Statistical modelling of determinants of contraceptive use in Kenya.

Abstract

The world population has been increasing as the years progresses. The world population as of 2020 stands at 7.8 billion people and is estimated to be approximately 9.9 billion people by 2050. Kenya’s population in the year 2019 stands at 47.7 million people (KNBS,2019). Previous research has established that population increase harms economic development due to changes in the population composition and size, leading to the rise in dependency level and poverty levels. Eighteen percent of currently married women are experiencing an unmet need for limiting and spacing (KDHS,2014). The government and non-governmental organisations have made a deliberate e_ort to introduce and expose those women with the unmet need for family planning to minimize unwanted pregnancies, leading to mortality or morbidity. This project aims to determine the Socioeconomic and Social demographic factors associated with the use and non-use of contraceptives. The study used secondary data obtained from the Kenya Demographic Health Survey 2014. Based on previous studies, the outcome variable was the use or non-use of contraceptives while the independent variables were; current marital status, highest education level, religion, total children ever born, type of place of residence, frequency of listening to the radio, age in _ve-year interval and wealth index. A multicollinearity test was performed, and it was established that there was no collinearity since the variance in_ation factor was less than _ve for all the variables. The study used binary logistic regression to model the relationship between contraceptive use or non-use and the independent variables. The backward elimination method was adapted to terminate the non-signi_cant variables. Type of place of residence and age in _ve-year group intervals were eliminated from the model. The variables that had a statistically signi_cant relationship with the use and non-use of contraceptives were; current marital status, highest education level, religion, the total number of children ever born, frequency of listening to the radio, and wealth index. Roman catholicwomenwere 2.802 more likely to use contraceptives than those with other religions rather than Muslims, protestant, with no religion. Thewomen with no education were 0.188 less likely to adopt contraceptives than those with a higher education level. Married women were 2.058 times more likely to use contraceptives than those who were no longer living together with their spouses. In terms of wealth, the poorest women are 0.46 times less likely to adopt contraceptives than the wealthiest women under the study. Thosewomen who do not listen to the radio are 0.703 times less likely to use contraceptive than those who listen to the radio at least once a week.