Forecasting Delivery Time of Low-Risk Pregnant Women by Applying Linear Regression

The period of normal childbirth is the shortest, lasting no more than 24 hours, but it is the most important because up to 1 in 3 fetal deaths occur during birth. Accurate predictions of the time of birth can help health professionals provide effective care for the women during the time they give birth.

The aim of this research is to investigate the influence of cervical dilatation, the effacement of the cervix, station of the presentation, body mass index, maternal height, fetal weight, dose and duration of oxytocin exposure, and parity in low-risk pregnant women and to create a mathematical equation model for use in predicting the time to delivery.

This study is a retrospective descriptive study conducted from July 2023 to December 2023 at Thammasat Hospital. One hundred and eight low-risk pregnant women who had 37^{+ 0} to 41^{+ 6} weeks of gestation were selected by stratified random-sampling technique and systematic random sampling technique. The sample size was 108 participants. The research tool consisted of observation sheets and questions. Data analysis was obtained using multiple linear regression with the Stepwise regression method to examine the factor that influenced the time to delivery and create the equation.

The obtained model had an R² value of 0.316. The significant variables that mostly influence the time of delivery were the timing of oxytocin exposure (β = 0.31, p < .01) and cervical dilatation (β = -31.51, p < .01). The explanatory power of the regression model was statistically significant at 31.03%.

This study was designed for improving the prediction of time to delivery, which can be useful for enhancing the preparation pathways of normal childbirth. In this way, multiple regression analysis showed that the timing of oxytocin exposure and cervical dilatation can predict the time of birth.