INDEED Collaborative
Here I provide an overview of my involvement in the INDEED Collaborative Project. This page includes an introduction to periviable delivery and the purpose of INDEED, the study methodology, and current findings from the most recent data analysis.
Background
What is periviable delivery?
Approximately 0.5% of births occur before the third trimester of pregnancy, and 40% of these births account for infant death.1 Infants that are born within the periviable gestational age (22-24 weeks of age) face significant morbidity and mortality rates since they are not developed enough to live outside of the womb without life-sustaining measures. Parents of periviable infants are tasked with making end of life decisions, often having to make these decisions quickly while in the delivery room. There are two courses of action that can be taken, resuscitation or comfort care. Resuscitation involves life-saving procedures that may be painful and many infants can experience long term cognitive and physical disabilities. Parents can also opt for comfort care, which does not involve any life-saving interventions, and let their baby pass away naturally.
The severity of disability and survival varies significantly by gestational age. For instance, infants who are born at 22 weeks have a NICU survival rate of 23–27% compared to 67–76% for those who are born at 25 weeks.1 A study of 4,000 periviable births between 2001 and 2011 found that the rate of survival and survival without neurodevelopmental impairment increased as gestational age increased; however the survival rate of infants who did develop neurodevelopmental impairments did not increase.1 This finding argues that an increase in survival rate was not simply a tradeoff for infants to live life with significant neurodevelopmental impairments. Figure 1 depicts survival rates by gestation age throughout time based upon studies conducted from 2010-2017 and Figure 2 illustrates the percentage of infants who survive with moderate to severe impairments.1
Investigating Neonatal Decisions for Extremely Early Deliveries” (INDEED)
The significant morbidity and mortality rates of periviable neonatal infants pushes for the need to identify factors that may contribute to neonatal survival rates, such as obstetric interventions and systematic guidelines. Obstetrical interventions (e.g. steroids tocolytics, magnesium sulfate, and antibiotics) are used to increase the chances of prolonging the pregnancy and to improve neonatal outcomes. However, because there is a lack of national standard guidelines for periviable treatment, the use of obstetrical interventions can vary by institution. The “Investigating Neonatal Decisions for Extremely Early Deliveries” (INDEED) multicenter collaborative sought to examine how obstetrical interventions and institutional practice guidelines impact neonatal survival and outcomes through a retrospective cohort study of 6 centers using data from 2011-2015. Women were more likely to receive obstetrical interventions when postnatal resuscitation was planned at 22 and 23 weeks, with survival rates increasing for infants born at 23 weeks gestation. Additionally, providers from centers that conducted high rates of resuscitation had a less favorable outlook on interventions for these newborns than those at low resuscitation centers, suggesting that attitudes of medical team members may be more closely associated with experiences of clinical outcomes.3 It was also found that mothers who received neonatal consultation when admitted into the hospital for periviable birth had a longer median admission-to-delivery intervals compared to mothers who did not. This finding concluded that Wide-spread consultations modeled by experienced neonatologists can offer better care approaches to increase the chances of a positive clinical outcome.2
This study is now undergoing a second initiative, “INDEED 2.0”, to utilize an updated and larger dataset, including five additional years of data (2016-2020) data and nine more centers, creating a combined total of 15 institutions in the collaborative. Data collection is currently underway. For purposes of this project, we report findings from our institution here at Indiana University and combine data from both INDEED 1.0 and 2.0 in order to compile new connections and discoveries by comparing both studies findings.
Methodology
The Collaborative
The INDEED 2.0 collaborative comprises 15 institutions: University of Chicago, Indiana University, Medical College of Wisconsin, University of Mississippi Medical Center at Jackson and University of Michigan, Johns Hopkins, Brandon Regional HCA, St. Vincent, Vanderbilt University, Madigan Army Hospital, University of Hawaii, University of Virginia, Columbia University, Northwestern University Prentice Women’s Hospital, and Advocate Christ Hospital. Each institution is responsible for collecting data from medical charts of periviable deliveries between 2011-2020. Because Indiana University was part of the original INDEED collaboration and contributed data from 2011-2015, only data from 2016-2020 was collected for this current project.
Eligibility & Data Collection
Indiana Health conducted a query of all periviable deliveries between 2016 and 2020 and provided the list of medical record numbers to the study team to be further evaluated for eligibility. The qualifications for the study population included adult (ages 18+ years) pregnant women who delivered at IU Health between 22 0/7 and 24 6/7 weeks gestational age. Babies with major congenital anomalies and life-limiting genetic conditions were excluded from the study, and only mothers presenting with live fetuses were included. Other exclusion factors included: patients who had Intrauterine fetal death (IUFD), patients with delayed interval delivery outside the periviable age, and patients who delivered outside of the IU medical facilities, infants with genetic/anatomic anomalies, and perinatal issues such as TTTS and fetal surgery.
Both the mother and baby/babies records were evaluated. Babies were divided into two groups: infants who were resuscitated in the delivery room and those who underwent comfort care measures. At our institution, the 2016-2020 data abstraction was a collaborative effort between myself, Catherine Groden, a neonatology fellow, and Jasmine Soo, a third-year Ob/Gyn resident. Each person was assigned 10-12 cases per year. I also made a Cerner guide that future interns can utilize for possible INDEED updates. The following variables were abstracted from the mother/baby charts: clinical circumstances; sociodemographics, aspects of antenatal counseling, the birth center, obstetric interventions, neonatal disposition, hospital length of stay, consultative services, and zip code. The medical charts were accessed via Cerner and data was directly recorded into REDCap, a secure and encrypted data management system.
Data Analysis
Data analysis for the entire INDEED 2.0 project is currently underway, as some sites are in the process of obtaining IRB approval and others are conducting data collection. For purposes of this project, we present an updated analysis that combined Indiana-specific data from INDEED 1.0 and 2.0 years (2011-2020). This analysis followed the same data analysis methods that were used in the "Intention to treat: obstetrical management at the threshold of viability" study that was led by our IU investigators using the INDEED 1.0 dataset. A univariate analysis and a mutlivariate analysis was conducted using linear regression models in order to control variables in order to find statistically significant findings regarding obsetric interventions and the effects on overall neonatal survival rates as well as being broken down by gestational age. Mothers who received the Bundled Interventions which included; Antenatal Steriods, Tolcolytics, Magnesium Sulfate, and GBS ppx were compared with mothers who did not.
Click the button below to visit the manuscript for specifics on data analysis methodology. The characteristics of mothers being a recipient of the following obstetric interventions was analyzed: Antenatal Steroids, Tocolytics, Magnesium Sulfate, GBS PPX, and Mode of Delivery (Vaginal versus Cesarean Section).
Results
Demographics of Study Population
There were a total of 183 study participants from the combined datasets. A total of 173 eligible women were identified for the INDEED 2.0 dataset. One record was excluded for not have any data variables and eight records were excluded for having deliveries outside of the periviable gestational window. The average maternal age was 27.8 years. The racial demographics of these mother’s included: 98 Caucasian (56.64%), 62 African American (35.83%), and 13 (7.51%) who were considered of other race/ethnicity. There were 152 singleton pregnancies (87.86%) and 21 multiple pregnancies (12.13%). 90 mothers were nulliparous (52.0%), and 80 mother’s were multiparous (46.24%). The findings for the type of healthcare insurance mothers had were the following: 83 mothers had Medicaid (47.97%), 74 (42.77%) had private healthcare, and 16 (9.24%) mothers were uninsured. 130 mothers had a planned resuscitation (75.14%) and 43 mothers (24.85%) did not.
Maternal age and race did not seem to affect if mothers were recipients of medication interventions; however it was found that planned resuscitation had a significant effect on whether mothers received obstetrical interventions. Mothers who had planned resuscitation were significantly more likely to receive magnesium sulfate and GBS PPX than mothers who did not have a planned resuscitation. Mode of delivery was also impacted by planned resuscitation. There were a total of 58 mothers who underwent a Cesarean. 54 mothers who had a planned resuscitation underwent a Cesarean (93.1%), while only 4 (6.9%) of mothers who did not have a planned resuscitation underwent a Cesarean. Mothers who had a parity of singletons were found to be more significantly likely to be offered GBS prophylaxis (GBS PPX). It was found that mothers who were uninsured were less likely to receive magnesium sulfate, only 5 out of the 16 uninsured mothers received it (3.15%).
Univariate Anaylsis (Tables 1 &2)
A univariate analysis was conducted in order to compare the survival rates of periviable neonates based upon administration of the following obstetric interventions: Antenatal Steroids, Tocolytics, Magnesium Sulfate, GBS ppx, and Mode of Delivery (Cesarean vs vaginal). It was found that administration of all obstetrics increased survival rates; however the most statistically significant interventions (P-Value = <0.01) that increased neonatal survival included the following: Antenatal Steroids (5x increased rate of neonatal survival), Magnesium sulfate (4x increased survival rate), and infants born via Cesarean Section (3x increased survival rate).
Obstetric Interventions administered to neonates of 22 and 23 weeks of gestational age were found to have the most significant effect on their survival rates, whereas neonates of 24-25 weeks did not have a significant increase in their survival rate. Neonates who were of 22 weeks gestational age that received: Magnesium Sulfate were 9x more likely to survive, Antenatal Steroids were 4x more likely to survive, and those born via Cesarean Section were 8x more likely to survive. It was found that at 23 weeks, Antenatal Steroids were the most statistically significant intervention, increasing survival rate by 4x.
Multivariate Analysis (Table 3)
A multivariable analysis was conducted through the use of a regression model in order to review survival rates while controlling variables such as age, race, insurance, chronicity, and parity. After controlling these variables, significant results in survival rates were still present within the following obstetric interventions: Antenatal Steroids (6x increased rate of neonatal survival), Magnesium sulfate (5x increased survival rate), and infants born via Cesarean Section (3x increased survival rate), and GBS ppx (2x increased survival rate).
Bundled Interventions (Tables 4 &5)
It was found that mothers and their neonates who received the Bundled Intervention had an overall increased survival rate of neonates across all gestational ages was increased by 3x (<0.01). More specifically, it was found that neonates of 23 weeks gestational age were found to be 2x more likely to survive when receiving this bundle compared to other mothers who did not. A multivariate analysis was conducted upon the bundled intervention as well, which resulted in almost a 4x increase in survival rate after controlling for maternal age, race, insurance, chorionicity, and parity. This multivariate analysis found that the bundled interventions were associated with a 3.9x increase in odds of survival in the 23-week gestation neonates after controlling the multitude of variables as described above, and this was found to be statistically significant (p = 0.04). All of the gestational ages had an increase in odds of survival, however they were not found to be statistically significant.
Learn More about the INDEED 1.0 Collaborative and the resulting publications below
Cerner Data Extraction Guide
I created a Chart Extraction Guide that can aid future interns in navigating Cerner for future INDEED data updates. It can also be used to aid interns in finding specific data from patient charts for other projects.
Cerner Guide edited.pdfReferences
1. Ecker, J. L., Kaimal, A., Mercer, B. M., Blackwell, S. C., deRwgnier, R. A., Farrel, R. M., ... Sciscione, A. C. (2017, October). American College of Obstetricians and Gynecologists and the Society for Maternal–Fetal Medicine . Periviable Birth, 130(4), e187-e199. Retrieved from https://www.acog.org/clinical/clinical-guidance/obstetric-care-consensus/articles/2017/10/periviable-birth
2. Feltman, D. M., Fritz, K. A., Datta, A., Carlos, C., Hayslett, D., Tonismae, T., Lawrence, C., Batton, E., Coleman, T., Jain, M., Andrews, B., Famuyide, M., Tucker Edmonds, B., Laventhal, N., & Leuthner, S. (2020). Antenatal Periviability Counseling and Decision Making: A Retrospective Examination by the Investigating Neonatal Decisions for Extremely Early Deliveries Study Group. American journal of perinatology, 37(2), 184–195. https://doi.org/10.1055/s-0039-1694792
3. Lawrence, C., Laventhal, N., Fritz, K. A., Carlos, C., Famuyide, M., Tonismae, T., Hayslett, D., Coleman, T., Jain, M., Edmonds, B. T., Leuthner, S., Andrews, B., & Feltman, D. M. (2020). Ethical Cultures in Perinatal Care: Do They Exist? Correlation of Provider Attitudes with Periviability Practices at Six Centers. American journal of perinatology, 10.1055/s-0040-1709128. Advance online publication. https://doi.org/10.1055/s-0040-1709128
4. Tonismae, T. R., Tucker Edmonds, B., Bhamidipalli, S. S., Fadel, W. F., Carlos, C., Andrews, B., Fritz, K. A., Leuthner, S. R., Lawrence, C., Laventhal, N., Hayslett, D., Coleman, T., Famuyide, M., & Feltman, D. (2020). Intention to treat: obstetrical management at the threshold of viability. American journal of obstetrics & gynecology MFM, 2(2), 100096. https://doi.org/10.1016/j.ajogmf.2020.100096
Acknowlegments
Special thanks to the following people who made my internship and the INDEED Initiative Possible: Brownsyne Tucker-Edmonds, Shelley Marie Hoffman, Catherine Groden, Jasmine Soo, Sruthi Bhamidipalli, Dalia Feltman, and the INDEED Collaborative.
