholar]
- 39. Hreinsson J. P., Kalaitzakis E., Gudmundsson S., and Björnsson E. S., “Upper Gastrointestinal Bleeding: Incidence, Etiology and Outcomes in a Population‐Based Setting,” Scandinavian Journal of Gastroenterology 48, no. 4 (2013): 439–447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Lee H. A., Jung H. K., Kim T. O., et al., “Clinical Outcomes of Acute Upper Gastrointestinal Bleeding According to the Risk Indicated by Glasgow‐Blatchford Risk Score‐Computed Tomography Score in the Emergency Room,” Korean Journal of Internal Medicine 37, no. 6 (2022): 1176–1185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Wang C. Y., Qin J., Wang J., Sun C. Y., Cao T., and Zhu D. D., “Rockall Score in Predicting Outcomes of Elderly Patients With Acute Upper Gastrointestinal Bleeding,” World Journal of Gastroenterology 19, no. 22 (2013): 3466–3472. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Kalula S. Z., Swingler G. H., Louw J. A., and Chb M. B., “Clinical Predictors of Outcome in Acute Upper Gastro‐Intestinal Bleeding,” South African Medical Journal 93, no. 4 (2003): 286–290. [PubMed] [Google Scholar]
- 43. Wilcox C. M., Cryer B. L., Henk H. J., Zarotsky V., and Zlateva G., “Mortality Associated With Gastrointestinal Bleeding Events: Comparing Short‐Term Clinical Outcomes of Patients Hospitalized for Upper GI Bleeding and Acute Myocardial Infarction in a US Managed Care Setting,” Clinical and Experimental Gastroenterology 2 (2009): 21–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data is available upon reasonable request to the corresponding author.