The present study was designed to predict type of childbirth using input variables to the Decision Tree model. This longitudinal study was conducted on 170 pregnant mothers with inclusion criteria were referred to Shahroud Health Care Centers for pregnancy care during the third trimester of pregnancy from 2018 to 2019. Inclusion criteria were: Iranian citizenship, single pregnancy, having electronic records in the health system, no history of illness and no onset of labor pain. A questionnaire including demographic and socio-economic variables, pregnancy history and WHO-5 well-being index were completed.
Measurements:
1- WHO-5 well-being index
The WHO-5 questionnaire consists of five questions about the feelings of the participant during the previous two weeks, with each item being scored on a 6-points Likert scale of 0–5. Raw ratings are converted onto a scale from zero to 100. The validity and reliability of Persian version of this scale was assessed in the Iranian population (7). The validity of the WHO-5 questionnaire on pregnant mothers up to 8 weeks postpartum was assessed in a study, the Cronbach's alpha coefficient for the WHO-5 items was 0.85 and a score of 50 or less with a good sensitivity of 84% and a specificity of 59% for psychological symptoms identification used (8). A 2015 review identified WHO-5 as a valid tool for screening for depression (9).
2- Socio-economic index
The socio-economic status was constructed using Principal component analysis (PCA) with combining 3 main factors including economic indicators (occupation, spouse occupation, homeownership status, a separate bedroom for couples, number of bedrooms, indoor bathroom, cooking area), asset-index (including refrigerator, freezer, Color TV, washing machine, dishwasher, microwave, vacuum cleaner, personal car, landline, mobile phone, computer or laptop, internet access) and social factors (education, spouse education, family members, family supplementary health insurance). Finally, 15 variables were used in constructing socioeconomic status variables, which explained 17% of the total variance in this new variable. After calculating a variable called socioeconomic status, based on the twenty-fifth and seventy-fifth percentiles of these variables, the population was grouped into three socio-economic groups with high (41 persons), medium (84 persons) and low (38 persons).
The height and weight of the mothers and attendance in childbirth preparation classes was recorded and then 3cc blood samples were taken from non-fasting antecubital vein (often 9 am to 11 am) to measure estrogen hormone. The blood samples were transferred to the laboratory immediately and centrifuged by the laboratory officer and, before analysis; the samples were frozen at -80 ° C after plasma separation. Serum estradiol E2 levels were assayed by enzyme-linked immunosorbent assay (ELISA) method (Monobind kit, China).
It should be noted that in coordination with mothers during pregnancy and Estimated Date of Confinement (EDC) estimates indicating the probable date of childbirth, we were informed of the time of delivery and childbirth information was recorded at the time of referral to the health centers at 30–42 days postpartum. It should be noted that for 4 of the samples due to emigration from Shahroud and 10 of them due to non-cooperation, the required information was obtained by telephone from 11 of them and finally, 163 samples were analyzed (Fig. 1). Participants' blood samples examined at the end of the second stage and after ensuring that the type of childbirth data was collected, Serum estradiol E2 levels were assayed by ELISA method (Monobind kit, China).
Data analysis
The data were analyzed with SPSS-21 and MATLAB softwares. Chi-square and t-test were used for initial comparisons between the two groups. P-value < 0.05 was considered as the significant level. A C4.5 decision tree algorithm was developed by using input variables and the determination of the target variable.
The Input Variables including
maternal age, family socioeconomic status, previous type of childbirth, maternal mental health status during pregnancy, maternal birth body mass index (BMI) and maternal serum estradiol E2.
The Output Variables including
Type of childbirth (Vaginal delivery and cesarean section were coded 0 and 1, respectively)
Design of model
At this point, the data is divided into training and testing datasets using the 70 − 30% method, 70% of which is the training data and 30% of the data is considered as the test dataset. The decision tree algorithm is implemented and formed on the training data. The decision tree is then evaluated by training and test dataset. Figure 2 shows the structure of the prediction algorithm.