MODELLING THE BLOOD PRESSURE OF PATIENTS WITH SOME RISK FACTORS USING MULTINOMIAL LOGISTIC REGRESSION

Two hundred and two (202) out - patients from the General Hospital Ikot Abasi were tested of their blood pressure (BP) and the data collected. The data was used to model the blood pressure and the risk factors of patients using the Logistic Regression Model. The risk factors that were considered include age, gender, physical activity, and smoking status, thus, the bioda of the patients collected includes age groups, genders, levels, physical activity and smoking habits. Analysis of the data using Logistic Regression Models were carried out. The results show that the regression model obtain is a good fit model. It was further observed that out of the four risk factors, age and gender have p-values greater than 0.05, hence, they are not statistically significant while physical exercise has p-value less than 0.05, hence, they are statistically significant. Furthermore, the result shows that for age, gender, physical exercise, we expect a 0.057, 1.074, and 1.217 unit increase in the Log odds respectively but for smoking status, we expect 24.481 unit of decrease in the Log odds. The test of association between the blood pressure and all the risk factors shows that the factors affects the blood pressure which implies that all the factors correlate with blood pressure.