Modeling Mode Of Childbirth Delivery Using Dummy Dependent Variable Models
Abstract
In many applications, the dependent variable is nominal or ordinal, thereby, standard linear model is not suitable for modeling such dependent variables. Mode of delivery is nominal; it is either by spontaneous vaginal delivery or by caesarean section (CS). This study then investigates factors affecting mode of delivery using binary dummy dependent models. Data on mode of delivery and risk factors such as Mother’s age (years), child's sex, child's weight (kg), child's length (cm), child's head circumference (cm) and mode of delivery were collected from State Hospital Ota, Ogun State over a period of eight months in order to identify factors that affect mode of delivery. Binomial logistic and probit regression models were fitted into the data to determine the preferred model for the analysis of mode of delivery. AIC was used to determine the competence of the models. Probit model with the lower AIC was found to be more adequate for the data and was used to identify the factors affecting mode of delivery. The result showed that maternal age and child’s weight have significant effects on the probability of having a child through CS. Also, children with head circumference less than 35 cm are less likely to deliver via CS than those with head circumference between 20-35 cm, though not significant. It is recommended that the health practitioners should take cognizance of the factors affecting mode of delivery as identified by this research work.
Keywords: Binomial logistic regression, Caesarean section, Probit regression, Risk factors, Vaginal delivery
for the data and was used to identify the factors affecting mode of delivery. The result showed that maternal age and
child’s weight have significant effects on the probability of having a child through CS. Also, children with head
circumference less than 35 cm are less likely to deliver via CS than those with head circumference between 20-35 cm,
though not significant. It is recommended that the health practitioners should take cognizance of the factors affecting
mode of delivery as identified by this research work.
Keywords: Binomial logistic regression, Caesarean section, Probit regression, Risk factors, Vaginal delivery
