Identifying significant diagnosis and endoscopy factors for upper gastrointestinal bleeding

Ho, Jia Yiun (2015) Identifying significant diagnosis and endoscopy factors for upper gastrointestinal bleeding. Universiti Malaysia Sabah. (Unpublished)

[img]
Preview
Text
Identifying significant diagnosis and endoscopy factors.pdf

Download (10MB) | Preview

Abstract

Background: Upper gastrointestinal bleeding (UGIB) is one of the most common medical emergencies associated with significant morbidity and mortality. Bleeding peptic ulcer is the most common disease for UGIB. The incidence of UGIB is twice as high in men as in women Lost-man-hours were increased due to the cost of peptic ulcer disease in Malaysia. Main Method: Multiple binary logistic was used in this study because the dependent variable, peptic ulcer was in binary form. Four phases were involved in model building approach. Independent variables with multicollinearity sources were removed and insignificant variables were also eliminated from the test. The best model was selected through the Modified Eight Selection Criteria (MSSC) and pseudo-R² which had achieved the objectives in this study. Main Findings: Factor analysis was used to determine the independent variables from a raw data set. This research found that the variable for the best model had been reduced to two variables which were Significant to the model. As a result, the best model was the interaction between Blatchford score and Forrest ulcer classification while another was the interaction between Blatchford score and rapid urea test. The best model was only left with two interaction variable after the process of elimination of insignificant variable and removal of multicollinearity source variables. Conclusion and Interpretation: In conclusion, the problem statement had been achieved by finding the best model which was the interaction between Blatchford score, Forrest ulcer classification and rapid urea test. Those significant variables could help in diagnosis and endoscopy for the treatment of upper gastrointestinal bleeding more accurately and precisely.

Item Type: Academic Exercise
Uncontrolled Keywords: Multiple binary logistic regression, multicollinearity, modified eight selection criteria (MSSC), best model, upper gastrointestinal bleeding
Subjects: Q Science > QA Mathematics
R Medicine > RC Internal medicine
Divisions: FACULTY > Faculty of Science and Natural Resources
Depositing User: Munira
Date Deposited: 20 Nov 2017 02:04
Last Modified: 20 Nov 2017 02:04
URI: http://eprints.ums.edu.my/id/eprint/17496

Actions (login required)

View Item View Item

Browse Repository
Collection
   Articles
   Book
   Speeches
   Thesis
   UMS News
Search
Quick Search

   Latest Repository

Link to other Malaysia University Institutional Repository

Malaysia University Institutional Repository