Clinical Profile and Endoscopic Findings in Patients with Upper Gastrointestinal Bleed Attending a Tertiary Care Hospital: A Descriptive Cross-sectional Study

ABSTRACT Introduction: Upper gastrointestinal bleeding is a common acute medical emergency. Endoscopy is the gold standard diagnostic and therapeutic tool in the management of upper gastrointestinal bleed. This study was undertaken to address the clinical profile, endoscopic profile, and outcomes in patients with upper gastrointestinal bleed. Methods: A descriptive cross-sectional study was conducted in a tertiary care teaching hospital in Gandaki Province, Nepal from January 2018 to December 2019 after obtaining ethical clearance from Institutional Review Committee (MEMG/IRC/291/GA) and informed consent from the patient or patient relatives. The sample size was calculated. Six hundred and sixty patients with upper gastrointestinal bleed were included in the study. Data entry was done in Statistical Packages for the Social Sciences version 20. Results: Peptic ulcers and ruptured oesophageal varices are the common aetiologies of upper gastrointestinal bleed. Inpatient mortality was seen in 98 (14.8 %) patients. Upper gastrointestinal bleed of variceal etiology presents with a higher Rockall score and has more chances of rebleeding and has higher mortality than those with non-variceal aetiologies. Bad prognostic factors were rebleeding, variceal etiology, and comorbidities including cirrhotic and Rockall score > 6. Conclusions: Upper gastrointestinal bleeding is a common acute medical emergency. Early upper gastrointestinal endoscopy preferably within 24 hours is recommended for diagnosis, timely intervention, and management of the patients with an upper gastrointestinal bleed that helps in reducing morbidity and mortality.


INTRODUCTION
Upper gastrointestinal (UGI) bleed is defined as bleeding proximal to the ligament of Treitz. 1 It is one of the common medical emergencies that have an incidence of 50 to 150 cases per 100 000 population and hospital mortality of approximately 7% to 10% even in developed countries. 2,3 Endoscopy is the gold standard diagnostic and therapeutic tool in the management of UGI bleed. used to predict the rebleeding and prognosis in patients with UGI bleed. 5 A descriptive cross-sectional study regarding upper gastrointestinal bleeding is scanty from this region of Nepal.
In the present study, we will try to identify the prevalence, etiology of acute UGI bleed, clinical profiles of the patients including clinical outcomes and mortality in patients admitted in a tertiary care Hospital in Gandaki Province, Nepal.

METHODS
This descriptive cross-sectional study was carried out in the Unit of Medical Gastroenterology , Department of Medicine at a tertiary care Teaching Hospital at Pokhara, Gandaki Province, Nepal of 24 months duration period from January 2018 to December 2019. The sample size was collected using the formula, n= (Z 2 x p x q)/e 2 = (1.96 2 x 0.5 x 0.5)/0.05 2 = 384.16 ~ 385 where, Z= 1.96 (At 95% confidence interval) p= taking 50% prevalence q= 1-p e= margin of error, 5% The calculated minimum sample size was 385. A total of 2880 endoscopies were performed in 24 months which was the sample size more than adequate for the study.
Cases were studied from the records of the endoscopy unit and departmental records of admission and discharge summaries. Patients who presented with a history suggestive of acute upper gastrointestinal bleeding i.e. haematemesis, melena, or syncope were hospitalized randomly irrespective of age, sex, or comorbidities in the emergency department, the hemodynamic assessment was done with careful measurement of pulse and blood pressure including orthostatic changes and urine output. Patients were first hemodynamically stabilized; blood transfusion was given when required. Data considering demographic variables, clinical features, bleeding characteristics were collected and alongside blood investigations like complete blood count, platelets count, blood grouping, liver function test, prothrombin time/international normalized ratio (PT/ INR), coagulation profile and ultrasonography of abdomen reports were recorded.
After hemodynamic stability, usually within 24 hours, each patient had undergone endoscopic investigation by standard flexible gastro duodenal endoscope (PENTAX EPK 700, PENTAX JAPAN Inc) and diagnostic findings were documented. The lesion with recent stigmata of hemorrhage or active bleeding was considered the cause of bleeding when various lesions were found in UGI endoscopy. For control of bleeding, pharmacologic and endoscopic treatments were used. Non-variceal bleeders were treated with continuous intravenous infusion of proton pump inhibitor for at least 72 hours and ulcer healing agents. Variceal bleeders were treated with injection terlipressin/octreotide +/-variceal band ligation. The biopsy was performed in all patients with growth, ulcers, mucosal erosions, and areas suspicious to harbor neoplasia.
Finally, the clinical presentation of patients, their endoscopic findings, complications during hospitalization including rebleeding, days of ICU stay, total hospital stay, outcomes, and mortality were assessed. All patients with acute upper GI bleed aged above 16 irrespective of sex and presentation were included in the study. The following cases were excluded from the study: 1. Patients with incomplete records 2. The same patient readmitted and endoscopy showing the same endoscopic finding in readmissions.
The study was approved and verified by the Institutional Review Committee and the work was following the rules and regulations laid down by the Institutional Review Committee. Informed consent was obtained from patients or patient relatives. Data were collected on a structured proforma covering the relevant subjects of the study and entry was done in Statistical Packages for the Social Sciences version 20. All categorical data were expressed in percent and absolute number. All numerical continuous data were expressed in mean±SD. The data analysis was done using SPSS version 20. All tests were analyzed with a 95% confidence interval.

RESULTS
The total upper GI endoscopies performed was 2880 during the 24 months of the study period from January 2018 till December 2019. Out of these 2880 sample study, 720 patients presented with an upper GI bleed, detecting a prevalence of 25%.
But 24 patients were taken away to home or elsewhere by patient relatives against medical advice despite initial management, UGI endoscopy, and few days of admission, and 16 were excluded because of inadequate data. Twenty patients were readmitted with an upper GI bleed but had the same endoscopic findings as in previous endoscopies and admissions. Finally In the majority comprising of 580 (87.9%) study participants, random blood sugar levels were normal, and blood urea and creatinine were raised in almost half of the patients. Liver function tests revealed hyperbilirubinemia in 300 (45.4 % ) patients and increased ALT, AST, and prolonged prothrombin time by more than 4 sec in 350 (53%), 380 (57.6 %), and 320 (48.5%) patients respectively.
Peptic ulcer disease was the most common cause of upper GI bleed and was detected in 236 (35.7%) patients (Table 3). Peptic ulcer was seen almost equally among both the sexes with a mean age of 36±8.25 years. Duodenal ulcer was more common than gastric ulcer in both the sexes. They were further classified according to Forrest classification.     7 and UK audit 10 respectively. The male predominance of 75% and 78.4% were reported by Dewan et al. 12 and Kashyap et al. 13 in Nepal. The current study also showed the male predominance of 70% which was marginally less compared to the previous studies in Nepal.
Peptic ulcer disease was the most common cause of upper GI bleed and was detected in 236 (35.7%) patients in the current study. Whereas, studies by Anand et al., 6 Jain et al. 14 and Rao et al. 7 reported esophageal varices as the most common cause of UGI bleed and found that the incidence was 45.5%, 47.4%, and 51% respectively. The study by Dewan et al. 12 in Chitwan, Nepal also revealed that esophageal varices (47.5%) were the most common cause of UGI bleed. This was followed by peptic ulcer disease (33.3%), and cardiogenic shock, congestive cardiac failure, spontaneous bacterial peritonitis, and aspiration pneumonia were less common causes of mortality and seen in 9 (9.1%) patients. All of the patients who died had Rockall score >6. More mortalities were observed with rebleeding, variceal bleeders, comorbidities including cirrhotics, GI malignancies, and Rockall score >6 (Table 4).
Endoscopic interventions in the form of injecting adrenaline were done in 6 peptic ulcer bleeders and endoscopic oesophageal variceal band ligation in 133 patients. No surgical intervention of any kind was necessary for this study. The median hospital stay was 6.178+2.11 days. The median ICU stay was 3.21+1.37days. Five hundred and sixty-two (85.2%) patients recovered after needful management and treatment and were discharged from the hospital. Male predominance (M: F=2.3:1) was noted in subjects with variceal bleed. Patients with the variceal bleed group presenting in shock were double than those with a non-variceal bleed group. Mean complete Rockall score after UGI endoscopy was 5.66±1.20 in a variceal group and 4.06±0.98 in the non-variceal group. Rockall score more than or equal to 5 was 3 times more in the variceal group compared to no variceal bleeders. Rebleeding within 7 days was higher and 3 times more with variceal group compared to that with a non-variceal group (30 % vs 10 %). Similarly, mortality was also higher and almost 5 times more with variceal group compared to that with a non-variceal group (31.25% vs 6.4%). (Table 5) The incidence of peptic ulcer bled reported by Anand et al. 6 and Rao et al. 7 were 38.5% and 28% respectively. Peptic ulcer bleeds accounting to 41% was the common cause reported by Kaliamurthy et al. 9 from Jamaica. The incidence of peptic ulcer bleed was 36% in a UK audit 10 and 30.6% by Bhutta et al. 15 Peptic ulcer disease detected in 35.7% was the most common cause of upper GI bleed followed by 33.9%of variceal bleeds in the current study.
Bleeding from erosive mucosal disease was 9.4% in the current study. Anand et al. 6 and Rao et al. 7 found the erosive mucosal disease in 8.5% and 9% respectively. The incidence rates were almost similar in the abovementioned studies. But Bhutta et al. 15 found a higher incidence rate of 18.4%. Bleeding from Mallory Weiss tear was found to be 3.3% in the current study. Kashyap et al. 13 reported Mallory Weiss to tear as a cause of upper GI bleeding in 12 patients (10.8%) and Bhutta et al. 15 reported it in 2 patients (1.4%).
Bleeding from upper GI malignancy was 6.4% in the current study. It was reported to be 5.9 % and 3.3 % in the studies by Jain et al. 14  Among patients with chronic liver disease and variceal bleed, this subset of patients presented with severe anemia, shock, and more complications. Rockall score was higher, rebleeding rates were more (30 % vs 10 %) and higher mortality almost 5-fold (31.25% vs 6.4%) was recorded in variceal bleeders on comparison with non-variceal bleed in the current study. Similar were the findings in the study done by Bhattarai et al. 16 in central Nepal with increased mortality amongst variceal bleeders (27.5 % vs. 7.5%). Similarly, mortality was higher in variceal bleed group (20.2% vs 6.9 %) in the study by Svoboda et al. 17 In the current study, the overall mortality was seen in 98 (14.8 %) patients. Salih et al. 8 investigated 238 patients and found a mortality rate of 3.4%. Kashyap et al. 13 studied 111 patients in India and showed an overall mortality rate of 3.6% and all patients had esophageal variceal bleed. Kaliamurthy et al. 9 reported a mortality rate of 5.7%, whereas UK Audit 2007 10 reported a mortality of 10%. Mortality of 4.2% was reported in the study by Dewan et al. in Chitwan, Nepal. The mortality rate was higher in the current study when compared to the above mentioned studies.
The median hospital stay was 6.178±2.11 days in the current study whereas, it was 7.28±3.18 days in the study by Dewan et al. 12 Rebleeeding, presence of comorbidities and Rockall score >6 were the identifiable risk factors for mortality in the study by Dewan et al. 12 These risk factors were also associated with mortality in our study patients. Advanced age >60 years, variceal bleeders, deranged coagulation profile, prolonged prothrombin time, thrombocytopenia were the other factors associated with mortality in the current study.
This study had its limitations. The study reflects a certain geographical area. Most of the patients had no repeat endoscopies records on follow-ups. Etiology of UGI bleed could not be identified in some subjects because of the unavailability of enteroscopy, capsule endoscopy, and other needful modalities. Early and appropriate fluid and pharmacological management alongside upper GI endoscopy preferably within 24 hours are recommended for diagnosis, timely intervention, and management of the patients with UGI bleed that helps in reducing morbidity and mortality. More expertise with training and availability of diagnostic and therapeutic endoscopies in the majority of hospitals