Developing and refining methods to assess medical product utilization, safety, and effectiveness during pregnancy is a focus of FDA’s Sentinel System. The Sentinel Common Data Model (SCDM) includes a Mother-Infant Linkage (MIL) table that enables routine evaluation of the effects of medical product exposures during pregnancy on infant outcomes. Descriptions of efforts led by the Center for Drug Evaluation and Research are shown below. Please visit the links to learn more about each area of activity.
The US Food and Drug Administration established the Sentinel System to monitor the safety of medical products. A component of this system includes parameterizable analytic tools to identify mother-infant pairs and evaluate infant outcomes to enable the routine monitoring of the utilization and safety of drugs used in pregnancy. We assessed the feasibility of using the data and tools in the Sentinel System by assessing a known association between topiramate use during pregnancy and oral clefts in the infant. We identified mother-infant pairs using the mother-infant linkage table from six Data Partners contributing to the Sentinel Distributed Database from January 1, 2000, to September 30, 2015. We compared mother-infant pairs with first-trimester exposure to topiramate to mother-infant pairs that were topiramate-unexposed or lamotrigine-exposed and used a validated algorithm to identify oral clefts in the infant. We estimated adjusted risk ratios through propensity score stratification.
Traditional surveillance of adverse infant outcomes following maternal medication exposures relies on pregnancy exposure registries, which are often underpowered. We characterize the statistical power of TreeScan™, a data mining tool, to identify potential signals in the setting of perinatal medication exposures and infant outcomes. We used empirical data to inform background incidence of major congenital malformations and other birth conditions. Statistical power was calculated using two probability models compatible with TreeScan, Bernoulli, and Poisson, while varying the sample size, magnitude of the risk increase, and incidence of a specified outcome. We also simulated larger exposure to referent matching ratios when using the Bernoulli model in the setting of fixed N:1 propensity score matching. Finally, we assessed the impact of outcome misclassification on power.
This presentation identifies and describes pregnancies in claims-based data sources which presents logistical and inferential challenges. FDA’s Sentinel System included data from six sites with linked infant and maternal data, and reusable, parameterizable tools to identify and describe these cohorts. These tools and the methods for identifying pregnancies are publicly available and may be used by anyone following Sentinel’s Common Data Model. In this symposium, Sentinel Operation Center presenters discuss the publicly available analytic tools and recent pharmacoepidemiologic analyses related to pregnancies and outcomes within Sentinel.
It was presented at the Society for Pediatric and Perinatal Epidemiologic Research (SPER) webinar on September 13, 2022.
It is a priority of the US Food and Drug Administration (FDA) to monitor the safety of medications used during pregnancy. Pregnancy exposure registries and cohort studies utilizing electronic health record data are primary sources of information but are limited by small sample sizes and limited outcome assessment. TreeScan™, a statistical data mining tool, can be applied within the FDA Sentinel System to simultaneously identify multiple potential adverse neonatal and infant outcomes after maternal medication exposure. We implemented TreeScan using the Sentinel analytic tools in a cohort of linked live birth deliveries and infants nested in the Merative™ MarketScan® Research Databases. As a case study, we compared first trimester fluoroquinolone use and cephalosporin use. We used the Bernoulli and Poisson TreeScan statistics with compatible propensity score-based study designs for confounding control (matching and stratification) and used multiple propensity score models with various strategies for confounding control to inform best practices. We developed a hierarchical outcome tree including major congenital malformations and outcomes of gestational length and birth weight.
This presentation provides an overview of the Sentinel Distributed Database and its use of real-world data for the surveillance of medications in pregnancy.
It was presented at the FDA Office of Women's Health (OWH) and Johns Hopkins University Center of Excellence in Regulatory Science and Innovation (JH-CERSI) collaborative workshop, held on May 9, 2022.
This presentation provides an overview of TreeScan signal identification methods and discusses how such methods can be used for surveillance of potential adverse infant events following maternal medication exposure during pregnancy. Methods and results from a simulation analysis to assess the performance of TreeScan under known conditions are reviewed, as well as a case study which aimed to demonstrate the use of TreeScan in real-world data, in a cohort of pregnant women linked to their live-born infants.
It was presented at the Canadian Network for Observational Drug Effect Studies (CNODES) Virtual Seminar, held on March 9, 2022.
In this report we describe exposure to multiple sclerosis (MS) drugs before, during and after pregnancies resulting in a live-born delivery among women in the Sentinel Distributed Database (SDD). We also describe MS drug use among matched women during episodes without evidence of a live-birth delivery.
The study period includes data from January 1, 2001 to December 31, 2020. We distributed this request to six Sentinel Data Partners on June 25, 2021.
This is a joint agency project involving the Food and Drug Administration (FDA) and National Institutes of Health/ National Library of Medicine (NIH/NLM) that addresses the Patient-Centered Outcomes Research (PCOR) priority to expand data capacity or data infrastructure for conducting research that informs decisions about the effectiveness of health interventions used in the Medicaid and Children's Health Insurance Programs.
This project aims to develop and publish open source code to format T-MSIS Research Identifiable File (RIFs) to the Sentinel CDM in tandem with NLM’s formatting of the T-MSIS RIFs to the OMOP CDM. Data quality will be characterized using a harmonized Data Quality Assessment Framework for electronic healthcare data that was developed in a previously funded OS-PCORTF project. To illustrate the benefits of CDM transformation, open source tools will be used to conduct several studies of public health importance, such as studies of maternal mortality and adherence to clinical guidelines for perinatal testing. Additionally, the project will assess the feasibility of using Fast Healthcare Interoperability Resources (FHIR) APIs (Application Programming Interfaces) to link Electronic Health Record (EHR) data to T-MSIS claims data. Lastly, the project will ensure stakeholder engagement and sustainability via a Technical Expert Panel (TEP) to provide guidance on the data quality metrics, the PCOR demonstration projects, and development of a series of training materials. The goal of the training materials is to ensure findings are optimally disseminated to the Medicaid research community to lead to greater utilization of the CDM translation code and use of T-MSIS for patient-centered outcomes research in Sentinel and OMOP ecosystems.
This project covers several major tasks:
Task 1: Develop open source code to format the T-MSIS data into the Sentinel and Observational Medical Outcomes Partnership (OMOP) Common Data Models (CDMs)
Task 2: Leverage the ASPE funded data quality metrics model to characterize each CDM-formatted version
Task 3: Develop open-source code to create a mother-infant linkage in the Sentinel CDM to support several Sentinel analyses on maternal health
Task 4: Conduct one or more demonstration studies assessing the harms and benefits of an aspect of maternal health using the transformed T-MSIS dataset and Sentinel analytic tools. The FDA-led workgroup will include participation from CDC/NCHHSTP, CDC/NCBDDD, NIH/NICHD and HRSA
Task 5: Evaluate the feasibility of implementing FHIR APIs to link T-MSIS data with EHR data
Task 6: Engage stakeholders and create sustainability by:
- Establishing a TEP with relevant expertise in the Medicaid program, T-MSIS data, or patient-centered outcomes research to provide non-binding guidance from an end-user’s perspective on the data quality metrics selected in Task 2, potential PCOR demonstration projects using the new T-MSIS dataset in Task 4, and the structure of the training materials
- Developing a webinar series to train Medicaid researchers on the new data transformation tools and disseminate major project findings
This study aims to describe the feasibility and generalizability of a mobile app for capturing data on health conditions, medication use, and other relevant exposures during pregnancy and linking those data to the patient’s medical record research database.
The following presentation was presented at the 37th International Conference on Pharmacoepidemiology and Therapeutic Risk Management in 2021.
The goals of this methods project are to evaluate the performance of TreeScan to assess maternal and infant outcomes, test signal identification methods in a pregnancy setting, and evaluate methods performance using older drugs with relatively well-characterized safety profiles.
A Propensity Score-Based TreeScan approach will be the primary analysis for this project. There will be two analyses: 1) empirical (real) data set and a 2) simulated data set with investigator-inserted risks. Two protocols will be written for Aims 2 and 3, respectively, and posted for public comment.
Aim 1:
- a) Develop a mother-infant linkage programming module, to be executed in Merative™ MarketScan® Research Databases, that produces an SCDM-compliant mother-infant linkage table using a published method.
- b) Develop a general propensity score for pregnant women.
- c) Develop tree of infant outcomes in ICD-10-CM and develop tree of maternal outcomes in ICD-10-CM.
Aim 2:
- a) Assess the performance of TreeScan to detect infant outcomes using empirical data with the following outcomes: congenital malformations as the primary objective and a selection of additional outcomes including small for gestation age, low birth weight, and increased rate of premature birth as the secondary objectives.
- b) Perform a simulation study with investigator-injected risks to develop data on the power to detect risk under ideal circumstances, using empirical data to develop background rates.
Aim 3:
- a) Assess the performance of TreeScan to detect maternal outcomes using empirical data among pregnant women with live birth outcomes.
- b) Perform a simulation study with investigator-injected risks to develop data on the power to detect risk under ideal circumstances, using empirical data to develop background rates.
This infant outcomes protocol was posted for public comment from June 19, 2020 through July 6, 2020. The public comment period is now closed. A revised version was approved for implementation by FDA on August 17, 2020. A second revised version was approved for implementation by FDA on February 9, 2021. A log of changes is included in the revised protocol (v3).
This maternal outcomes protocol was posted for public comment from June 29, 2021 through July 13, 2021. The public comment period is now closed. A revised version was approved for implementation by FDA on August 17, 2022. A log of changes is included in the revised protocol (v2).