Negative Binomial Regression Model of Typhoid Fever Trajectory in Saboba District of Ghana.

Abstract

Typhoid fever has always been a major health problem, and therefore, timely and accurate information about its occurrence rate is imperative. The main objective of the research was to model the prevalence of typhoid fever in Saboba district given the demographic age groupings, gender and time. Data on the incidence of typhoid fever in Saboba district (2009 - 2013) were obtained from the Saboba District Health Information Management database. Statistical Analysis System (SAS) was used to analyze the acquired data. Based on the model diagnostics, the Negative Binomial regression model was emerged as superior to the Poisson regression model. This is because it was able to accommodate the over-dispersion that existed in the dataset. The results from the Negative Binomial regression model indicated that, typhoid fever in Saboba District is independent of gender. With the demographic age grouping of typhoid patients, it was established that, the infectious rate of typhoid fever in Saboba District was comparatively mild or nonthreatening in young children and the aged. Young adults or the economically active age groups suffer more of typhoid infection in Saboba District. The Mann –Kendall test results showed downward trend of the prevalence of typhoid fever among the observed years. More cases of typhoid fever were recorded in the year 2009 followed by a sharp decrease until 2012. Between 2012 and 2013, the incidence of typhoid fever in the district assumed an upward trend.