Amanda Speller

Dissertation Title: "Essays on Maternal Health"

This dissertation focuses on exploring the maternity episode and it's associations with healthcare use, healthcare spending, and social policies. The first chapter uses survey data and a difference-in-differences analytic strategy to estimate the effect of a state-level paid family leave policy. The second chapter documents ultrasound and antenatal fetal surveillance use during the prenatal period. The third chapter estimates the incremental difference in health care use and expenditures between pregnant women with and without diabetes, hypertension, or mood and anxiety disorders compared to non-pregnant women with and without those conditions.

Chapter One: The association between New York's Paid Family Leave Policy and Use of Paid Leave After Childbirth (with Jessica Cohen and Maria Steenland)

In this paper, we examine whether New York's paid family leave policy increased paid leave use and leave duration among new mothers. We used the Pregnancy Risk Assessment Monitoring System (2016-2019), a population-representative survey of postpartum respondents, to compare trends in paid leave and leave duration in New York to trends in states that did not implement paid family leave during the study period. The policy was associated with a 9.1 percentage point increase in paid leave use, but was not associated with a change in leave duration. The policy did not decrease disparities in paid leave by race/ethnicity. However, there is some evidence that the policy decreased disparities between Medicaid-covered mothers and non-Medicaid-covered mothers. New York's PFL policy was associated with increased use of paid leave after childbirth but the policy did not reduce substantial disparities in paid leave between racial/ethnic groups. State-level PFL policies may be an effective policy option to increase paid leave use but additional emphasis on equity in program design may be needed to reduce disparities in postpartum paid leave.

Chapter Two: Title: Utilization and Spending on Prenatal Ultrasound and Antenatal Fetal Surveillance in a Commercially Insured Population (2017-2022) (with Kathe Fox, Joanne Armstrong, and Anna Sinaiko)

Ultrasound and fetal surveillance are important parts of prenatal care and have led to improvements in maternal and infant outcomes. This study aims to determine whether ultrasound use continues to increase, whether this increase is seen among those at high risk for adverse pregnancy outcomes, and assess the association between use and demographic factors. Diagnostic ultrasound utilization, AFS utilization, and expenditures were totaled for deliveries identified from 2017-2022 using the Health Care Cost Institute's commercial claims database. We describe ultrasound use by clinical and sociodemographic factors using descriptive statistics and negative binomial regression. Our findings are reported for the full cohort and for subgroups at relatively higher or lower risk for stillbirth. The majority of high-risk pregnancies did not receive any AFS monitoring. We document higher than expected use of CPT code 76816 (detailed anatomic scan) and meaningful increases in billing for CPT codes 76816 (follow-up ultrasound) and 76819 (biophysical profile). Although we estimate an increase in diagnostic ultrasound and AFS use, over half of pregnancies identified as being at high-risk for stillbirth do not receive the recommended AFS at any point during pregnancy. Policy makers should ensure that increased monitoring is accessible to women at higher risk for adverse pregnancy outcomes.

Chapter Three: Medical and Pharmacy Expenditures Among Reproductive Age Women with and without Diabetes, Hypertension, or Mood and Anxiety Disorders (with Anna Sinaiko)

Diabetes, hypertension, and mood/anxiety disorders diagnoses are increasing among reproductive age women. The purpose of this study is to describe the incremental difference in medical and pharmaceutical expenditures for each health condition among pregnant and non-pregnant women. We conducted a retrospective analysis of the Health Care Cost Institutes Commercial Insurance claims database to quantify health care spending and use in pregnant women and non-pregnant women from 2016-2021. We use two-part generalized linear models to estimates expenditures (medical, pharmaceutical, and total) and health care utilization across each condition for a 12-month period. The incremental medical cost of diabetes and hypertension was higher in pregnant women compared to non-pregnant women while the incremental cost of mood/anxiety diagnosis was lower. The incremental pharmaceutical cost was lower for pregnant women compare to non-pregnant women across all diagnoses. The incremental differences in therapy visits (among those with mood/anxiety disorders) and the number of unique prescriptions (across all conditions) were the greatest in magnitude. The incremental differences in glucose and A1C testing were negligible. Overall, we find that diabetes and hypertension were associated with higher total cost for pregnant women compared to the cost associated with these conditions for non-pregnant women. Mood and anxiety disorders were associated with a decrease in incremental total expenditures. Further research on the incrementally lower medical cost, pharmaceutical cost, and therapy visit rate compared with non-pregnant women among diagnosed with mood/anxiety disorders may provide information on whether women with these conditions have access to quality care.