Anticipated Impact on Veterans Health Care: Women are a growing share of the VA population, especially among younger veterans. A better understanding of the differential risk for reproductive outcomes among veterans will help VA provide better reproductive care to veterans. For example, this study will provide a better understanding of the preterm delivery risk associated with PTSD, which will help VA design programs to help mitigate this risk. It will also provide VA with a wealth of information about how the pregnancy risks of VA enrollees compare to the general population, including data on VA enrollees who use other forms of insurance for their pregnancy care. Project Background: Reproductive outcomes are sensitive to many factors, including environmental exposures and maternal co-morbidities. For example, there are small studies that suggest that service in the first Gulf War by either the mother or father was associated with increased risk for congenital anomalies. The PI has conducted a preliminary study with VA fee data that found that maternal PTSD is associated with spontaneous preterm delivery. We also found that this effect was larger in women who also screened positive for military sexual trauma (MST). The PTSD study only had data on the fact that a spontaneous preterm delivery occurred, with no data on extent of prematurity, other causes of preterm delivery, or infant outcomes. This study will fill a significant gap in our knowledge of the effectof maternal PTSD on preterm delivery and add a novel exploration of the effects of paternal PTSD on preterm delivery. Project Objectives: The project will use linked VA and California data to:
Aim 1 : Elucidate the effect of maternal PTSD on birth outcomes, confirming the increased risk of preterm birth, quantifying the distribution and character of such preterm births, and exploring other outcomes such as maternal complications (e.g., preeclampsia) and neonatal mortality.
Aim 2 : Determine if paternal PTSD status affects preterm delivery.
Aim 3 : Determine how pregnancy outcomes for Veterans (males and females) differ from those of the general population and provide information on the pregnancy risks and outcomes of veterans. Project Methods: We will link VA data with the California birth cohort linked data (VS-PDD) using birth certificates, infant death certificates, fetal death certificates, discharge abstracts for the mothrs and infants, including prenatal hospitalizations and infant transfers. The PI led the NIH-funded project that created the initial 10 years of the California VS-PDD data and has confirmed that sufficient data elements are available to link to VA data. We will expand this methodology to identify deliveries to VA enrollees (mothers and fathers) and to add VA data about them (e.g., OEF/OIF deployment, PTSD diagnosis). These linked data will allow for a more complete understanding of pregnancy outcomes than is possible from any of the component datasets. Projections from VA fee data and Medicaid data from the VA Information Resource Center (VIREC) indicate that this linkage will yield at least 13,000 deliveries to women veterans enrolled in VA, with the likely number being much larger. The comparison population is very large at over 500,000 birth/year. We will use regression analysis to control for medical (e.g., PTSD, hypertension) and social- demographic (e.g., age, race/ethnicity) risk factors to examine reproductive outcomes for different populations for a wide range of reproductive outcomes (e.g., preterm delivery, birth weight, pregnancy complications such as preeclampsia, congenital anomalies). We will compare VA enrollees to the non-VA population and examine special high risk groups (e.g., PTSD). The data will be maintained on the secure research server at the Palo Alto VA. CRITIQUE 1 1. Significance. This study is measuring the association between PTSD and pregnancy outcomes. It does so using three aims: * Relationship between maternal PTSD and birth outcomes; * Relationship between paternal PTSD and birth outcomes; * Difference in pregnancy outcomes between veterans and the general population. This is a resubmission of a previous grant proposal. The reviewers were responsive to feedback from earlier reviewers. The VA is increasingly purchasing/providing obstetrical care, and this is a relevant and feasible study with potentially significant implications for the VA. The proposal provides a thoughtful conceptual model which provides guidance regarding how they believe military enrollment and veteran-specific factors may impact pregnancy outcomes. This study leverages the linkability of veteran's VA-based health data and obstetric health services provided in non-VA settings. This linkable data can provide unique insights, on a population-level in California, into the relationship between birth outcomes and PTSD. Evidence in non-veteran studies has pointed to an effect of PTSD on pregnancy outcomes.
The first aim appears to be incremental, as this aim proposes to carefully examine the relationship between maternal PTSD and birth outcomes. The VS-PDD linkage provides a number of specific birth outcomes not routinely collected in VA data. This proposal extends the science by outlining a careful analysis of the roles of PTSD and deployment history in outcomes of: preterm birth, maternal complications, birth defects and death. The VA is likely the best place for this avenue of research since veterans have a higher prevalence of PTSD than the general population and routinely screens for PTSD.
The second aim examining the association between birth outcomes and paternal factors is novel. Since there is a basis in evidence paternal factors affecting pregnancy outcomes, the findings from this study will contribute to the literature. The objective of the third aim is to provide a basis of evidence regarding the pregnancy outcomes of veterans relative to non-veterans. There is a gap in the strength of this evidence, and the findings of this study will be important to the VA for policy and planning purposes. 2. Approach (including Feasibility). There are a number of technical questions. These questions are not major. How is deployment history (OEF/OIF registry) categorized and included in the model; how is the sensitivity of the categorization of this variable assessed? For instance, how is duration of service reflected in the model? The regional pollution levels provide insight into environmental factors that may impact birth outcomes, but they are poorly described. For instance, are the regional pollution levels variable by day or month, and how are temporal data incorporated into the models? The proposal describes linkage to census data to obtain neighborhood characteristics, but these are poorly described. The proposal describes that many missing paternal SSNs are related to undocumented immigration status; does this affect the generalizability or bias the results in a specific directin? 3. Impact and Innovation. The impact of this proposal is significant and addresses an area where research is sparse. 4. Investigator Qualifications, and Facilities and Resources. Excellent. 5. Multiple PI Leadership Plan. (if applicable) 6. Adequacy of Response to Previous Feedback Provided by HSR&D Regarding the Proposed Study. Very thorough response to previous feedback. 7. Protection of Human Subjects from Research Risk. Adequate for secondary data analyses. 8. Inclusion of Women and Minorities in Research. Adequate. 9. Budget. Adequate - modest budget for two year study. 10. Overall Impression. 11. Key Strengths. 1. Linking VA data with California's VS-PDD. 2. Investigators with excellent experience with these data. 3. Data driven study will provide results that address gaps in VA's understanding of the roles of maternal/paternal PTSD and obstetric outcomes. 12. Key Weaknesses. 1. Impact of the potential findings on VA policy is unclear. 2. The handling of temporal effects of some variables in the model is unclear.
Women, especially younger women, are a growing share of the VA population. A better understanding of the differential risk for reproductive outcomes among Veterans will help VA provide better reproductive care to Veterans. Reproductive outcomes are sensitive to many factors that are more common in Veterans, such as environmental exposures and co-existing maternal medical and mental health conditions. A specific focus of this study will be to build off of a study the PI has conducted to better understand the increased risk for preterm delivery associated with PTSD. This study will link VA data with detailed data on pregnancy from California to conduct analyses that are not possible with VA data alone. The investigators are partnered with the VHA Office of Women's Health Services to facilitate the translation of research findings into policy actions to help improve the reproductive outcomes of Veterans.
