The Cosmos Collaborative: A Vendor-Facilitated Electronic Health Record Data Aggregation Platform

Mental health in persons with von Willebrand disease in the United States - a large national database study

BACKGROUND

There are very few large population-based studies studying mental health in persons with von Willebrand disease (PwVWD).

OBJECTIVES

We aim to assess prevalence of depression and anxiety in PwVWD over a period of 20 years and identify bleeding symptoms that may be more likely associated with depression and anxiety in PwVWD.

METHODS

This is a retrospective cohort study using a deidentified national dataset from 1118 hospitals with 176 million patients. Cases were defined as patients aged 0-110 years, both male and female, with von Willebrand disease (VWD), without hemophilia. Controls were defined as patients aged 0-110 years, both male and female, without VWD or hemophilia. We compared rates of depression and anxiety in cases and controls and by type of bleeding symptoms.

RESULTS

We identified 66 367 PwVWD and 183 890 766 controls. The prevalence of depression (23.12% vs 8.62%; p ≤ .00093; relative risk = 2.68) and anxiety (32.90% vs 12.29%; p ≤ .00093; relative risk = 2.68) was higher in PwVWD. Most of the bleeding symptoms were associated with higher rates of depression and anxiety in PwVWD with the highest rates with abnormal uterine bleeding, hematemesis, hemoptysis, hematuria, and melena.

CONCLUSION

Our study shows that mental health disorders in PwVWD are a significant health burden, and that burden is increased with documented bleeding symptoms. It is important that primary care physicians and hematologists caring for this population recognize this increased risk and appropriately screen and refer to mental health professionals.

Advancing diabetes surveillance ecosystems: a case study of India

Professional society and expert guidelines recommend the achievement of glycaemic, blood pressure, and cholesterol targets to prevent the microvascular and macrovascular complications of diabetes. The WHO Diabetes Compact recommends that countries meet and monitor these targets for diabetes management. Surveillance-ie, continuous, systematic measurement, analysis, and interpretation of data-is a crucial component of public health. In this Personal View, we use the case of India as an illustration of the challenges and future directions needed for a diabetes surveillance system that documents national progress and persistent gaps. To address the growing burdens of diabetes and cardiometabolic diseases, the Government of India has launched programmes such as the National Programme for Prevention and Control of Non-Communicable Diseases. Different surveys have provided estimates of the diabetes care continuum of awareness, treatment, and control at the national, state, and, very recently, district level. We reviewed the literature to analyse how these surveys have varied in both their data collection methods and the reported estimates of the diabetes care continuum. We propose an integrated surveillance and monitoring framework to augment decentralised decision making, leveraging the complementary strengths of different surveys and electronic health record databases, such as data obtained by the National Programme for Prevention and Control of Non-Communicable Diseases, and building on methodological advances in model-based small-area estimation and data fusion. Such a framework could aid state and district administrators in monitoring the progress of diabetes screening and management initiatives, and benchmarking against national and global standards in all countries.

Autoimmune, Autoinflammatory Disease and Cutaneous Malignancy Associations with Hidradenitis Suppurativa: A Cross-Sectional Study

BACKGROUND

Hidradenitis suppurativa (HS) is a debilitating cutaneous disease characterized by severe painful inflammatory nodules/abscesses. At present, data regarding the epidemiology and pathophysiology of this disease are limited.

OBJECTIVE

To define the prevalence and comorbidity associations of HS.

METHODS

This was a cross-sectional study of EPICTM Cosmos© examining over 180 million US patients. Prevalences were calculated by demographic and odds ratios (OR) and identified comorbidity correlations.

RESULTS

All examined metabolism-related, psychological, and autoimmune/autoinflammatory (AI) diseases correlated with HS. The strongest associations were with pyoderma gangrenosum [OR 26.56; confidence interval (CI): 24.98-28.23], Down syndrome (OR 11.31; CI 10.93-11.70), and polycystic ovarian syndrome (OR 11.24; CI 11.09-11.38). Novel AI associations were found between HS and lupus (OR 6.60; CI 6.26-6.94) and multiple sclerosis (MS; OR 2.38; CI 2.29-2.48). Cutaneous malignancies were largely not associated in the unsegmented cohort; however, among Black patients, novel associations with melanoma (OR 2.39; CI 1.86-3.08) and basal cell carcinoma (OR 2.69; CI 2.15-3.36) were identified.

LIMITATIONS

International Classification of Diseases (ICD)-based disease identification relies on coding fidelity and diagnostic accuracy.

CONCLUSION

This is the first study to identify correlations between HS with melanoma and basal cell carcinoma (BCC) among Black patients as well as MS and lupus in all patients with HS.

Quantifying effects of blood pressure control on neuroimaging utilization in a large multi-institutional healthcare population

OBJECTIVES

Essential hypertension is a common chronic condition that can exacerbate or complicate various neurological diseases that may necessitate neuroimaging. Given growing medical imaging costs and the need to understand relationships between population blood pressure control and neuroimaging utilization, we seek to quantify the relationship between maximum blood pressure recorded in a given year and same-year utilization of neuroimaging CT or MR in a large healthcare population.

METHODS

A retrospective population-based cohort study was performed by extracting aggregate data from a multi-institutional dataset of patient encounters from 2016, 2018, and 2020 using an informatics platform (Cosmos) consisting of de-duplicated data from over 140 academic and non-academic health systems, comprising over 137 million unique patients. A population-based sample of all patients with recorded blood pressures of at least 50 mmHg DBP or 90 mmHg SBP were included. Cohorts were identified based on maximum annual SBP and DBP meeting or exceeding pre-defined thresholds. For each cohort, we assessed neuroimaging CT and MR utilization, defined as the percentage of patients undergoing ≥1 neuroimaging exam of interest in the same calendar year.

RESULTS

The multi-institutional population consisted of >38 million patients for the most recent calendar year analyzed, with overall utilization of 3.8-5.1% for CT and 1.5-2.0% for MR across the study period. Neuroimaging utilization increased substantially with increasing annual maximum BP. Even a modest BP increase to 140 mmHg systolic or 90 mmHg diastolic is associated with 3-4-fold increases in MR and 5-7-fold increases in CT same-year imaging compared to BP values below 120 mmHg / 80 mmHg.

CONCLUSION

Higher annual maximum recorded blood pressure is associated with higher same-year neuroimaging CT and MR utilization rates. These observations are relevant to public health efforts on hypertension management to mitigate costs associated with growing imaging utilization.

Contemporary Prevalence of Oral Clefts in the US: Geographic and Socioeconomic Considerations

Background: Socio-economic status, living environments, and race have been implicated in the development of different congenital abnormalities. As orofacial clefting is the most common anomaly affecting the face, an understanding of its prevalence in the United States and its relationship with different determinants of health is paramount. Therefore, the purpose of this study is to determine the modern prevalence of oral-facial clefting in the United States and its association with different social determinants of health. Methods: Utilizing Epic Cosmos, data from approximately 180 US institutions were queried. Patients born between November 2012 and November 2022 were included. Eight orofacial clefting (OC) cohorts were identified. The Social Vulnerability Index (SVI) was used to assess social determinants of health. Results: Of the 15,697,366 patients identified, 31,216 were diagnosed with OC, resulting in a prevalence of 19.9 (95% CI: 19.7-20.1) per 10,000 live births. OC prevalence was highest among Asian (27.5 CI: 26.2-28.8) and Native American (32.8 CI: 30.4-35.2) patients and lowest among Black patients (12.96 CI: 12.5-13.4). Male and Hispanic patients exhibited higher OC prevalence than female and non-Hispanic patients. No significant differences were found among metropolitan (20.23/10,000), micropolitan (20.18/10,000), and rural populations (20.02/10,000). SVI data demonstrated that OC prevalence was positively associated with the percentage of the population below the poverty line and negatively associated with the proportion of minority language speakers. Conclusions: This study examined the largest US cohort of OC patients to date to define contemporary US prevalence, reporting a marginally higher rate than previous estimates. Multiple social determinants of health were found to be associated with OC prevalence, underscoring the importance of holistic prenatal care. These data may inform clinicians about screening and counseling of expectant families based on socio-economic factors and direct future research as it identifies potential risk factors and provides prevalence data, both of which are useful in addressing common questions related to screening and counseling.

Detection and Grading of Radiographic Hand Osteoarthritis Using an Automated Machine Learning Platform

OBJECTIVE

Automated machine learning (autoML) platforms allow health care professionals to play an active role in the development of machine learning (ML) algorithms according to scientific or clinical needs. The aim of this study was to develop and evaluate such a model for automated detection and grading of distal hand osteoarthritis (OA).

METHODS

A total of 13,690 hand radiographs from 2,863 patients within the Swiss Cohort of Quality Management (SCQM) and an external control data set of 346 non-SCQM patients were collected and scored for distal interphalangeal OA (DIP-OA) using the modified Kellgren/Lawrence (K/L) score. Giotto (Learn to Forecast [L2F]) was used as an autoML platform for training two convolutional neural networks for DIP joint extraction and subsequent classification according to the K/L scores. A total of 48,892 DIP joints were extracted and then used to train the classification model. Heatmaps were generated independently of the platform. User experience of a web application as a provisional user interface was investigated by rheumatologists and radiologists.

RESULTS

The sensitivity and specificity of this model for detecting DIP-OA were 79% and 86%, respectively. The accuracy for grading the correct K/L score was 75%, with a κ score of 0.76. The accuracy per DIP-OA class differed, with 86% for no OA (defined as K/L scores 0 and 1), 71% for a K/L score of 2, 46% for a K/L score of 3, and 67% for a K/L score of 4. Similar values were obtained in an independent external test set. Qualitative and quantitative user experience testing of the web application revealed a moderate to high demand for automated DIP-OA scoring among rheumatologists. Conversely, radiologists expressed a low demand, except for the use of heatmaps.

CONCLUSION

AutoML platforms are an opportunity to develop clinical end-to-end ML algorithms. Here, automated radiographic DIP-OA detection is both feasible and usable, whereas grading among individual K/L scores (eg, for clinical trials) remains challenging.

Socioeconomic Disparities in Glaucoma Severity at Initial Diagnosis: A Nationwide Electronic Health Record Cohort Analysis

PURPOSE

To assess disparities in initial disease severity among open-angle glaucoma (OAG) patients.

DESIGN

Cross-sectional study.

METHODS

In this analysis of Epic Cosmos, an aggregated electronic health record dataset encompassing >213 million patients, OAG patients examined in ophthalmology or optometry clinics between January 1, 2013, and June 1, 2023, were evaluated. OAG severity at presentation was classified as mild, moderate, or severe using International Classification of Disease-10 codes. Demographics, social vulnerability index (SVI) scores, and rural-urban commuting area codes were evaluated as predictors of disease stage using ordinal logistic regression. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated.

RESULTS

Of 245,669 patients, 38.1% had mild, 32.5% moderate, and 29.3% severe disease at presentation. In multivariable analyses, significant determinants of worse severity included older age (OR: 1.23 per decade, 95% CI: 1.22-1.23), male sex (OR: 1.37, 95% CI: 1.35-1.39), Black race (OR: 1.61, 95% CI: 1.58-1.65), Hispanic ethnicity (OR: 1.15, 95% CI: 1.11-1.18), non-commercial insurance or uninsured status (OR: 2.53, 95% CI: 2.33-2.74), secondary OAGs (eg, pseudoexfoliative glaucoma - OR: 1.65, 95% CI: 1.58-1.72), and higher socioeconomic SVI scores (OR: 1.25 for highest versus lowest quartile, 95% CI: 1.22-1.28). Black and Hispanic patients were diagnosed at younger ages compared to White patients (mean ages: 67.8 ± 12.3 and 68.1 ± 12.8 vs 73.3 ± 11.8 years respectively, P < .001).

CONCLUSIONS

Worse OAG at presentation was associated with older age, male sex, Black race, Hispanic ethnicity, non-commercial insurance or uninsured status, secondary OAGs, and greater socioeconomic vulnerability in this nationwide cohort. These findings can help tailor screening programs towards vulnerable populations.

How Are Leading Research Institutions Engaging with Data Sharing Tools and Programs?

With widespread electronic health record (EHR) adoption and improvements in health information interoperability in the United States, troves of data are available for knowledge discovery. Several data sharing programs and tools have been developed to support research activities, including efforts funded by the National Institutes of Health (NIH), EHR vendors, and other public- and private-sector entities. We surveyed 65 leading research institutions (77% response rate) about their use of and value derived from ten programs/tools, including NIH's Accrual to Clinical Trials, Epic Corporation's Cosmos, and the Observational Health Data Sciences and Informatics consortium. Most institutions participated in multiple programs/tools but reported relatively low usage (even when they participated, they frequently indicated that fewer than one individual/month benefitted from the platform to support research activities). Our findings suggest that investments in research data sharing have not yet achieved desired results.

Design, implementation, and inferential issues associated with clinical trials that rely on data in electronic medical records: a narrative review

Real world evidence is now accepted by authorities charged with assessing the benefits and harms of new therapies. Clinical trials based on real world evidence are much less expensive than randomized clinical trials that do not rely on "real world evidence" such as contained in electronic health records (EHR). Consequently, we can expect an increase in the number of reports of these types of trials, which we identify here as 'EHR-sourced trials.' 'In this selected literature review, we discuss the various designs and the ethical issues they raise. EHR-sourced trials have the potential to improve/increase common data elements and other aspects of the EHR and related systems. Caution is advised, however, in drawing causal inferences about the relationships among EHR variables. Nevertheless, we anticipate that EHR-CTs will play a central role in answering research and regulatory questions.

How Electronic Medical Record Integration Can Support More Efficient Critical Care Clinical Trials

Large volumes of data are collected on critically ill patients, and using data science to extract information from the electronic medical record (EMR) and to inform the design of clinical trials represents a new opportunity in critical care research. Using improved methods of phenotyping critical illnesses, subject identification and enrollment, and targeted treatment group assignment alongside newer trial designs such as adaptive platform trials can increase efficiency while lowering costs. Some tools such as the EMR to automate data collection are already in use. Refinement of data science approaches in critical illness research will allow for better clinical trials and, ultimately, improved patient outcomes.

Big Data in Stroke: How to Use Big Data to Make the Next Management Decision

The last decade has seen significant advances in the accumulation of medical data, the computational techniques to analyze that data, and corresponding improvements in management. Interventions such as thrombolytics and mechanical thrombectomy improve patient outcomes after stroke in selected patients; however, significant gaps remain in our ability to select patients, predict complications, and understand outcomes. Big data and the computational methods needed to analyze it can address these gaps. For example, automated analysis of neuroimaging to estimate the volume of brain tissue that is ischemic and salvageable can help triage patients for acute interventions. Data-intensive computational techniques can perform complex risk calculations that are too cumbersome to be completed by humans, resulting in more accurate and timely prediction of which patients require increased vigilance for adverse events such as treatment complications. To handle the accumulation of complex medical data, a variety of advanced computational techniques referred to as machine learning and artificial intelligence now routinely complement traditional statistical inference. In this narrative review, we explore data-intensive techniques in stroke research, how it has informed the management of stroke patients, and how current work could shape clinical practice in the future.

Data-driven longitudinal characterization of neonatal health and morbidity

Although prematurity is the single largest cause of death in children under 5 years of age, the current definition of prematurity, based on gestational age, lacks the precision needed for guiding care decisions. Here, we propose a longitudinal risk assessment for adverse neonatal outcomes in newborns based on a deep learning model that uses electronic health records (EHRs) to predict a wide range of outcomes over a period starting shortly before conception and ending months after birth. By linking the EHRs of the Lucile Packard Children's Hospital and the Stanford Healthcare Adult Hospital, we developed a cohort of 22,104 mother-newborn dyads delivered between 2014 and 2018. Maternal and newborn EHRs were extracted and used to train a multi-input multitask deep learning model, featuring a long short-term memory neural network, to predict 24 different neonatal outcomes. An additional cohort of 10,250 mother-newborn dyads delivered at the same Stanford Hospitals from 2019 to September 2020 was used to validate the model. Areas under the receiver operating characteristic curve at delivery exceeded 0.9 for 10 of the 24 neonatal outcomes considered and were between 0.8 and 0.9 for 7 additional outcomes. Moreover, comprehensive association analysis identified multiple known associations between various maternal and neonatal features and specific neonatal outcomes. This study used linked EHRs from more than 30,000 mother-newborn dyads and would serve as a resource for the investigation and prediction of neonatal outcomes. An interactive website is available for independent investigators to leverage this unique dataset: https://maternal-child-health-associations.shinyapps.io/shiny_app/.

Technology Platforms and Approaches for Building and Evaluating Machine Learning Methods in Healthcare

BACKGROUND

Artificial intelligence (AI) methods are becoming increasingly commonly implemented in healthcare as decision support, business intelligence tools, or, in some cases, Food and Drug Administration-approved clinical decision-makers. Advanced lab-based diagnostic tools are increasingly becoming AI driven. The path from data to machine learning methods is an active area for research and quality improvement, and there are few established best practices. With data being generated at an unprecedented rate, there is a need for processes that enable data science investigation that protect patient privacy and minimize other business risks. New approaches for data sharing are being utilized that lower these risks.

CONTENT

In this short review, clinical and translational AI governance is introduced along with approaches for securely building, sharing, and validating accurate and fair models. This is a constantly evolving field, and there is much interest in collecting data using standards, sharing data, building new models, evaluating models, sharing models, and, of course, implementing models into practice.

SUMMARY

AI is an active area of research and development broadly for healthcare and laboratory testing. Robust data governance and machine learning methodological governance are required. New approaches for data sharing are enabling the development of models and their evaluation. Evaluation of methods is difficult, particularly when the evaluation is performed by the team developing the method, and should ideally be prospective. New technologies have enabled standardization of platforms for moving analytics and data science methods.

Racial and Ethnic Differences in Rates and Age of Diagnosis of Autism Spectrum Disorder

This cohort study uses electronic health record data to assess racial and ethnic disparities in prevalence or median age of diagnosis of autism spectrum disorder in children.

Phenotypic prevalence of obesity and metabolic syndrome among an underdiagnosed and underscreened population of over 50 million children and adults

Background: Metabolic syndrome is a phenotypic condition associated with a variety of genotypes. Studies of rare genotypes can be made more difficult by clinical underscreening of the population for the phenotypic traits that define metabolic syndrome to clinicians. Studies have demonstrated underdiagnosis of pediatric obesity, as well as reduced rates of pediatric screening for obesity related conditions, including conditions leading to a diagnosis of metabolic syndrome. If true, there may be a significant underdiagnosis of metabolic syndrome among the pediatric population compared to the adult population.

Methods: Using Epic’s Cosmos Data Network aggregated, de-identified patient data collected from healthcare organizations using the Epic electronic health record (EHR), we examined obesity and metabolic syndrome rates among adult and pediatric patients. We also examined screening rates for obesity related conditions and metabolic syndrome among adult and pediatric patients across the United States. We also sought to compare rates between subgroups within the population including age, sex, and race.

Results: In our population, 45% of adults and 27% of pediatric population were obese by age and gender specific BMI criteria. 38% of the obese adult population had an ICD-10 code associated with the diagnosis vs. 52% of the pediatric population. Of adults meeting obesity criteria, 36% had results for appropriate, guideline-based blood laboratory testing for insulin resistance, 40–42% for dyslipidemia, and 55% for hepatic steatosis. 36% of obese adult patients had none of the recommended blood laboratory testing. 31% of the adult population met diagnostic criteria for metabolic syndrome. Of pediatric patients meeting obesity criteria, 27% had results for appropriate blood laboratory testing for insulin resistance, 28% for dyslipidemia, and 33% for hepatic steatosis. 59% of obese pediatric patients had none of the recommended blood laboratory testing. 3% of the pediatric population met criteria for diagnosis of metabolic syndrome.

Discussion: This study represents one of the largest multicenter national cohorts assembled for studying metabolic syndrome (over 50 million patients) and demonstrates the power of emerging aggregated EHR tools for research. Although obesity is better diagnosed in pediatric patients than in adult patients, significantly lower screening rates for obesity related conditions occurred in pediatric patients compared to adults. Statistically significant, but clinically negligible differences in screening rates were found by race and gender. These results support smaller prior studies that suggest that obesity is under-diagnosed and obesity related conditions underscreened in pediatric and adult populations, and additionally suggests underdiagnosis of metabolic syndrome among United States pediatric and adult patients.

Effectiveness of COVID-19 mRNA Vaccination in Preventing COVID-19-Associated Hospitalization Among Adults with Previous SARS-CoV-2 Infection - United States, June 2021-February 2022

Previous infection with SARS-CoV-2, the virus that causes COVID-19, has been estimated to confer up to 90% protection against reinfection, although this protection was lower against the Omicron variant compared with that against other SARS-CoV-2 variants (1-3). A test-negative design was used to estimate effectiveness of COVID-19 mRNA vaccines in preventing subsequent COVID-19-associated hospitalization among adults aged ≥18 years with a previous positive nucleic acid amplification test (NAAT) or diagnosis of COVID-19.^† The analysis used data from Cosmos, an electronic health record (EHR)-aggregated data set (4), and compared vaccination status of 3,761 case-patients (positive NAAT result associated with hospitalization) with 7,522 matched control-patients (negative NAAT result). After previous SARS-CoV-2 infection, estimated vaccine effectiveness (VE) against COVID-19-associated hospitalization was 47.5% (95% CI = 38.8%-54.9%) after 2 vaccine doses and 57.8% (95% CI = 32.1%-73.8%) after a booster dose during the Delta-predominant period (June 20-December 18, 2021), and 34.6% (95% CI = 25.5%-42.5%) after 2 doses and 67.6% (95% CI = 61.4%-72.8%) after a booster dose during the Omicron-predominant period (December 19, 2021-February 24, 2022). Vaccination provides protection against COVID-19-associated hospitalization among adults with previous SARS-CoV-2 infection, with the highest level of protection conferred by a booster dose. All eligible persons, including those with previous SARS-CoV-2 infection, should stay up to date with vaccination to prevent COVID-19-associated hospitalization.

Postpartum Length of Stay and Hospital Readmission Before and During the Coronavirus Disease 2019 (COVID-19) Pandemic

OBJECTIVE

To compare postpartum hospitalization length of stay (LOS) and hospital readmission among obstetric patients before (March 2017-February 2020; prepandemic) and during the coronavirus disease 2019 (COVID-19) pandemic (March 2020-February 2021).

METHODS

We conducted a retrospective cohort study, using Epic Systems' Cosmos research platform, of obstetric patients who delivered between March 1, 2017, and February 28, 2021, at 20-44 weeks of gestation and were discharged within 7 days of delivery. The primary outcome was short postpartum hospitalization LOS (less than two midnights for vaginal births and less than three midnights for cesarean births) and secondary outcome was hospital readmission within 6 weeks of postpartum hospitalization discharge. Analyses compared outcomes before and during the pandemic using standardized differences and Bayesian logistic mixed-effects models, among all births and stratified by mode of delivery.

RESULTS

Of the 994,268 obstetric patients in the study cohort, 742,113 (74.6%) delivered prepandemic and 252,155 (25.4%) delivered during the COVID-19 pandemic. During the COVID-19 pandemic, the percentage of short postpartum hospitalizations increased among all births (28.7-44.5%), vaginal births (25.4-39.5%), and cesarean births (35.3-55.1%), which was consistent with the adjusted analysis (all births: adjusted odds ratio [aOR] 2.35, 99% credible interval 2.32-2.39; vaginal births: aOR 2.14, 99% credible interval 2.11-2.18; cesarean births aOR 2.90, 99% credible interval 2.83-2.98). Although short postpartum hospitalizations were more common during the COVID-19 pandemic, there was no change in readmission in the unadjusted (1.4% vs 1.6%, standardized difference=0.009) or adjusted (aOR 1.02, 99% credible interval 0.97-1.08) analyses for all births or when stratified by mode of delivery.

CONCLUSION

Short postpartum hospitalization LOS was significantly more common during the COVID-19 pandemic for obstetric patients with no change in hospital readmissions within 6 weeks of postpartum hospitalization discharge. The COVID-19 pandemic created a natural experiment, suggesting shorter postpartum hospitalization may be reasonable for patients who are self-identified or health care professional-identified as appropriate for discharge.

Coronavirus Disease 2019 (COVID-19) Pandemic and Pregnancy Outcomes in a U.S. Population

OBJECTIVE

To examine whether the coronavirus disease 2019 (COVID-19) pandemic altered risk of adverse pregnancy-related outcomes and whether there were differences by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection status among pregnant women.

METHODS

In this retrospective cohort study using Epic's Cosmos research platform, women who delivered during the pandemic (March-December 2020) were compared with those who delivered prepandemic (matched months 2017-2019). Within the pandemic epoch, those who tested positive for SARS-CoV-2 infection were compared with those with negative test results or no SARS-CoV-2 diagnosis. Comparisons were performed using standardized differences, with a value greater than 0.1 indicating meaningful differences between groups.

RESULTS

Among 838,489 women (225,225 who delivered during the pandemic), baseline characteristics were similar between epochs. There were no significant differences in adverse pregnancy outcomes between epochs (standardized difference<0.10). In the pandemic epoch, 108,067 (48.0%) women had SARS-CoV-2 testing available; of those, 7,432 (6.9%) had positive test results. Compared with women classified as negative for SARS-CoV-2 infection, those who tested positive for SARS-CoV-2 infection were less likely to be non-Hispanic White or Asian or to reside in the Midwest and more likely to be Hispanic, have public insurance, be obese, and reside in the South or in high social vulnerability ZIP codes. There were no significant differences in the frequency of preterm birth (8.5% vs 7.6%, standardized difference=0.032), stillbirth (0.4% vs 0.4%, standardized difference=-0.002), small for gestational age (6.4% vs 6.5%, standardized difference=-0.002), large for gestational age (7.7% vs 7.7%, standardized difference=-0.001), hypertensive disorders of pregnancy (16.3% vs 15.8%, standardized difference=0.014), placental abruption (0.5% vs 0.4%, standardized difference=0.007), cesarean birth (31.2% vs 29.4%, standardized difference=0.039), or postpartum hemorrhage (3.4% vs 3.1%, standardized difference=0.019) between those who tested positive for SARS-CoV-2 infection and those classified as testing negative.

CONCLUSION

In a geographically diverse U.S. cohort, the frequency of adverse pregnancy-related outcomes did not differ between those delivering before compared with during the pandemic, nor between those classified as positive compared with negative for SARS-CoV-2 infection during pregnancy.