Selecting the best statistical model for analyzing the mortality of preterm infants (applied study)
A Thesis
Submitted to the Council of Faculty of Administration and Economics at the University of Karbala as a partial Fulfilment of the Requirements for the Degree of Master in statistics science
By
Maithem Abdel Wahab Saleh
Under supervision of
PROF. Dr . Adnan Karim Najm ALdin
Abstract
The level of premature mortality is an important indicator of the progress of health services and health education in any society. This topic has been selected due to a few studies that have been examined in this area, especially in the province of Babylon and in Iraq in general.
To achieve this, the researcher relied on the data obtained from the Department of Health of Babylon, which included the monthly totals of the number of preterm infants for the period 2014-2016, and included data for each child in the unit of preterm infants and the condition of exit (alive or dead).
The hypothesis of the study was to determine the effect of a number of factors (pregnancy, child weight, age of the child, age of the mother, Birth status , type of birth, sterilization) on the dependent variable. With a view to developing solutions to reduce this phenomenon and to develop measures to counter or control it in the future through the formulation of medical and health policies that contribute to reducing mortality, as well as the introduction of maternal and newborn health programs, continuity of care including prenatal services, the adoption of qualified staff during childbirth and follow-up during the first month of life .
The purpose of this thesis is to analyze this phenomenon and to show its side effects, in addition to building a suitable statistical model for the nature of the data, through which we can know the effect of these factors on the mortality of premature infants of each variable, and predict the number of deaths in the next stage , Such as multiple regression model, logarithmic regression model, exponential regression model for quantitative data study, logistic regression model, differential function analysis for the study of descriptive response data.
The study found a significant relationship between the different variables (influencing factors) and the dependent variable. The study also showed the suitability of statistical models based on a number of statistical indicators and tests.
The study finding some conclution, and recommendations , the most important of which is the need to pay attention to the level of medical care for preterm infants , Paying attention to factors affecting child mortality and developing appropriate solutions. The thesis also recommends the use of the logarithmic regression model in the case of quantitative data and logistic regression model in the case of dependent binary variable data.
