Induction Of Labor: Predictors of Outcomes
- Conditions
- Induction of Labor Affected Fetus / Newborn
- Interventions
- Drug: induction of labor
- Registration Number
- NCT04350437
- Lead Sponsor
- Assiut University
- Brief Summary
Induction of labor is a widely used intervention in OBGYN practice. Doctors still use the old Bishop score in patients' follow up. It remains difficult to anticipate the outcomes and the possibility of adverse effects during this process. In this large prospective multicentric interventional study, we aim to develop a more precise and sensitive score based on machine learning tools programmed on python 3.8
This new tool will account for many variables in patient demography(age, race, weight ... etc ) and medical history (previous OBGYN surgery, comorbidities .... etc). These variables not usually found in the classic bishop score. We predict that our analysis will aid doctors in making better decisions and efficiently predict the outcomes, need for switching to operative delivery and possible complications.
Machine learning and digital calculation of hazards will allow more precise assessment and more efficient management during IOL as it considers variables not included in clinical scores.
this study aims to provide modern and efficient assessment parameters to guide clinical decision making during the IOL process and help doctors predict its outcomes based on subtle factors not usually considered.
This will minimize the complications and allow more evidence-based practice.
- Detailed Description
the objective is to create a database registry documenting the induction of labor (IOL) process and apply machine learning tools to create a more precise assessment score for doctors as a contemporary follow-up method.
we will collect data from at least 12 centers worldwide describing the course, outcomes, maternal or fetal complications, and any related data. The data will be collected after ethical approval and from consenting patients in a prospective manner. during the period from July 1st, 2020 to June 30th, 2021 (anticipated dates).
each center will be responsible for quality assessment, data collection, and ensuring the data is accurate, complete, and representative.
Data collection includes baseline pelvic examination (cervical position, consistency, dilation, effacement, fetal position, and bishop score), method of induction and their time of administration in relation to index time (start of IOL), findings and time of serial pelvic examinations, fetal heart tone, and maternal vital signs. The entry of data from serial examinations will continue during active labor and fetal and maternal outcomes will be reported. If the diagnosis of failed IOL is made and obstetric team decides delivery by Cesarean section, criteria of diagnosis/indication of Cesarean delivery will be reported. Length of active labor and the second stage will be documented, and maternal/perinatal complications will be reported. the collectors must ensure patient confidentiality and safety.
Inclusion criteria:-
* Pregnant women admitted for IOL, aged between 18 to 40 years
* Term or late preterm pregnancy (gestational age at 34 weeks or beyond)
* A reassuring fetal heart tracing prior to IOL
Exclusion criteria:-
* Fetal growth restriction with abnormal Doppler indices
* Intrauterine fetal death
* Suspected intra-amniotic infection prior to IOL
* Fetal major congenital anomalies
* Patients who decline IOL in prior or during IOL without medical indication
statistical analysis :- Data will be described using (mean, median, standard deviation, range) in the final sample. Machine learning method is superior to traditional statistical methods as it provides robust and automatic estimation of complex relationships between different variables and clinical outcomes. Data will be utilized as xi and yi where xi presents input (features) and yi presents dependent variables (outcomes). Functional regression is based on support vector machine by regressing the outcomes yi on inputs xi. Model Validation will be performed via bootstrap estimation to evaluate the predictive ability of the functional regression models. Data will be split to training data (approximately 63% of the data) to create prediction model where bootstrapping will be applied, and testing data where prediction model will be validated. Machine learning models will be created using python 3.8.
Recruitment & Eligibility
- Status
- UNKNOWN
- Sex
- Female
- Target Recruitment
- 3000
- Pregnant women admitted for IOL, aged between 18 to 40 years
- Term or late preterm pregnancy (gestational age at 34 weeks or beyond)
- Reassuring fetal heart tracing prior to IOL
- Fetal growth restriction with abnormal Doppler indices
- Intrauterine fetal death
- Suspected intra-amniotic infection prior to IOL
- Fetal major congenital anomalies
- Patients who decline IOL in priori or during IOL without medical indication
Study & Design
- Study Type
- INTERVENTIONAL
- Study Design
- SINGLE_GROUP
- Arm && Interventions
Group Intervention Description induction of labor monitoring induction of labor meticulous data collection from patients and plotting that data in a machine learning model
- Primary Outcome Measures
Name Time Method Cesarean section rate Within 24 hours from start of induction of labor Incidence and indication of Cesarean section following induction of labor
- Secondary Outcome Measures
Name Time Method Suspected intraamniotic infection From start of induction of labor to 24 hours after delivery Maternal pyrexia \> 39 or \> 38 on 2 occasions
Postpartum hemorrhage From start of induction of labor to 24 hours after delivery Blood loss \> 1000 ml after delivery
Low neonatal APGAR Score 5 minutes after delivery APGAR score \< 7 at 5 minutes postpartum
Admission to neonatal intensive care unit Within 1 hour of delivery Admission of the newborn to intensive care unit and its indication