Abstract
Background and objective
Upper gastrointestinal bleeding (UGIB) is a major cause of morbidity and death on a global scale, with etiology differing between populations. Variceal bleeding predominates in patients with chronic liver disease (CLD), and the nonvariceal causes include peptic ulcer disease (PUD), malignancy, and corrosive injuries. This audit aimed to assess the clinical and endoscopic features, therapeutic interventions, and outcomes of patients presenting with UGIB over a 4-year period at a tertiary care center.
Materials and methods
This clinical audit comprised 3,448 patients who attended the hospital with UGIB between 2022 and 2025, and data were collected retrospectively. Patients of all ages were included in the study, whereas those with incomplete records, prior major gastrointestinal surgery, or unrelated coagulopathies were excluded. Data on demographics, etiology, clinical symptoms (hematemesis, melena, and hematochezia), endoscopic findings, interventions, and outcomes were obtained from hospital records. Endoscopic evaluation was carried out using high-definition video endoscopes. Therapeutic interventions consisted of variceal band ligation, sclerotherapy, adrenaline injection, and hemoclip placement. Data were analyzed using the statistical package SPSS (version 26). Quantitative variables were reported as mean and standard deviation, and categorical variables as frequencies and percentages. Chi-square test and analysis of variance were performed for year-wise comparisons. Statistical significance was determined at the level of p < 0.05.
Results
Hematemesis was the most common clinical presentation, observed in 2,855 (82.8%) of 3,448 patients, followed by melena in 408 (11.8%). Variceal bleeding was predominant, esophageal varices were in 1,374 (39.8%), and gastric varices were in 378 (11.0%). Peptic ulcer disease was the most common nonvariceal cause, 724 (21.0%). Variceal band ligation was the most commonly performed intervention, 1,625 (47.1%). Discharge was done in 2,950 (85.5%), while 341 (9.9%) died. Year-wise analysis revealed consistent trends in presentation and intervention, with significant differences by age, gender, and etiology (p < 0.05).
Conclusion
Upper gastrointestinal bleeding in this cohort is mainly related to variceal bleeding, mainly in patients with CLD. Endoscopic evaluation and early therapeutic intervention are pivotal to management and are associated with high discharge rates. Regular audits are essential in order to optimize patient care and outcomes.
Keywords: Peptic ulcer disease, Upper gastrointestinal bleeding, Variceal bleeding
Introduction
Upper gastrointestinal bleeding (UGIB) is a common and potentially life-threatening medical emergency caused by bleeding proximal to the ligament of Treitz (which includes bleeding from the esophagus, stomach, or duodenum). Patients usually present with clinical features such as hematemesis, melena, or hematochezia, which reflect the location and rate of bleeding. Hematemesis is usually suggestive of brisk upper tract bleeding, while melena is more typical of a slower transit time of blood through the gastrointestinal tract.1 These presenting symptoms require timely assessment, given that UGIB is linked to substantial morbidity, health resource utilization, and mortality despite improvements in diagnostics and therapeutic strategies.2
The etiology of the UGIB spectrum can generally be categorized as variceal or nonvariceal bleeding. Variceal bleeding is due to portal hypertension within a setting of chronic liver disease (CLD) and is especially prevalent in 20th-century populations with high endemic rates of viral hepatitis.3 Nonvariceal upper gastrointestinal bleeding (NVUGIB) is usually caused by peptic ulcer disease (PUD), erosive gastritis, Mallory-Weiss tears, vascular anomalies, and gastrointestinal malignancies.4 The prevalence of these causes depends on geographic location, and variceal bleeding is frequently more common in regions with high hepatitis B virus (HBV) and hepatitis C virus (HCV) burden, and this can contribute significantly to the overall UGIB burden and can affect both the clinical presentation and clinical outcome of UGIB.5
The primary mode of management of UGIB is endoscopic. It not only assists in precisely identifying the location of bleeding but also in selecting the most appropriate hemostatic treatments, including endoscopic variceal band ligation (EVBL), sclerotherapy, adrenaline injection, and hemoclip placement. High-definition video endoscopes allow proper characterization of lesions such as varices, ulcers, strictures, and masses in the esophagus, stomach, and duodenum.6 Variceal bleeding has been reported as the most frequent endoscopic finding in tertiary centers with a higher prevalence of CLD, whereas the most common nonvariceal finding is PUD.7
Results of UGIB are influenced by a number of interrelated factors, including age, underlying comorbid conditions, extent of bleeding, and efficacy of endoscopic treatment. Mortality rates are also a significant concern, although there has been an increase in endoscopic hemostasis and supportive care, and the rates of mortality vary between approximately 2 and 10% in present-day audits.8 It has been established that early endoscopy is also associated with reduced rebleeding, fewer transfusions, shorter hospital stays, and improved survival compared with late procedures.9 The use of risk stratification instruments like the Glasgow Blatchford Score and the AIMS65 score has also been found to be effective predictors of how such patients may need intervention and how they will respond in clinical care, and can act as a guideline in making the clinician's decision on what they need in terms of management triage.10
Open-ended data from local audits on the epidemiology, clinical presentations, endoscopic findings, and treatment measures and outcomes of UGIB are vital, particularly in areas with high hepatitis viral and CLD prevalence. This kind of data will provide insights into the disease burden, resource consumption, and the actual efficacy of treatment plans within the healthcare system in question.11 Nevertheless, long-term large-scale audits involving both variceal and nonvariceal UGIB in tertiary care remain relatively lacking.
The aim of this audit was to characterize clinical features, endoscopic findings, therapeutic interventions, and outcomes of patients presenting with UGIB at a tertiary care center from 2022 to 2025, with an emphasis on etiological patterns, symptom profiles, endoscopic diagnoses, and outcome measures. This audit aims to inform local clinical practice and quality improvement efforts by identifying trends in presentation and prognosis in a high-burden population.
Materials and Methods
Study Design and Setting
This clinical audit was conducted at the Gastroenterology ward of Jinnah Post Graduate Medical Center over a 4-year period from January 2022 to December 2025, and the data were collected retrospectively from the patients' records. The study included patients presenting with symptoms of UGIB who underwent esophagogastroduodenoscopy (EGD) during the study period. A non-probability convenience sampling technique was employed to include all eligible cases. A total of 3,448 patients were enrolled, comprising the entire cohort that met the study criteria within the defined timeframe.
The study was conducted after obtaining ethical approval in accordance with the principles outlined in the Declaration of Helsinki and applicable institutional ethical guidelines.
Patient Selection
Patients of all ages presenting with clinical features suggestive of UGIB, including hematemesis, melena, or hematochezia, were included in the study. Patients were excluded if they had a history of major gastrointestinal surgical interventions, incomplete or missing medical records, documented coagulopathies unrelated to liver disease, or if they died before complete clinical data could be obtained.
Laboratory Investigations and Etiological Classification
All patients underwent relevant laboratory investigations, including complete blood count, liver function tests, hepatitis B surface antigen (HBsAg), and anti-HCV antibodies.
Based on the underlying etiology, patients were categorized into decompensated chronic liver disease (DCLD) secondary to HCV, HBV, HBV/HCV coinfection, non-B non-C (NBNC), metabolic dysfunction-associated fatty liver disease (MAFLD), alcoholic liver disease (ALD), Autoimmune hepatitis (AIH), PUD, corrosive injury, and gastrointestinal malignancy.
Endoscopic Evaluation and Therapeutic Interventions
Endoscopic evaluation was performed using a high-definition video endoscope (Olympus GIF-Q180 or equivalent) by experienced endoscopists. Endoscopic findings involving the esophagus, stomach, and duodenum, including varices, ulcers, strictures, and masses, were systematically documented.
Therapeutic endoscopic interventions were recorded, including EVBL, Histoacryl® injection sclerotherapy, adrenaline injection therapy, and hemoclip placement, depending on the underlying pathology and bleeding source.
Outcome Measures
Patient outcomes were recorded and analyzed to assess the clinical course and effectiveness of interventions. These outcomes included discharge after clinical stabilization, in-hospital mortality, leaving against medical advice (LAMA), referral to another facility, and discharge on request (DOR).
Audit Standards and Benchmarks
In addition to descriptive analysis of clinical and endoscopic findings, selected quality indicators for the management of UGIB were evaluated as part of an institutional audit. These benchmarks were defined according to recommendations from the Baveno VII Consensus Workshop on Portal Hypertension and other established international guidelines for the management of UGIB.
Key audit parameters included time to endoscopy, defined as the interval between hospital presentation and performance of EGD. For patients with suspected UGIB, endoscopy within ≤24 hours of presentation was considered the recommended benchmark.
For patients with suspected variceal bleeding, the administration of vasoactive pharmacological therapy (Terlipressin or Octreotide) prior to or during endoscopic intervention was evaluated as an essential management indicator.
Similarly, for cases of nonvariceal bleeding, particularly those related to PUD, the use of proton pump inhibitor (PPI) therapy before and after endoscopic evaluation was assessed to determine adherence to recommended treatment protocols.
Statistical Analysis
Data analysis was performed using IBM SPSS Statistics 26. Quantitative variables, such as age and follow-up duration, were expressed as mean ± standard deviation (SD). Categorical variables, including symptoms, etiologies, endoscopic findings, interventions, and outcomes, were presented as frequencies and percentages.
Comparisons of categorical variables across study years and patient groups were performed using the Chi-square test, while continuous variables were compared using one-way analysis of variance (ANOVA). A p-value < 0.05 was considered statistically significant, indicating meaningful differences between groups or trends during the study period.
Results
The baseline characteristics and etiological distribution of the 3,448 patients included in the study are mentioned in Table 1 and Figure 1. The mean age was 52.11 ± 14.98 years, with a mean follow-up difference of 3.44 ± 2.43 days. The largest proportion of patients was in the older age group (>50 years), 1,902 (55.2%). A male predominance was observed with 2,035 (59.0%) males and 1,413 (41.0%) females. Regarding etiology, DCLD secondary to HCV was the most common cause.
Table 1.
Demographic characteristics and etiological distribution of the study population
| Variable (n = 3,448) | Mean ± SD n (%) |
|---|---|
| Age (years) | 52.11 ± 14.98 |
| Follow-up difference (Days) | 3.44 ± 2.43 |
| Age group | |
| Young (>30) | 405 (11.7%) |
| Middle (31–50) | 1,141 (33.1%) |
| Older (>50) | 1,902 (55.2%) |
| Gender | |
| Male | 2,035 (59.0%) |
| Female | 1,413 (41.0%) |
| Etiology | |
| DCLD-HCV | 1,898 (55.0%) |
| DCLD-HBV | 174 (5.0%) |
| DCLD HBV + HCV | 102 (3.0%) |
| DCLD (Non B/C) | 93 (2.7%) |
| DCLD secondary to HBV + HDV | 36 (1.0%) |
| DCLD secondary to ALD | 31 (0.9%) |
| DCLD secondary to AIH | 9 (0.3%) |
| DCLD/HCV/HBV/HDV | 12 (0.3%) |
| DCLD/MAFLD | 10 (0.3%) |
| Drugs | 17 (0.5%) |
| Peptic ulcer disease | 622 (18.0%) |
| Portal vein thrombus | 2 (0.1%) |
| BCS | 15 (0.4%) |
| NCPH | 12 (0.3%) |
| Corrosive injury | 33 (1.0%) |
| Cancer | 73 (2.1%) |
| Systemic vasculitis | 6 (0.2%) |
| Chronic pancreatitis | 4 (0.1%) |
| Primary sclerosing cholangitis (PSC) | 8 (0.2%) |
| Splenic vein thrombus -SVT | 5 (0.1%) |
| Mallory-Weiss tear | 15 (0.4%) |
| Other | 184 (5.3%) |
| Esophagitis | 32 (0.9%) |
| Erosive gastritis | 55 (1.6%) |
DCLD-HBV, decompensated chronic liver disease-hepatitis B virus; DCLD-HCV, decompensated chronic liver disease-hepatitis C virus; DCLD HBV + HCV, decompensated chronic liver disease hepatitis B virus + hepatitis C virus
Fig. 1.
Temporal trends in the clinical features of the patients (2022–2025)
Table 2 and Figure 2 present a year-by-year comparison of demographic characteristics and etiological patterns from 2022 to 2025. The annual distribution included 862 patients each in 2022 and 2023, 781 in 2024, and 950 in 2025. The mean age showed a statistically significant variation across years, with values of 51.58 ± 15.62 in 2022, 54.52 ± 14.54 in 2023, 52.35 ± 14.21 in 2024, and 50.19 ± 15.12 in 2025 (p < 0.001). Similarly, the mean follow-up difference declined significantly from 3.70 ± 2.80 days in 2022 to 3.66 ± 2.91 days in 2023, 3.44 ± 2.28 days in 2024, and 3.0 ± 1.43 days in 2025 (p < 0.001). The older age group (>50 years) predominated each year with 454 (52.7%) in 2022, 519 (60.2%) in 2023, 434 (55.6%) in 2024, and 495 (52.5%) in 2025, whereas the young group (<30 years) accounted for 108 (12.5%), 57 (6.6%), 84 (10.8%), and 156 (16.5%), respectively (p < 0.001). Across all years, DCLD due to HCV remained the leading etiology. PUD was consistently the second most common cause. DCLD/HBV accounted for 54 (6.3%), 55 (6.4%), 53 (6.8%), and 12 (1.3%) cases across the four years, while DCLD/HBV/HCV showed a marked rise in 2025 with 58 (6.2%) compared to 15 (1.7%), 15 (1.7%), and 14 (1.8%) in preceding years. Less frequent etiologies demonstrated minor year-to-year variation, though notable increases in 2025 were observed for esophagitis 32 (3.4%), persistent vomiting 8 (0.8%), DCLD/MAFLD 8 (0.8%), PSC 5 (0.5%), and splenic vein thrombus 3 (0.3%).
Table 2.
Year-wise comparison of demographic and etiological distribution of patients
| Variable | Year 2022 Mean ± SD n = 862 | Year 2023 Mean ± SD n = 862 | Year 2024 Mean ± SD n = 781 | Year 2025 Mean ± SD n = 943 | p-value |
|---|---|---|---|---|---|
| Age (years) | 51.58 ± 15.62 | 54.52 ± 14.54 | 52.35 ± 14.21 | 50.19 ± 15.12 | <0.001 |
| Follow-up difference (days) | 3.70 ± 2.80 | 3.66 ± 2.91 | 3.44 ± 2.28 | 3.0 ± 1.43 | <0.001 |
| Age group | |||||
| Young (<30) | 108 (12.5%) | 57 (6.6%) | 84 (10.8%) | 156 (16.5%) | <0.001 |
| Middle (31–50) | 300 (34.8%) | 286 (33.2%) | 263 (33.7%) | 292 (31.0%) | |
| Older (>50) | 454 (52.7%) | 519 (60.2%) | 434 (55.6%) | 495 (52.5%) | |
| Gender | |||||
| Male | 549 (63.7%) | 460 (53.4%) | 474 (60.7%) | 552 (58.5%) | <0.001 |
| Female | 313 (36.3%) | 402 (46.6%) | 307 (39.3%) | 391 (41.5%) | |
| Etiology | |||||
| DCLD/HCV | 466 (54.1%) | 463 (53.7%) | 418 (53.5%) | 551 (58.4%) | <0.001 |
| DCLD/HBV | 54 (6.3%) | 55 (6.4%) | 53 (6.8%) | 12 (1.3%) | |
| DCLD/HBV/HCV | 15 (1.7%) | 15 (1.7%) | 14 (1.8%) | 58 (6.2%) | |
| DCLD/NBNC | 30 (3.5%) | 32 (3.7%) | 21 (2.7%) | 10 (1.1%) | |
| Drugs | 5 (0.6%) | 5 (0.6%) | 5 (0.6%) | 2 (0.2%) | |
| Peptic ulcer disease | 165 (19.1%) | 165 (19.1%) | 152 (19.5%) | 140 (14.8%) | |
| DCLD/HBV/HDV | 13 (1.5%) | 13 (1.5%) | 10 (1.3%) | 0 (0.0%) | |
| Portal vein thrombus | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 2 (0.2%) | |
| BCS | 5 (0.6%) | 5 (0.6%) | 5 (0.6%) | 0 (0.0%) | |
| DCLD/ALD | 6 (0.7%) | 6 (0.7%) | 6 (0.8%) | 13 (1.4%) | |
| DCLD/AIH | 3 (0.3%) | 4 (0.5%) | 0 (0.0%) | 2 (0.2%) | |
| NCPH | 3 (0.3%) | 3 (0.3%) | 3 (0.4%) | 3 (0.3%) | |
| Corrosive injury | 11 (1.3%) | 11 (1.3%) | 9 (1.2%) | 2 (0.2%) | |
| Cancer | 16 (1.9%) | 16 (1.9%) | 20 (2.6%) | 21 (2.2%) | |
| Systemic vasculitis | 1 (0.1%) | 1 (0.1%) | 1 (0.1%) | 3 (0.3%) | |
| Chronic pancreatitis | 1 (0.1%) | 1 (0.1%) | 0 (0.0%) | 2 (0.2%) | |
| PSC | 1 (0.1%) | 1 (0.1%) | 1 (0.1%) | 5 (0.5%) | |
| Splenic vein thrombus | 1 (0.1%) | 1 (0.1%) | 0 (0.0%) | 3 (0.3%) | |
| Persistent vomiting | 2 (0.2%) | 2 (0.2%) | 3 (0.4%) | 8 (0.8%) | |
| Other | 47 (5.5%) | 47 (5.5%) | 43 (5.5%) | 47 (5.0%) | |
| DCLD/HCV/HBV/HDV | 2 (0.2%) | 2 (0.2%) | 2 (0.3%) | 6 (0.6%) | |
| DCLD/MAFLD | 1 (0.1%) | 0 (0.0%) | 1 (0.1%) | 8 (0.8%) | |
| Esophagitis | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 32 (3.4%) | |
| Erosive gastritis | 14 (1.6%) | 14 (1.6%) | 14 (1.8%) | 13 (1.4%) | |
Fig. 2.
Temporal trends in outcomes (2022–2025)
Table 3 summarizes the clinical presentation, endoscopic findings, therapeutic interventions, annual distribution, and outcomes of the 3,448 patients. Hematemesis was the predominant presenting symptom in 2,855 (82.8%) patients, followed by melena in 408 (11.8%). On endoscopic evaluation, variceal bleeding was common, with esophageal varices observed in 1,374 (39.8%), esophagogastric varices in 460 (13.3%), and gastric varices in 378 (11.0%). Among nonvariceal findings, gastric ulcer was the most frequent, 724 (21.0%). Regarding endoscopic interventions, EVBL was performed in 1,625 (47.1%) patients, Histoacryl® sclerotherapy in 378 (11.0%), combined EVBL with Histoacryl in 279 (8.1%), adrenaline sclerotherapy in 48 (1.4%), and adrenaline with hemoclip in 23 (0.7%), while 1,095 (31.8%) required no intervention. In terms of outcomes, the majority were discharged 2,950 (85.5%).
Table 3.
Clinical presentation, endoscopic findings, therapeutic interventions, and outcomes of the study population (n = 3,448)
| Variable (n = 3,448) | Mean ± SD n (%) |
|---|---|
| Symptoms | |
| Hematemesis | 2,855 (82.8%) |
| Melena | 408 (11.8%) |
| Hematochezia | 47 (1.4%) |
| Hematemesis and melena | 138 (4.0%) |
| Findings | |
| Esophagus | |
| Esophageal ulcer | 85 (2.5%) |
| Post-Banding ulcer | 18 (0.5%) |
| Esophageal mass | 29 (0.8%) |
| Gastric | |
| Gastric ulcer | 724 (21.0%) |
| Gastric mass | 33 (1.0%) |
| Portal hypertensive gastropathy | 55 (1.6%) |
| Duodenum | |
| Duodenum ulcer | 242 (7.0%) |
| Duodenal stricture | 38 (1.1%) |
| Duodenal mass | 12 (0.3%) |
| Variceal bleeding | |
| Esophageal varices | 1,374 (39.8%) |
| Gastric varices | 378 (11.0%) |
| Esophagogastric varices | 460 (13.3%) |
| EGD intervention | |
| EVBL | 1,625 (47.1%) |
| Histoacryl sclerotherapy | 378 (11.0%) |
| EVBL and histoacryl sclerotherapy | 279 (8.1%) |
| Adrenaline sclerotherapy | 48 (1.4%) |
| Adrenaline and hemoclip | 23 (0.7%) |
| None | 1,095 (31.8%) |
| Outcome | |
| Discharge | 2,950 (85.5%) |
| Expired | 341 (9.9%) |
| LAMA | 133 (3.9%) |
| Referred | 19 (0.6%) |
| DOR | 5 (0.1%) |
DOR, discharges on request; LAMA, left against medical advice
Table 4 and Figure 2 present the year-wise trends in symptoms, endoscopic findings, interventions, and outcomes from 2022 to 2025. Hematemesis remained the most common presenting symptom, reported in 711 (82.5%) patients in 2022, 711 (82.5%) in 2023, 634 (81.2%) in 2024, and 799 (84.7%) in 2025. Significant variations were noted in endoscopic findings, particularly esophageal ulcer, which increased from 17 (2.0%) in 2022 and 2023 and 14 (1.8%) in 2024 to 37 (3.9%) in 2025 (p < 0.001). Variceal bleeding showed a marked rise in esophageal varices from 307 (35.6%) in 2022 and 2023 and 270 (34.6%) in 2024 to 490 (52.0%) in 2025, whereas gastric varices decreased from 102 (11.8%), 102 (11.8%), and 97 (12.4%) to 77 (8.2%), and esophagogastric varices declined from 137 (15.9%), 137 (15.9%), and 116 (14.9%) to 70 (7.4%). Among interventions, EVBL remained the most frequently performed intervention. Outcome trends showed that the majority of patients were discharged successfully each year. Mortality was recorded in 126 (14.6%) patients in 2022, 126 (14.6%) in 2023, 89 (11.4%) in 2024, and 135 (14.3%) in 2025. Patients leaving against medical advice were 47 (5.5%) in 2022, 47 (5.5%) in 2023, 39 (5.0%) in 2024, and 20 (2.1%) in 2025. Referrals to other centers were observed in 5 (0.6%) patients in 2022, 5 (0.6%) in 2023, 9 (1.2%) in 2024, and 7 (0.7%) in 2025. Overall, these findings indicate relatively stable outcomes over the years, with a slight increase in discharge rates and a reduction in the number of patients leaving against medical advice.
Table 4.
Year-wise trends in symptoms, endoscopic findings, interventions, and outcomes of patients (2022–2025)
| Variable | Year 2022 Mean ± SD n = 862 | Year 2023 Mean ± SD n = 862 | Year 2024 Mean ± SD n = 781 | Year 2025 Mean ± SD n = 943 | p-value |
|---|---|---|---|---|---|
| Symptoms | |||||
| Hematemesis | 711 (82.5%) | 711 (82.5%) | 634 (81.2%) | 799 (84.7%) | 0.652 |
| Melena | 102 (11.8%) | 102 (11.8%) | 98 (12.5%) | 106 (11.2%) | |
| Hematochezia | 12 (1.4%) | 12 (1.4%) | 11 (1.4%) | 12 (1.3%) | |
| Hematemesis and melena | 37 (4.3%) | 37 (4.3%) | 38 (4.9%) | 26 (2.8%) | |
| Findings | |||||
| Esophagus | |||||
| Esophageal ulcer | 17 (2.0%) | 17 (2.0%) | 14 (1.8%) | 37 (3.9%) | <0.001 |
| Post-banding ulcer | 4 (0.5%) | 4 (0.5%) | 3 (0.4%) | 7 (0.7%) | |
| Esophageal mass | 6 (0.7%) | 6 (0.7%) | 8 (1.0%) | 9 (1.0%) | |
| Gastric | |||||
| Gastric ulcer | 190 (22.0%) | 190 (22.0%) | 176 (22.5%) | 168 (17.8%) | |
| Gastric mass | 8 (0.9%) | 8 (0.9%) | 9 (1.2%) | 8 (0.8%) | |
| Portal hypertensive gastropathy | 15 (1.7%) | 15 (1.7%) | 15 (1.9%) | 10 (1.1%) | |
| Duodenum | |||||
| Duodenal ulcer | 62 (7.2%) | 62 (7.2%) | 58 (7.4%) | 60 (6.4%) | |
| Duodenal stricture | 12 (1.4%) | 12 (1.4%) | 12 (1.5%) | 2 (0.2%) | |
| Duodenal mass | 2 (0.2%) | 2 (0.2%) | 3 (0.4%) | 5 (0.5%) | |
| Variceal bleeding | |||||
| Esophageal varices | 307 (35.6%) | 307 (35.6%) | 270 (34.6%) | 490 (52.0%) | |
| Gastric varices | 102 (11.8%) | 102 (11.8%) | 97 (12.4%) | 77 (8.2%) | |
| Esophagogastric varices | 137 (15.9%) | 137 (15.9%) | 116 (14.9%) | 70 (7.4%) | |
| EGD intervention | |||||
| EVBL | 398 (46.2%) | 398 (46.2%) | 342 (43.8%) | 487 (51.6%) | 0.001 |
| Histoacryl sclerotherapy | 102 (11.8%) | 102 (11.8%) | 97 (12.4%) | 77 (8.2%) | |
| EVBL and histoacryl sclerotherapy | 71 (8.2%) | 71 (8.2%) | 67 (8.6%) | 70 (7.4%) | |
| Adrenaline sclerotherapy | 9 (1.0%) | 9 (1.0%) | 7 (0.9%) | 23 (2.4%) | |
| Adrenaline and hemoclip | 8 (0.9%) | 8 (0.9%) | 7 (0.9%) | 0 (0.0%) | |
| None | 274 (31.8%) | 274 (31.8%) | 261 (33.4%) | 286 (30.3%) | |
| Outcome | |||||
| Discharge | 684 (79.3%) | 684 (79.3%) | 639 (81.8%) | 778 (82.5%) | <0.001 |
| Expired | 126 (14.6%) | 126 (14.6%) | 89 (11.4%) | 135 (14.3%) | |
| LAMA | 47 (5.5%) | 47 (5.5%) | 39 (5.0%) | 20 (2.1%) | |
| Referred | 5 (0.6%) | 5 (0.6%) | 9 (1.2%) | 7 (0.7%) | |
| DOR | 0 (0.0%) | 0 (0.0%) | 5 (0.6%) | 3 (0.3%) | |
Discussion
This 4-year retrospective audit of 3,448 patients with UGIB at a tertiary care center demonstrated that variceal bleeding, primarily related to CLD, was the predominant etiology, with PUD as the leading nonvariceal cause. Hematemesis and melena were the most frequent clinical manifestations. Endoscopic assessment enabled both diagnosis and therapeutic intervention, with variceal band ligation most commonly performed. The findings emphasize the importance of timely endoscopy and systematic audits in guiding evidence-based management and improving patient outcomes.
In a recent investigation of 383 patients with UGIB, esophageal varices predominated (78.3%) and were strongly associated with HCV (70.7%, p < 0.001), while HBV showed minimal involvement, and acid peptic disease was less frequent (11%).3 In comparison, our audit of 3,448 patients similarly demonstrated variceal bleeding, predominantly linked to HCV DCLD (55.0%), as the leading cause of UGIB. Nonvariceal causes, including PUD, were less frequent (18.0%), and HBV-related cases were slightly higher, reflecting regional variations.
A recent retrospective study of 142 Pakistani patients with PUD (PUD)-related UGIBU reported a mean age of 53 years, 68.3% male predominance, 1-week and 1-month mortality rates of 10.6% and 14.8%, and early re-bleeding rates of 14–18%, with NSAID and antiplatelet use associated with worse outcomes.12 In comparison, our audit found PUD in 18% of 3,448 patients, with a similar mean age (52.1 years) and male predominance (59%), highlighting a comparable demographic pattern and the clinical significance of PUD in our population.
In an international multicenter cohort of 1,535 cirrhotic patients with acute gastrointestinal bleeding, peptic ulcer bleeding accounted for 4.7% of cases and was not independently associated with increased in-hospital mortality or 5-day failure to control bleeding compared with variceal bleeding.13 Correspondingly, in our audit, PUD constituted 18% of cases within a predominantly CLD population, highlighting that nonvariceal etiologies such as PUD represent a substantial clinical burden and may have outcomes comparable to variceal bleeding in cirrhotic patients.
A retrospective study from Ayub Teaching Hospital, including 105 patients with CLD with UGIB, showed a mean age of 55.3 years, with a preponderance of males (61.9%), leading etiology of 42.9% (viral hepatitis), and 33.3% with PUD.7 In comparison, our audit demonstrated a similar demographic pattern (mean age 52.1 years; 59% male) with viral hepatitis constituting the predominant cause of DCLD, and PUD accounting for 18% of cases overall. These findings collectively highlight the consistent burden of PUD within a predominantly viral hepatitis–driven CLD population in our regional setting.
A recent meta-analysis of 18 studies (n = 19,040) reported a pooled prevalence of DCLD of 15.6% in cirrhotic patients in Pakistan, with HCV (63.5%) and HBV (18.7%) as the leading causes and an emerging contribution from MASLD (12.1%), alongside high mortality driven by variceal bleeding and related complications.14 In comparison, our audit similarly demonstrated a persistent predominance of HCV-related DCLD (53–58% annually), followed by HBV and mixed viral etiologies, with a gradual rise in MAFLD cases in later years, reinforcing a consistently viral hepatitis–driven burden with an evolving metabolic component in our audit.
A recent global analysis reported HCV-related cirrhosis as the leading cause of CLD (63.5%), with NAFLD contributing 12.1% and HBV 18.7%, while showing regional and sociodemographic variations in trends and increasing metabolic etiologies over time.15 In comparison, our audit showed a similar trend with HCV-associated DCLD in 53-58% each year, MAFLD 0.1–0.8% by 2025, and HBV and mixed viral etiologies between 1.3 and 6.8%, supporting both the ongoing viral hepatitis burden contribution and the new metabolic burden contribution in our regional group.
A study from King Fahad Central Hospital reported hematemesis (67.5%) as the predominant UGIB presentation, with esophageal varices (52.2%) and duodenal ulcers (21.7%) as main endoscopic findings, EVBL and Adrenaline/Endoclip therapy performed in 48–37% of cases, and rebleeding and mortality rates of 7 and 6%, respectively.1 Similarly, our audit found hematemesis in 82.8%, esophageal varices in 39.8%, gastric and duodenal ulcers in 21 and 7%, EVBL/sclerotherapy in 47.8%, with 14.3% rebleeding and 9.9% in-hospital mortality, reflecting comparable clinical presentations, interventions, and outcomes.
The Lady Reading Hospital study reported esophageal varices (87.5%), portal gastropathy (37.5%), and gastric varices (27%) as the main endoscopic findings in 152 cirrhotic patients with UGIB, with peptic ulcers (13.8%) and gastritis (12.5%) as nonvariceal lesions.16 Similarly, our audit found esophageal varices (39.8%), gastric ulcers (21%), duodenal ulcers (7%), and portal gastropathy (1.6%), confirming that variceal lesions predominate while nonvariceal sources remain significant in cirrhotic UGIB.
This Karachi study of 112 high AIMS 65 score UGIB patients indicated duodenal ulcers (46.4%) to be the most frequent endoscopic observation, followed by gastric ulcer (42%), and Forrest class IIC was highest; other findings included esophagitis (10.7%), Mallory-Weiss tears (13.4%), and Dieulafoy lesions (5.4%).17 In the same way, gastric ulcers (21%) and duodenal ulcers (7%) were the most common nonvariceal lesions in our audit group of 3,448 patients, on top of nonvariceal bleeding and portal hypertensive gastropathy. Both articles highlight the significance of nonvariceal ulcers in a considerable number of UGIB cases, underscoring the relevance of endoscopic analysis for diagnosis, Forrest classification, and treatment.
In the mentioned study, the 13-year-old cohort, nonvariceal UGIB (NVUGIB) accounted for 80.5% of cases, followed by gastroduodenal erosions (23.8%), as well as duodenal (23.5%) and gastric ulcers (10.8%) as the main lesions.18 Likewise, through our audit of 3,448 patients (2022–2025), we found that hematemesis was the most common (82–85%); other top endoscopic findings included esophageal varices (3,552), gastric ulcers (1,722), and duodenal ulcers (67).
According to the above-listed 13-year study, NVUGIB led in 80.5 cases; gastroduodenal erosions were the most frequent, at 23.8 and 23.5, respectively; and variceal bleeding constituted 13.1 cases per 100 cases.18 Equally, in our audit of 3,448 patients (2022–2025), the most common symptom is hematemesis (82–85%), and the major endoscopy findings were variceal (35–52%) and nonvariceal (6–7%) sources of UGIB.
A recent cross-sectional survey in rural and peri-urban Sindh reported a high HCV burden, with 17.1% of participants testing anti-HCV positive and 7.7% viremic, disproportionately affecting women ≥30 years, rural residents, and individuals with limited education. These findings underscore the persistent challenge of HCV in underserved populations with limited healthcare access, delayed diagnosis, and low awareness.19 In comparison, our study similarly observed that variceal bleeding secondary to hepatitis C–related CLD was a predominant cause of UGIB, highlighting the clinical consequences of the high HCV burden and the importance of early endoscopic intervention in resource-limited settings.
This study has several limitations. Its retrospective design and single-center setting restrict the generalizability of the findings to other populations and healthcare environments. The study also did not include long-term follow-up, preventing assessment of delayed complications or recurrent bleeding. Despite these limitations, the study provides valuable insights into the demographic and clinical characteristics of patients with UGIB, highlighting the prominence of variceal bleeding in HCV-related liver disease and the importance of early endoscopic intervention in improving patient outcomes.
Limitations of the Study
Single-center study.
Conclusion
In this audit, UGIB was largely due to variceal bleeding, with PUD being the major nonvariceal cause. Hematemesis and melena were the most common clinical manifestations, and endoscopic intervention (especially variceal band ligation) was associated with better clinical outcomes. These findings reveal the importance of prompt endoscopic assessment and demonstrate the significance of systematic audits in terms of informing evidence-based management and optimizing patient care.
Clinical Significance
A local audit creates Pakistan-specific data on etiology patterns (variceal vs PUD vs malignancy) and rebleeding and mortality rates with current practice. Systematically reviewing every case of UGIB in your hospital against standards, tracking outcomes, and feeding that data back to improve care. In resource-limited settings like Pakistan, it matters more, not less, than in high-income countries.
In resource-limited countries, UGIB audits don't just measure quality; they create quality by identifying cheap, fixable gaps and providing evidence to fix them.
International guidelines assume 24/7 availability of endoscopy, interventional radiology, and the ICU. That's not reality here.
Disclosure Statements
None.
Ethical Approval
Study approved by the Institutional Research Board (NO.F.2-81/2024-GENL/33/JPMC).
Data Availability Declaration
Data can be provided upon request.
Artificial Intelligence (AI) Disclosure
No AI-assisted technologies were used.
Acknowledgments
The authors would like to acknowledge our endoscopy nurses and staff for their relentless dedication and vital contributions.
Authors' Contributions
Nazish Butt, corresponding author, supervisor, analysis, review. Omer Idris, Ghulam Mohiuddin, Ali Akbar, Bushra Shahid, and Sabir Ali: Endoscopies, data collection. Nimra Mehak and Hamza Reaz: Data entry and analysis.
Orcid
Omar Idris https://orcid.org/0009-0007-8345-5420
Ghulam Mohiuddin https://orcid.org/0009-0004-8509-329X
Bushra Shahid https://orcid.org/0009-0001-3031-2051
Hamza R Aziz https://orcid.org/0009-0000-0056-8032
Footnotes
Source of support: Nil
Conflict of interest: None
References
- 1.Someili AM, Mobarki SJ, Moafa RH, et al. Upper Gastrointestinal bleeding: A retrospective, single-center experience on the role of endoscopy and outcomes. J Clin Med Res. 2025;17(1):22. doi: 10.14740/jocmr6134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Shung DL, Laine L. Upper gastrointestinal bleeding–review of current evidence and implications for management. Aliment Pharmacol Ther. 2024;59(9):1062–1081. doi: 10.1111/apt.17949. [DOI] [PubMed] [Google Scholar]
- 3.Ali M, Naveed H, Ahmad MH, et al. Endoscopic evaluation of upper gastrointestinal bleeding in high-risk populations for hepatitis B and C: A cross-sectional study. Ann Med Surg. 2025;87(6):3157–3161. doi: 10.1097/MS9.0000000000003305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mohammed MF, Gurunathan RK, Eltanikhy NN, et al. Clinical profile, endoscopic findings, and outcomes among inpatients with upper gastrointestinal bleeding in a United Kingdom District General Hospital. Cureus. 2025;17(11):e97644. doi: 10.7759/cureus.97644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Elbadry M, El-Raey F, Alboraie M, et al. Clinical and endoscopic characteristics of patients undergoing gastrointestinal endoscopic procedures in Egypt: A nationwide multicenter study. BMC Gastroenterol. 2024;24(1):186. doi: 10.1186/s12876-024-03262-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Suliman O, Alotaibi F, Alhoivi Y, et al. Evaluation and management of upper gastrointestinal bleeding in adults: A systematic review. J Adv Med Med Res. 2025;37(12):444–458. doi: 10.9734/jammr/2025/v37i126030. [DOI] [Google Scholar]
- 7.Akbar H, Iqbal N, Khan J, et al. Frequency of PUD in patients with chronic liver disease presenting with upper gastrointestinal bleeding. J Health Rehabil Res. 2024;4(3):1–7. doi: 10.61919/jhrr.v4i3.1347. [DOI] [Google Scholar]
- 8.Nagesh VK, Varughese VJ, Musalli J, et al. Readmission events following EGD for upper gastrointestinal bleed: An analysis using the National Readmission Database. Med Sci. 2025;13(2):45. doi: 10.3390/medsci13020045. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ibrahim SA, Sidahmed HA, Ahmed IE, et al. Minutes to save a life in acute GI bleeding: A systematic review of early vs. delayed endoscopy. Cureus. 2025;17(12):e100176. doi: 10.7759/cureus.100176. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Gouda TE, Fikry AA, Mousa SA, et al. Predictors of mortality in patient with upper gastro-intestinal tract bleeding. Egyptian J Hospital Med (April 2025) 2025;99:2153–2157. [Google Scholar]
- 11.Zada I, Dar SI, Shah FU, et al. Frequency of different etiologies in patients of upper GI bleed. J Islamabad Med Dent College. 2024;13(Suppl):568–573. doi: 10.35787/jimdc.v13i(Suppl.).1084. [DOI] [Google Scholar]
- 12.Butt N, Usmani MT, Mehak N, et al. Risk factors and outcomes of peptic ulcer bleed in a Pakistani population: A single-center observational study. World J Gastrointest Pharmacol Ther. 2024;15(3):92305. doi: 10.4292/wjgpt.v15.i3.92305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Yin Y, Ji F, Romeiro FG, et al. Impact of peptic ulcer bleeding on the in-hospital outcomes of cirrhotic patients with acute gastrointestinal bleeding: An international multicenter study. Expert Rev Gastroenterol Hepatol. 2024;18(8):473–483. doi: 10.1080/17474124.2024.2387823. [DOI] [PubMed] [Google Scholar]
- 14.Rind S, Ali M, Jamali GM, et al. Burden of decompensated liver disease in Pakistan: An epidemiological evidence. J Peoples Univ Med Health Sci Nawabshah. 2025;15(01):229–239. doi: 10.46536/jpumhs/2025/15.01.616. [DOI] [Google Scholar]
- 15.Zhang Y, Luo M, Ming Y. Global burden of cirrhosis and other chronic liver diseases caused by specific etiologies from 1990 to 2021. BMC Gastroenterol. 2025;25(1):641. doi: 10.1186/s12876-025-04264-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Afridi MA, Rashid US, Hassan MK, et al. Acute upper gastrointestinal bleeding in patients with cirrhosis: Endoscopic findings. J Postgrad Med Inst. 2023;37(1):67–72. doi: 10.54079/jpmi.37.1.3130. [DOI] [Google Scholar]
- 17.Kumar D, Kalwar HA, Khalid MT, et al. Endoscopic findings in non-variceal upper gastrointestinal bleed patients with high AIMS 65 score. Pak Armed Forces Med J. 2024;74(4):1102. doi: 10.51253/pafmj.v74i4.8725. [DOI] [Google Scholar]
- 18.Alruzug IM, Aldarsouny TA, Semaan T, et al. Time trends of causes of upper gastrointestinal bleeding and endoscopic findings. Saudi J Gastroenterol. 2021;27(1):28–34. doi: 10.4103/sjg.SJG_378_20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Lotia-Farrukh I, Maniar R, Dharejo ZA, et al. Hepatitis C burden in rural and periurban Sindh, Pakistan: Insights into prevalence and the urgent need for targeted health services. BMJ Public Health. 2026;4(1):e003512. doi: 10.1136/bmjph-2025-003512. [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 can be provided upon request.