Risk factors and predictors of occurrence of preterm birth in women with threatened preterm birth and validation of some recommended prognostic methods for probability of its occurrence

Abstract

ABSTRACT

Introduction: The prevalence of preterm birth has not decreased significantly in recent years. Preterm birth is a complex obstetrics complication, predicting preterm birth remains a medical dilemma. Nomograms have good predictive power due to the simultaneous use of several available clinical variables. No nomogram in Iran has been designed and validated to predict the occurrence of preterm birth. Objective: To determining risk factors and predictors of preterm birth in women threatened with preterm labor and validating some recommended methods to predict the probability of its occurrence. Methods: The present study was a prospective hospital-based cohort study, that was performed in Alzahra and Taleghani educational and medical centers in Tabriz. Written informed consent was obtained from all women participating in the study. During 14 months of recruiting and a total of 17 months of recruiting and follow-up of samples included in the cohort, 409 pregnant women were included in the analysis. Risk factors and predictors of preterm birth were determined within 48 hours, less than 34 weeks and less than 37 weeks. Alluoche nomograms were validated in predicting preterm birth at 48 hours, less than 34 weeks, and less 37 weeks. Binary logistic regression model and multivariate logistic regression were used to determine risk factors and predictors of preterm birth, respectively. STATA software also used to determine sensitivity, specificity, positive and negative predictive value, positive and negative likelihood ratios. We used R software to validate the Alluoche nomogram.

Results: Predictors of preterm birth within 48 hours include: rupture of the amniotic sac (AOR 4.71; 95% CI, 2.3-9.4), uterine contractions in tococardiography (AOR 6.69; 95% CI, 2.1-20.7), and dilatation of the cervix (AOR 3.41; 95% CI, 1.4-8.0), at the time of admission, gestational age 28-32 weeks (AOR 0.35; 95% CI, 0.1-0.9), living in the rural area (AOR 2.04; 95% CI, 1.1-3.9). These factors explained 30.5% of the observed variance. Predictors of preterm birth under 34 weeks include: rupture of the amniotic sac (AOR 15.85; 95% CI, 7.7-32.6), multiple pregnancies (AOR 5.56; 95% CI, 2.3-13.3), dilatation of the cervix at the time of admission (AOR 3.53; 95% CI, 1.8-6.8), having white blood cell counts of 16,000 and more at the time of admission (AOR 5.33; 95% CI, 1.9-14.6) and gestational age less than 28 weeks (AOR 3.24; 95% CI, 1.4-7.6), which explains 45.6% of the observed variance. Predictors of preterm birth under 37 weeks included: ruptured amniotic sac (AOR 7.99; 95% CI, 3.8-16.7), dilatation of the cervix at the time of admission (AOR 3.38; 95% CI, 2.1-5.4), pregnancy using assisted reproduction techniques (AOR 2.72; 95% CI, 1.5-4.8), gravidity fifth or more (AOR 4.89; 95% CI, 1.5-16.0), vaginal bleeding at the time of admission (AOR 1.77; 95% CI, 1.0-3.1) and experience of emotional violence during pregnancy (AOR 1.73; 95% CI, 1.1-2.7). These factors explained 29.7% of total observed variance. Among diagnostic tests; Cervical length disproportionate to gestational age showed the highest sensitivity (83.3%) and the lowest negative likelihood ratio (0.23) in predicting birth within 48 hours. The highest specifity (98.0%), positive predictive value (77.4%) and the highest positive likelihood ratio (19.35) were seen in Bishop score test seven and above. The highest negative predictive value (96.4%) was seen in the cervical funnelling test. Among diagnostic tests; Cervical length disproportionate to gestational age showed the highest sensitivity (67.9%), highest negative predictive value (86.4%) and lowest negative likelihood ratio (0.40) in predicting birth less than 34 weeks. The highest specifity (99.0%), positive predictive value (92.9%) and the highest positive likelihood ratio (24/10) were seen in Bishop score test seven and above. Among diagnostic tests; Cervical length disproportionate to gestational age showed the highest sensitivity (55.5%), highest negative predictive value (68.1%) and lowest negative likelihood ratio (0.49) in predicting birth less than 37 weeks. The highest characteristic (99.5%), positive predictive value (96.8%) and the highest positive likelihood ratio (26.60) were seen in Bishop score test seven and above. The results of the predictors of birth whithin 48 hours, less than and less than 37 weeks in the Allouche nomogram showed; decreasing the length of the cervix to millimeters on vaginal ultrasound, rupture of the amniotic sac, vaginal bleeding, and the presence of uterine contractions increase the probablity of birth whithin 48 hours, less than 34 and less than 37 weeks. The results of nomogram validation in predicting preterm birth showed: Allouche nomogram with area under ROC curve and 95% confidence interval for birth within 48 hours 0.89 [0.80-0.97] and for birth less than 34 weeks 0.89 [0.83-0.95] and for birth less than 37 weeks 0.87 [0.83-0.92] has a good ability to predict preterm birth.

Conclusion: The Allouche nomogram in the Iranian female population is also a valid tool in predicting preterm birth. This nomogram can be easily used in all centers equipped with vaginal ultrasound technique to determine the length of the cervix and will help physicians and patients in clinical decision making. Some predictors of preterm birth can be controlled, such as multiple pregnancies, assisted reproductive techniques, gravidity fifth and more, and experience of emotional violence during pregnancy. It is necessary to avoid preterm births due to multiple pregnancies that are the result of non-protocol based managements in the treatment of infertility. Also, by promoting childbearing in time, it will prevent the delay of pregnancy and consequently the need for assisted reproductive techniques. Implementation of family planning program, based on scientific evidence, will prevent increasing the number of pregnancies to five and more. Also, by educating the community and families to reduce violent behaviors against pregnant women, a number of preterm births will be prevented.