An Evaluation of 605 Endoscopic Examination in a Rural Setting, Lacor Hospital in Northern Uganda

Introduction: The aim of this study was to evaluate the profile of esophagogastroduodenal (EGDS) diseases diagnosed by upper endoscopy in a rural area of Uganda in a retro-protective study of 605 patients. Results: The mean age of patients with digestive symptoms was 39.7yrs (SD +/-16.11) and female gender predominated by 60% compared to the male (P value 0.000). Peasant farmers were the commonest group with GI symptoms requiring EGDS compared to the rest 72.1% v 27.9%. Epigastric pain was the commonest indication (58%) for EGDS, followed by chest pain (11%), abdominal pain (8.8%), dyspahgia (7.6%) and hematemesis (7.3%). The commonest endoscopy finding was gastritis (47.9%) followed by esophagitis (14.4%), cancer esophagus (5.1%), esophageal varicose (4%), PUD (2.3%), gastric cancer (1%). However 19.5% of patients had normal EGDS. There was a significant correlation between the outpatient diagnosis and endoscopy finding (P value 0.01, r = 0.144) and between endoscopy finding and histology findings (P value 0.001, r = 0.236). H. pylori was positive in 53% of patients with gastritis. Conclusion: Gastritis is the commonest lesion (47.9%) of which 53% have H pylori and Cancer esophagus account for 5.1% of GI lesion in our setting. Cancer stomach is rare in our setting.


INTRODUCTION
Endoscopy is the ideal procedure for identifying organic diseases of the foregut and its practice is influenced by epidemiology of gastrointestinal (GI) diseases [1]. Results from endoscopic exam allow association of Esophagogastroduodenal (EGDS) findings with patient presentation and extrapolation of the findings to patient population in clinical practice, outside of traditional medical centers [2]. Although 39% of EGDS performed have inappropriate indications, better diagnostic yield depend appropriateness, patient gender and age, treatment setting, and symptoms [3]. Indications for EGDS vary from region to region and many authors have found varying proportions of indications for endoscopy.

Objective
To describe the profile of esophagogastroduodenal (EGDS) diseases diagnosed by upper digestive tract endoscopy (UDTE) in a rural area of Uganda and relate it to the histological findings and H. pylori.

METHODS AND MATERIALS
This retro-prospective study was carried out from Jan 2015 to July 2015 in order to investigate the endoscopic findings in 605 patients who presented to St Mary's Hospital Lacor a large rural based University teaching general hospital with 483 bed located in Northern Uganda. Currently the Hospital has two general surgeons who perform over 1200 upper GI endoscopy per year and each of the surgeons has over 10years experience in EGDS.
All patients with upper gastrointestinal symptoms were first assessed in the out patients unit by a clinician who discretionally decided to book and counsel him/her for EDGS as an elective case in manner standardized manner. On the day of scoping, inform consent was obtained from the patients priory. Those who consented were consecutively recruited and interviewed by the endoscopist to discern the main indication for EGDS and its duration. They were then given lidocaine 1% to gurgle in their throat for 5-10minutes as these causes local numbness in the pharynx and mouth. During the procedure, the endoscopist notes and records the findings and when necessary takes punch biopsy for immediate fixation in standard formalin and send for histology. The pathologist was required to examine and report on the histology finding as well as H. pylori presence after standard staining method with H&E and Giemsa stains. Only patients who could not tolerate the EDGS procedure were excluded. The following variables were obtained and analysed using SPSS version 15 to determine the pattern of esophagogastroduodenal presentation and diseases: Age, Gender, Occupation, indication, Outpatient diagnosis, Endoscopy finding, Histology findings, Presence of malignant cancer, and H Pylori presence.

Biographic Data
Approximate 605 patients were enrolled in the study. From Table 1, the age of patients ranged from 5 to 89 yrs old, mean age 39.7 year (SD+/-16.11).
The majority of patients with gastrointestinal complaints that required EGDS are the youth of 19-35 yrs old (39%, n=605) and Adults of age group 36-65 yrs (47%, n=605). For all the age groups, the female gender predominated in patients undergoing EGDS procedure and overall a significant majority of patient undergoing EGDS in our setting are females 60% compared with their male counterpart (40%), (P value 0.000). However regarding their occupation, 436 (72.1%) were peasant farmers, followed by students 10.7%, teachers (3.6%) and business men/women (3.5%). Therefore, peasant farmers significantly constitute the biggest proportion of patients who have GI symptoms requiring EGDS (P value 0.000, χ 2 = 3295.754).
Regarding indications for EGDS, the commonest indication was epigastric pain (58%), followed by chest pain (11%), abdominal pain (8.8%), dysphagia to solids/liquids (7.6%), and hematemesis (7.3%) [ Table 2]. Dyspepsia which is rather a vague presentation constituted only 0.7%, (n=605) of the indications. Dyspepsia, anemia, lower abdominal pain, hemoptysis and repeat endoscopy were the least common indication for EGDS. Table 3 shows the clinical diagnosis made by the examining clinician based on the medical presentation of the patients in the outpatients (OPD) or referring unit. The most common diagnosis made in medical clinical practice in our setting was gastritis (65%), followed by cancer esophagus (6.3%), Gastroesophageal Reflux Disease (GERD) (5.3%), peptic ulcer disease (PUD) (4.6%), esophagitis (4.6%) and esophageal varices (2.5%). Hence the more diagnosis of gastritis made, the more the outpatient move away from PUD. The least common OPD diagnosis made by clinicians was duodenitis. However in 3.5% of instances the clinician could not make a diagnosis.   Whereas the clinician made diagnosis basing on the clinical presentation and referred the patient for EDGS, out of the 605 patients studied, the endoscopist found gastritis (47.9%) to be the commonest condition affecting gastrointestinal symptoms patients in our community. This was followed by esophagitis (14.4%), cancer esophagus (5.1%), esophageal varicose (4%) and PUD (2.3%), (Table 4). Whereas 19.5% of patients who were scoped were found to be normal, hiatus hernia and cancer stomach accounted for the least lesion found that is, only 1% of the upper Gi diseases in our setting respectively. There was a significant correlation between the outpatient diagnosis and endoscopy finding, (P value 0.01, r = 0.144). Therefore in our clinical setting with paucity of resources, clinician could initiate treated without waiting for endoscopy results.

DISCUSSION
Endoscopy remains a powerful tool in diagnosis and management of GI conditions. In this study of Esophagogastroduodenoscopy (EGDS), the mean age of patients requiring the procedure was 39.7year (SD+/-16.11) which is similar to the mean age of 37.7 years (range: 15-84 years) reported by Djibril et al. 2009 [8]. Like in this study, many other others have found a higher female preponderance over their male counterpart in Nigeria, Togo and Uganda respectively [7,8,9]. Perhaps this relate to the female patients' better health seeking behavior compared to their male counterpart. Peasant farming was a significantly predominant occupation amongst the 605 patients studied (P value 0.000). In our setting peasant farmers are often poor people leaving in areas with poor sanitation and most often lack formal education. Tsongo et al. [13], found that poor sanitation, and lack of formal education constitute significant predisposition factors to H. Pylori infection.
The indications for EGDS vary with patients' description of what they feel. Whereas most studies found dyspepsia to be the leading indication [5,7,8] this study found epigastric pain to be the leading indication for EGDS (58%). Epigastralgia was also reported to be the leading indication of EGDS by Olokoba et al. [7] and Okello [9]. Most patients in our setting are able to finger point to their epigastrium as painful part and hence the reason for them requiring EGDS. With respect to the indication, the examining clinician had to make a diagnosis prior to referring the patient for EGDS. The most common diagnosis made in medical clinical practice in our setting was gastritis (65%), followed by cancer esophagus (6.3%) Gastroesophageal Reflux Disease (GERD) (5.3%), peptic ulcer disease (PUD) (4.6%), esophagitis (4.6%) and esophageal varices (2.5%). There was a positive significant correlation between OPD diagnoses (P value 0.01, r = 0.144) and EGDS findings. Accuracy of diagnosis relates to experience and expertise of the clinician and it helps to reduce patient anguish and save resources because treatment can be initiated based on clinical diagnosis.
Endoscopically, the commonest lesion identified was gastritis (47.9%) followed by esophagitis (14.4%), cancer esophagus (5.1%), esophageal varicose (4%), PUD (2.3%) and 19.5% (n=605) of patients who were scoped were normal. These findings had a positive and significant correlation to the histological findings, (P value 0.001, r = 0.236). Therefore endoscopist working in our setting could diagnose most of the GI disorders based on their examination alone. The rate of gastritis in this study is higher than most other studies, for example, in a student health center study by Schroeder [15] in Scandinavia, reported overall gastritis rate of 22.1%. The risk factors for H. Pylori infection [13] tend to mirror that of gastritis. Whilst Lodenyo, Rana, Mutuma, Kabanga, Kuria et al. [14] found gastritis rate of 25.8% in Kenya, 6% was reported in Ethiopia [6] and 12.6% in Uganda [9]. Whereas Gherasim and Dranga [16] attributes gastritis to chronic alcohol consumption, amount and frequency, our predominantly peasant community often suffer poor sanitation and lack of education on top of chronic alcohol ingestion in "sackets" as crude spirits. These risk factors combined predispose to high rates of H pylori infection [13], cancer esophagus as well as liver diseases. The low rate of PUD found in this study is congruent to fact that the disease has demonstrated a decreasing prevalence and incidence over the years globally as well as in the Philippines [17,18]. More than half (53%, n=88) of the gastric and PUD (50%, n=2) biopsies showed presence of H. Pylori. The association between H. pylori and gastritis has been found by other researchers in Brazil and Germany as well [19,20]. Furthermore, the severer the gastritis the more difficult to eradicate H. Pylori [21], yet their safe eradication is necessary for the prevention of gastric cancer [22]. The high prevalence of Hepatitis B infection in our community [23] makes liver disease, chronic alcoholism, esophageal lesion more precarious, hence the high rate of esophageal varicose However cancer stomach is rare in our setting (1%).

LIMITATION
The major limitation experienced in the study was related to 5 cases whose tissue histology could not be read, this was caused by a wrong formalin concentration supplied, but such cases were excluded from the study. The other limitation was related to difficulty in follow-up of the cases that met the inclusion criteria, but this was circumvented by using a cross-section survey.

CONCLUSION
The female gender, youth and adults who are poor peasants predominate amongst patients with upper digestive tract symptoms requiring EGDS. Whereas 19.5% their EGDS is normal, gastritis represent commonest lesion in them (47.9%) and 53% of the gastritis have H pylori. Cancer esophagus which accounted for 5.1% of GI lesion is also important ailment in our setting but cancer stomach is rare in our environment.

RECOMMENDATION
Clinician in resource limited areas could initiate treatment basing on clinical diagnosis since there is a correlation between clinical diagnosis, endoscopy and histological findings as this will reduce unnecessary referral and suffering of the patient. Treating of gastritis should aim at eradicating H pylori.

ETHICAL APPROVAL
It is not applicable.