Estimated Prevalence of Risk Factors for Preeclampsia Among Individuals Giving Birth in the US in 2019

IMPORTANCE Low-dose aspirin (LDA) is one of the few evidence-based interventions for preventing preeclampsia, which is a leading cause of maternal or fetal morbidity and mortality. Current guidelines recommend LDA based on the presence of risk factors for preeclampsia, but the population-based prevalence of these factors is unknown. OBJECTIVE To estimate population-level prevalence of preeclampsia risk factors used in prophylactic LDA guidelines for pregnant patients and the association of these risk factors with reported rates of pregnancy-related hypertension. DESIGN, SETTING, AND PARTICIPANTS A retrospective cohort study was conducted using National Center for Health Statistics birth certificate data to describe the frequency of pregnant individuals with moderate to high-risk factors for preeclampsia and pregnancy-related hypertension rates. The study used all birth records in the United States for the 2019 calendar year. EXPOSURES Documentation of preeclampsia risk factors: multifetal gestation, pregestational diabetes, chronic hypertension (high-risk factors) and nulliparity, a body mass index greater than 30, African American race, a maternal age 35 years or older, an interval of more than 10-years since last birth, and having low socioeconomic status (moderate risk factors). and cost-effectiveness of universal LDA are favorable compared to risk-factor based strategies given the risks associated with preeclampsia and the low cost and favorable safety profile of LDA. 5,13,14 Similarly complex risk stratification protocols for influenza vaccination were abandoned because of their complexity and lead to universal vaccination guidelines.


Introduction
Preeclampsia is a systemic hypertensive condition marked by elevated blood pressures and proteinuria that is unique to pregnancy. Preeclampsia increases the risk of fetal growth restriction, preterm birth, Cesarean delivery, long-term maternal cardiovascular disease, and is a leading cause of maternal and fetal morbidity and mortality. 1 In a recent meta-analysis of 16 clinical trials, low-dose aspirin (LDA) reduced the risk of preeclampsia by 15%, 1-3 especially when it was started before 16 weeks' gestation. The United States Preventive Services Task Force (USPSTF) recommends LDA (81 mg/d) for the prevention of preeclampsia in individuals with a preeclampsia risk of 8% or higher, based on the presence of certain known risk factors. 4 The 2014 USPSTF are complex, with a definite recommendation for LDA if 1 or more of 6 highrisk factors and consideration of LDA 2 or more of 11 moderate-risk factors are present. The recently updated 2021 guidelines are similarly complex. Similar to 2014, LDA is definitively recommended if 1 or more high-risk factors is present. The 2021 guidelines now recommend for LDA for a combination of moderate-risk factors and encourage consideration if a single moderate-risk factor is present.
Given the continued complexity of these recent guidelines, the overall safety of LDA in pregnancy, and the potential consequences of preeclampsia for maternal and fetal outcomes, some researchers have argued for universal use of LDA in all pregnant individuals. 5,6 Direct population-based data estimating the number of pregnant individuals who are candidates for LDA based on USPSTF guidelines and correlation with pregnancy-related hypertension outcomes are lacking. This study was conducted to determine population-level estimates of specific risk factors for preeclampsia and LDA candidacy.

Methods
This retrospective cohort study uses publicly available birth certificate data from the National Center for Health Statistics (NCHS) calendar year 2019. 7 The study protocol was approved by the institutional review board at Duke University and informed consent was waived because deidentified publicly available data were used in the analysis. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
All recorded births from 2019 were included in the analysis. The exposures of interest were high and moderate-risk factors for preeclampsia based on the USPSTF guidelines. 1,4 The standard birth certificate records 3 of the 6 high-risk factors for which LDA is definitively recommended: multifetal gestation, pregestational diabetes, and chronic hypertension. The remaining 3 high-risk factorspreeclampsia in a prior pregnancy, renal disease, and autoimmune disease-are not recorded in birth records. Among individuals without a documented high-risk factor, we assessed moderate-risk factors, including nulliparity, body mass index (BMI, calculated as weight in kilograms divided by height in meters squared) over 30, African American race, maternal age 35 years or older, more than a 10-year interval since last birth and low socioeconomic status (SES, defined by Medicaid as primary payer or participation in the Women, Infant, and Children [WIC] nutritional supplement program).
The 2021 USPSTF guidelines also include in-vitro conception as a moderate-risk factor and the NCHS data captures assisted reproductive technology, which includes in vitro fertilization and intracytoplasmic sperm injection. The USPSTF guidelines include data on the African American race, thus that data was collected and presented in the results. Having a family history of preeclampsia, personal history of low birth weight, and being small for gestational age are moderate-risk factors that are not captured in birth records. The primary outcomes were the frequencies of each risk factor alone and in combinations that would lead to recommendations for LDA. For estimation purposes, we assumed that missing data on a given risk factor meant that the factor was not present.
LDA is most effective when started before 16 weeks' gestation; therefore, we examined the frequency of patients meeting LDA criteria who initiated prenatal care by 16 weeks' gestation.
Although preeclampsia is not specifically recorded on the birth certificate, gestational hypertension (gHTN, including pregnancy-related blood pressure elevation and preeclampsia), and eclampsia (ie, seizures associated with preeclampsia) are recorded. We estimated the proportion of individuals with different combinations of risk factors who developed gHTN or eclampsia, collectively termed pregnancy-related hypertension. We estimated a number needed to treat (NNT) assuming LDA reduces overall pregnancy-related hypertension by an equivalent amount relative to the estimated relative risk reduction in preeclampsia used for developing the current guidelines of 0.85. 1 No formal statistical testing was done, however, we calculated 95% CIs assuming a range of preeclampsia reduction from 0.75 to 0.95 consistent with the 2021 guidelines. 1 All analyses were performed in JMP, version 15 (SAS Institute) and conducted from July to November 2021.

Results
There were 3 695 019 pregnancies that resulted in a live birth in 2019. The mean (SD) age of the cohort was 29.

Discussion
In 2019, more than 85% of pregnant individuals met the 2021 USPSTF criteria to recommend or consider LDA to prevent preeclampsia. More than 80% who met these criteria presented for prenatal care within the optimal time-period to initiate therapy. Birth records with documented conditions meeting current guidelines for LDA account for almost two-thirds of cases of gestational hypertension and eclampsia.
The single most commonly documented risk factor was low SES. Studies documenting an increased risk of preeclampsia with low SES use varying definitions. We used Medicaid as primary payer and participation in the WIC nutritional supplement program as markers of low SES, however the eligibility criterion for these programs varies by state. These strict criteria likely underestimate the extent of patients with low SES. Even with the likely underestimation, we observed a 7.8% and 4%  prevalence of pregnancy-related hypertension with low SES overall and low SES in isolation, suggesting the clinical importance of this factor.
The African American race is also considered a moderate-risk factor, and there is a growing understanding that race is not a genetic or biological construct, but rather race measures an individual's lived experiences with racism and discrimination. 8 Although self-report is the criterion standard 9 and there is likely some variation in how race is captured across hospitals, studies suggest that birth record data are highly accurate in race reporting compared with direct patient interviews. 10 The inclusion of SES and race in the USPSTF guidelines highlights the importance of ensuring clinicians ascertain data about these and other social determinants of health with a clear understanding that health inequities rather than biological differences drive these associations. In

Limitations
Although our findings add to the growing conversation about strategies to maximize LDA in pregnancy, several important limitations must be considered. Birth record data are entered by hand and subject to errors in reporting. Literature suggests birth record data accurately capture pregnancy-related hypertension yet may underestimate chronic hypertension. 16 Birth records also do not record several high and moderate-risk factors, which may have undercounted how many patients meet the criteria for LDA. The omission of these risk factors suggests we have underestimated the proportions of patients that meet LDA criteria. However, many of the conditions not captured are relatively rare (autoimmune and renal disease) or are rarely captured in the medical record (family history of preeclampsia), suggesting a limited impact on the estimates presented.
Approximately 20% of patients with a personal history of preeclampsia have recurrent preeclampsia in a future pregnancy. It is unclear whether a personal history of preeclampsia influences the incidence of future pregnancy; therefore, estimating the impact of this omission is challenging.
Furthermore, the prevalence of the high-and moderate-risk factors are not fixed, and many risk factors are increasing in prevalence. Based on likely increases in many of the risk factors over time, we suspect the current NNT estimates are likely overestimations.
The birth record documents pregnancy-related hypertension and is inclusive of individuals with gHTN who do not meet strict preeclampsia criteria. However, this broader definition is unlikely to have a meaningful impact on our estimates because gHTN alone is associated with adverse outcomes, and there is progression to preeclampsia in up to 50% of cases. ACOG suggests the distinction between gHTN and preeclampsia is "an exercise of nomenclature rather than a pragmatic one." 17