Racial/Ethnic Differences in Pre-Pregnancy Conditions and Adverse Maternal Outcomes in the nuMoM2b Cohort ========================================================================================================= * Meghan E. Meredith * Lauren N. Steimle * Kaitlyn K. Stanhope * Marissa H. Platner * Sheree L. Boulet ## Abstract **Background** Adverse maternal outcomes and multimorbidity affect a growing proportion of pregnant individuals in the United States and with racial and ethnic disparities in the rates of multimorbidity. There is limited research on multimorbidity, measured as the co-occurrence of two or more chronic conditions, and how it impacts adverse maternal outcomes. **Objectives** We sought to: (1) determine how pre-pregnancy conditions and their combined effects contribute to racial disparities in adverse maternal outcomes; and (2) incorporate conditions and their combined effects into predictive modeling algorithms to improve risk prediction using information available in the first trimester, specifically for racial minority subgroups. **Study Design** We used data from the “Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-to-be (nuMoM2b)” observational cohort study. We defined multimorbidity as the co-occurrence of two or more pre-pregnancy conditions. The primary outcomes of interest were severe preeclampsia, postpartum readmission, and blood transfusion during pregnancy or up to 14 days postpartum. We used weighted Poisson regression with robust variance to estimate adjusted risk ratios and 95% confidence intervals, and we used mediation analysis to evaluate the contribution of the combined effects of pre-pregnancy condition to racial/ethnic disparities. We also evaluated the performance of our models by racial subgroup using the area under the receiver operating characteristic curve (AUC) metric. **Results** In the nuMoM2b cohort (n=8729), 3.6% (n=318) experienced severe preeclampsia, 1.8% (n=157) experienced blood transfusion, and 1.8% (n=154) experienced postpartum readmission. 22.8% of all participants experienced two or more co-occurring pre-pregnancy conditions. Cardiovascular conditions were associated with an increased relative risk for severe preeclampsia (aRR, 1.71; CI, 1.59-1.84), and this risk was additionally exacerbated when hematologic conditions were co-occurring with cardiovascular conditions (aRR, 1.34; CI, 1.12-1.60). The mediation analysis results were not statistically significant; however, cardiovascular conditions explained 36.6% of non-Hispanic Black individuals’ association with increased risk for severe preeclampsia (p=0.07). The addition of pre-pregnancy conditions increased AUC for the prediction of blood transfusion and severe preeclampsia, and moreover, the multiracial subgroup experienced the greatest improvement in predictive performance of blood transfusion due to this addition (an increase in AUC 0.51 to 0.69). The best subgroup performance was in predicting severe preeclampsia in multiracial individuals using confounders and condition types, with an AUC of 0.74. **Conclusion** Though there is a greater burden of some pre-pregnancy conditions and co-occurring conditions among minority individuals, these conditions alone do not fully explain the higher rates of adverse maternal outcomes. However, some conditions and their combined effects are associated with adverse maternal outcomes. The incorporation of pre-pregnancy multimorbidity improves the performance of some risk prediction models, especially for correctly classifying individuals of minority race. Keywords * Maternal health * Pregnancy Complications * Maternal morbidity * Risk Assessment * Multimorbidity * Comorbidity * Chronic Conditions * Health Disparities ## Introduction The US maternal mortality rate, 23.8 per 100,000 live births, has worsened over the past 20 years as rates among peer nations have improved,1,2 and there are staggering racial and ethnic disparities in maternal mortality.3 Compared with non-Hispanic White individuals, non-Hispanic Black individuals have 3 to 4 times the risk of dying during childbirth.4 An estimated 80% of pregnancy related deaths are considered preventable,5 and individuals with adverse maternal outcomes have similar preventable factors including provider failure to identify high risk status and inappropriate management.6 Two potential targets for addressing pregnancy-related mortality and morbidity are better chronic condition management before and during pregnancy and improved prediction models for adverse outcomes which may disproportionally affect racial and ethnic minority individuals. Chronic conditions are associated with higher risk of adverse maternal outcomes, and deliveries among racial and ethnic minorities, compared with non-Hispanic white individuals, experience significantly higher prevalence of chronic conditions and adverse maternal morbidity.7 Pregnant individuals with co-occurring conditions, or *multimorbidity*, have a higher rate of experiencing severe maternal morbidity and postpartum readmission.8 While multimorbidity is increasing, apart from a few studies using administrative data, 9,10 and those creating comorbidity-based risk screening tools,11–13 obstetric research and practice remain largely focused on the impact of single conditions on maternal outcomes.14 When multimorbidity is accounted for, it is typically represented as a binary or count variable, while the co-occurrence of specific diagnoses may be more relevant for clinical management. In this study, we investigate the impact of co-occurring pre-pregnancy conditions and their combined effects on adverse maternal outcomes and whether these effects mediate the relationship between racial disparities and adverse maternal outcomes in a cohort of nulliparous individuals. A better understanding of which conditions and combinations thereof drive increase maternal risk could inform the development of new clinical care standards for multimorbid pregnant people and potentially improve the ability to risk-stratify pregnant people early in pregnancy. ## Materials and Methods ### Study Population We conducted a secondary analysis of data from the nuMoM2b prospective cohort study of nulliparous women.15,16 The study recruited women from hospitals affiliated with eight clinical centers and collected data on each participant over the course of four study visits via in-clinic interviews, self-administered questionnaires, clinical measurements, and chart abstractions. The cohort included 9,289 participants who consented to the release of their data. ### Data Preparation For this secondary data analysis, we chose to exclude participants that delivered at gestational age < 22 weeks or > 43 weeks and participants with key data missing. To address the remaining missing data, we created 5 imputed datasets using multivariate imputation by chained equations (MICE).17 Please see Appendix A for more details about the exclusion protocol and Appendix B.1 for more details about the data structure and preparation. #### Primary Outcomes The primary outcomes of this study were severe preeclampsia, blood transfusion, and postpartum readmission (up to 14 days). These outcomes are strongly associated with maternal mortality and morbidity and have high quality measurement in the nuMoM2b dataset. Appendix B.2 provides details about the definitions and collection of the primary outcomes in the nuMoM2b study. We fit prediction models to each of these outcomes individually to preserve interpretability of model results and compared results across each outcome. #### Pre-Pregnancy Conditions Pre-pregnancy conditions were collected by the nuMoM2b study team at each of the first three visits through an in-person interview with the participant, who indicated the presence of each condition anytime throughout their life. The nuMoM2b study team reconciled this data with chart abstraction. The nuMoM2b study team collected 41 pre-pregnancy conditions, and we categorized them into 12 condition types according to the framework of Tang et al.18 (Appendix Table B.2.1). We grouped the pre-pregnancy conditions based on similarities in treatments, clinical manifestation, or organization in the health care system. These condition types included: Autoimmune, Cardiovascular, Endocrine, Gastrointestinal, Gynecological, Hematologic, Kidney, Lung, Mental, and Neurological. This classification was developed and validated by three of our authors (MP, SB, KS) with extensive medical and specifically obstetric knowledge. #### Confounders We adjusted for confounding factors including maternal age, insurance, body mass index (BMI), and sociodemographic information such as income and education. Each of these factors were collected during Visit 1, which occurred between 6 and 14 weeks. ### Statistical Analysis We used Poisson regression models with robust standard errors to analyze the associations between pre-pregnancy condition types, self-reported race/ethnicity, and adverse maternal outcomes with the results reported as crude and adjusted risk ratios (RRs) and 95% confidence intervals (CIs). Poisson regression provides risk ratio estimates which aid clinical interpretation, and therefore was chosen instead of logistic regression which provides odds ratios. We used a class-weighting method which gives a higher priority to correctly classifying the subgroup of participants who experienced an adverse maternal outcome (see Appendix B.3). First, each adverse maternal outcome was regressed individually by race/ethnicity, confounders, and potential mediators (all pre-pregnancy condition types and all their combined effects) (hereafter referred to as the “Outcome Model”), and only combined effects and their components reaching statistical significance (p-value ≤ 0.05) in this model were selected for further analysis. We used a model-building approach starting with a standard individual-level Poisson model (Model 1, crude), followed by a multilevel Poisson model that adjusted for confounders (Model 2, adjusted). Next, we additionally controlled for significant pre-pregnancy condition types (Model 3), and finally, we additionally controlled for significant combined effects, or interaction terms, between co-occurring pre-pregnancy condition types (Model 4). Next, we sought to determine if pre-pregnancy condition types and their combined effects contributed to the racial disparities in adverse maternal outcomes using mediation analysis. To analyze mediation, the potential mediator condition types and their combined effects were individually regressed by race/ethnicity and adjusted for confounders (“Mediator Model”). The Outcome and Mediator models were combined to compute the mediation proportion, which estimates the proportion of the risk factor’s impact (race/ethnicity) on the outcome that is attributable to the mediator (pre-pregnancy condition type or combined effect).19 This analysis was conducted for each adverse maternal outcome and results were averaged across the five imputed datasets. ### Performance Analysis Finally, we compared the performance of an intercept (constant) model and our prediction models (Model 1, Model 2, Model 3, and Model 4) to understand which pre-pregnancy patient characteristics accurately classified adverse maternal outcomes. Performance of each model was evaluated for each racial/ethnic subgroup using the average area under the ROC curve (AUC) across 10-fold cross validation for the five imputed datasets. We used R, version 4.2.1, for all analysis, *caret* in R to train and test our prediction models and *mediate* in R to conduct mediation analysis. We use p-value threshold of 0.05 for statistical significance. This study was approved by the Georgia Institute of Technology Internal Review Board who deemed it exempt from review. ## Results After exclusions, the final study population included 8729 participants (Table 1). Of those, 61.0% (n=5322) identified as non-Hispanic white, 13.1% (n = 1145) non-Hispanic Black, 17.0% (n = 1487) Hispanic, 3.9% (n = 342) Asian, 4.0% (n = 348) multiracial, and 1.0% (n=85) “Other”. In total, this population included 157 cases (1.8%) of blood transfusion, 154 cases (1.8%) of postpartum readmission, and 318 cases (3.6%) of severe preeclampsia. The incidence of these adverse maternal outcomes varied by race and ethnicity (Table Appendix B.2.2). The incidence of blood transfusion and severe preeclampsia was highest in non-Hispanic Black participants and the incidence of all adverse maternal outcomes was lowest in Asian participants. View this table: [Table 1:](http://medrxiv.org/content/early/2022/11/02/2022.11.02.22281812/T1) Table 1: Descriptive Statistics of the nuMoM2b Study Population by race/ethnicity status (before imputing missing data). BMI, body mass index; Std, standard deviation; HS, high school The most common pre-pregnancy conditions were mental health conditions (14.4%, n=1261), hematologic conditions (13.8%, n=1206), neurological conditions (13.0%, n=1136), and lung conditions (12.5%, n=1091) (Table Appendix B.3.2). Non-Hispanic Black participants experienced the highest rates of lung and hematologic conditions, Hispanic participants experienced the highest rates of kidney conditions, and multiracial participants experienced the highest rates of cardiovascular conditions. Non-Hispanic White participants experienced the highest rates of neurological, gastrointestinal, and mental health conditions. Compared with non-Hispanic White participants, the unadjusted risk of most adverse maternal outcomes was higher in non-Hispanic Black participants, Hispanic participants, multiracial, and participants with “Other” race, and was lower in Asian participants. Statistical adjustment for age, sociodemographic information, insurance, and BMI attenuated but did not eliminate the elevated risk of adverse maternal outcomes in minority racial/ethnic groups. The adjusted relative risk (aRR) of blood transfusion was higher in non-Hispanic Black (aRR, 1.20; CI, 1.10-1.32), and Hispanic participants (aRR, 1.17; CI, 1.08-1.27) compared to non-Hispanic White participants (Table 2). The adjusted relative risk of postpartum readmission was higher in participants with “Other” race (aRR, 1.36; CI, 1.08-1.68) (Table 3) and the adjusted relative risk of severe preeclampsia was higher in non-Hispanic Black participants (aRR, 1.22; CI, 1.12-1.33) compared to non-Hispanic White participants (Table 4). View this table: [Table 2:](http://medrxiv.org/content/early/2022/11/02/2022.11.02.22281812/T2) Table 2: Multilevel Poisson regression with outcome blood transfusion. Bolding indicates statistical significance at 95% confidence. RR, risk ratio; CI, confidence interval; AUC, area under the receive operating characteristics curve View this table: [Table 3:](http://medrxiv.org/content/early/2022/11/02/2022.11.02.22281812/T3) Table 3: Multilevel Poisson regression with outcome postpartum readmission. Bolding indicates statistical significance at 95% confidence. RR, risk ratio; CI, confidence interval; AUC, area under the receive operating characteristics curve View this table: [Table 4:](http://medrxiv.org/content/early/2022/11/02/2022.11.02.22281812/T4) Table 4: Multilevel Poisson regression with outcome severe preeclampsia, bolding indicates statistical significance at 95% confidence RR, risk ratio; CI, confidence interval; AUC, area under the receive operating characteristics curve After adjusting for confounders, we controlled for pre-pregnancy condition types (Model 3) and their combined effects (Model 4). In these models, the significance of the risk ratios for each race/ethnicity did not change, however the adjusted risk ratios did shift. Therefore, controlling for pre-pregnancy condition types and their combined effects did not fully eliminate the elevated risk of adverse maternal outcomes in minority racial/ethnic groups. The increased risk of severe preeclampsia among non-Hispanic Black participants was partially attenuated by the pre-pregnancy condition types, where their risk ratio decreased from 1.22 (1.12-1.33) to 1.14 (1.04-1.24) (Table). Pre-pregnancy condition types and their combined effects were associated with adverse maternal outcomes (see Tables 2-4). Cardiovascular conditions significantly increased risk for all adverse outcomes. Hematologic, endocrine, and autoimmune conditions significantly increased risk for blood transfusion and postpartum readmission. Gynecological conditions significantly decreased risk for blood transfusion and severe preeclampsia. In models that included pre-pregnancy condition type combined effects, we found that neurological conditions increased relative risk for postpartum readmission (aRR, 1.18; CI, 1.06-1.32), and this risk was exacerbated when kidney conditions were also present (aRR, 2.12; CI, 1.36-3.37) (Table 3). This means that, compared to those with neither or only one condition, pregnant individuals with both kidney and neurological conditions are 2.12 times more likely to experience postpartum readmission. We found that cardiovascular conditions increased the relative risk for severe preeclampsia (aRR, 1.71; CI, 1.59-1.84), and this risk was exacerbated when hematologic conditions were also present (aRR, 1.34; CI, 1.12-1.60). While gastrointestinal and gynecological conditions individually decreased risk for severe preeclampsia (aRR, 0.48; CI, 0.35-0.66; aRR, 0.82; CI, 0.73-0.91; respectively), the co-occurrence of these two conditions was associated with increased risk (RR, 2.59; CI, 1.64-4.11) (Table 4). All other analyzed pre-pregnancy condition type combined effects did not impact or significantly decrease risk for adverse maternal outcomes. Mediation analysis was conducted on all statistically significant condition types and combined effects for each adverse maternal outcome. There were no significant results at p-value < 0.05; however, cardiovascular conditions accounted for 36.6% of the association between non-Hispanic Black race/ethnicity and severe preeclampsia at p-value 0.07. Finally, we compared the performance metrics, specifically the area under the receiver operating characteristics curve (AUC), of models with different feature sets to determine the value of different maternal characteristics in predicting the risk of experiencing each of the adverse maternal outcomes (Table 5). We found gains in overall AUC with the addition of condition types for the prediction of blood transfusion and severe preeclampsia (AUC, 0.54 to 0.58, 0.61 to 0.66, respectively). We computed AUC for each racial and ethnic subgroup to understand how the models performed across the subgroups. AUC varies significantly by race for each adverse maternal outcome. Model 3, which includes race, confounders, and condition types as features, has an AUC range of 0.46-0.64 for blood transfusion, 0.46-0.62 for postpartum readmission, and 0.46-0.74 for severe preeclampsia. The best subgroup performance was in predicting severe preeclampsia in multiracial individuals using confounders and condition types, with an AUC of 0.74. The largest improvement in AUC from adding condition types and combined effects was in predicting blood transfusion in multiracial individuals, from 0.51 to 0.69. View this table: [Table 5:](http://medrxiv.org/content/early/2022/11/02/2022.11.02.22281812/T5) Table 5: AUC by race for each model and adverse maternal outcome. AUC, area under the receive operating characteristics curve ## Discussion ### Principal Findings Although pre-pregnancy condition types and combined effects were associated with adverse maternal outcomes, they did not explain racial and ethnic disparities in adverse maternal outcomes. However, we found that pre-pregnancy conditions improved diagnostic ability (AUC) in predictive models for blood transfusion and severe preeclampsia, both for the overall cohort and for individual racial and ethnic groups. These results highlight the potential for risk prediction using pre-pregnancy conditions in a diverse, low-risk population. ## Results Consistent with previous studies,4,7,20–22 we found that identification as a minority race was associated with a higher risk of adverse maternal outcomes among a cohort of nulliparous individuals. Adding to some studies that have examined the impact of specific co-occurring condition combined effects,23,24 our study focused on exploring all potential pre-pregnancy condition type combined effects to understand their association with adverse maternal outcomes (specifically, severe preeclampsia, postpartum readmission, and blood transfusion) and quantified the value of including combined effects in predictive models. We found that the presence of any gynecological condition was associated with decreased risk of experiencing blood transfusion and severe preeclampsia, and the presence of hematologic and gastrointestinal conditions individually were associated with decreased risk of experiencing severe preeclampsia. We further found that the presence of kidney conditions alongside a neurological condition significantly exacerbated risk for postpartum readmission, and the presence of a hematologic condition alongside a cardiovascular condition exacerbated risk for severe preeclampsia. Our work adds to previous studies on the use of prediction models to predict various adverse outcomes, which have ranged from using symptoms and signs,11,13 laboratory tests and biomarkers,25,26 and demographics and medical history.27,28 With the inclusion of pre-pregnancy condition types, our model yields an AUC of 0.66 for predicting severe preeclampsia, which is comparable to the AUC of previously published obstetric comorbidity indices predicting SMM, which included severe preeclampsia (Bateman, et al.12: AUC, 0.65 and Easter, et al.11 : AUC, 0.70 as reported by Leonard et al.29). The prediction models for blood transfusion and severe preeclampsia improved with the addition of pre-pregnancy condition types and combined effects; however, the model for postpartum readmission did not improve. ### Clinical Implications Understanding the potential reasons for adverse maternal outcomes is an important pathway to understanding and reducing racial and ethnic disparities and high rates of maternal morbidity and mortality. We found that many pre-pregnancy condition types and combined effects were associated with a decreased risk for adverse maternal outcomes. For example, in predicting blood transfusion, having both cardiovascular and gynecological conditions as well as having both hematologic and gastrointestinal conditions decreased risk. A possible explanation for this association is that pregnant individuals with one or multiple pre-pregnancy condition types are placed into a high-risk category in which they receive more frequent and higher quality care, such as maternal or fetal monitoring, frequent laboratory assessments, and referrals to subspecialists. This care may effectively lower their risk compared to pregnant individuals without pre-pregnancy conditions. Future studies may wish to control for the frequency of prenatal care, but this was not possible in our study due to the nature of the data collection process in nuMoM2b. In predicting severe preeclampsia, we found that having both cardiovascular and hematologic conditions as well as having both gastrointestinal and gynecological conditions increased risk. Thus, the presence of these co-occurring conditions may warrant more intensive clinical management. Further discussion of pre-pregnancy condition type combined effect results and their clinical implications can be found in Appendix C.1. ### Research Implications While the presence of pre-pregnancy conditions alone does not explain the observed racial and ethnic disparities in adverse maternal outcomes, there are likely other factors that do explain the observed disparities (condition management, access to care, social determinants of health, structural racism, etc.). The management of these pre-existing conditions before, during, and between pregnancies could be an important consideration. Thus, future work may examine the extent to which multimorbidity is managed during the preconception, interconception, and postpartum periods and whether differential management explains the observed disparities. Our work suggests that multimorbidity and combined effects among pre-pregnancy conditions can be a useful in improving the ability to risk-stratify individuals. The use of predictive modeling may be useful for further exploring these complex relationships. However, more research is necessary to inform best clinical practice for use of predictive models with a focus on mitigating unintended consequences and preventing the exacerbation of disparities.30 ### Strengths and Limitations Our study has several strengths. Using a large and comprehensive dataset, we evaluated the association between race and ethnicity, pre-pregnancy conditions, and adverse maternal outcomes. This dataset contains thoroughly collected data that goes beyond a typical electronic health record including health history and conditions, demographics, and survey questionnaires. Our condition type groupings allowed for larger sample sizes of conditions and more accurate estimates of risk ratios. In addition, the use of feature selection algorithms allowed for the exploration of combined effects to improve model performance for predicting adverse maternal outcomes. Our study has several important limitations. Our analysis was limited to the data collected in the nuMoM2b dataset, which only includes nulliparous, predominantly non-Hispanic white women who received care at academic medical centers and began care before 13 weeks and 6 days gestation; thus, our findings may not be generalizable to different birthing populations. Also, our dataset only collected data up to 14 days postpartum, although many postpartum readmissions occur after this timespan.31 ## Conclusions In addition to describing associations between race and ethnicity, pre-pregnancy condition types and their combined effects, and adverse maternal outcomes, our findings indicate that the data collected at an initial prenatal care visit has utility for predicting the risk of experiencing an adverse maternal outcome. Though there is a greater burden of some pre-pregnancy conditions and co-occurring conditions on minority individuals, multimorbidity alone does not account for the higher rates of adverse maternal outcomes. However, the incorporation of pre-pregnancy conditions improved the performance of risk prediction models. Our study findings have important implications for the preconception care as well as antepartum care of individuals with multimorbidity as adequately assessing a patient’s risk is essential to providing risk-appropriate and equitable care to prevent adverse maternal outcomes. ## Data Availability The data is not available to be shared per our data use agreement. ## Acknowledgements Research reported in this publication was supported in part by Imagine, Innovate and Impact (I3) from the Emory School of Medicine, Georgia Tech, and through the Georgia CTSA NIH award (UL1-TR002378) and by the National Science Foundation under grant number DGE-2039655 (Meredith); any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation. ## Appendix ### Appendix A. Exclusion Protocol For this secondary data analysis, we chose to exclude participants with pregnancy outcomes including fetal death before 20 weeks gestation, elective termination, indicated termination, unknown outcomes, and if the participant refused to release their pregnancy outcome. We also excluded participants that delivered at gestational age < 22 weeks or > 43 weeks. We excluded participants that had an incomplete labor, delivery, and postpartum form, dataset CMA in the nuMoM2b database, and those with an incomplete across visit medical conditions and medications form, dataset VXX in the nuMoM2b database.16 Finally, we excluded participants without self-reported race/ethnicity status and those without mode of delivery recorded (Figure A.1). ![Figure A.1:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2022/11/02/2022.11.02.22281812/F1.medium.gif) [Figure A.1:](http://medrxiv.org/content/early/2022/11/02/2022.11.02.22281812/F1) Figure A.1: Exclusion Criteria ### Appendix B: Methods #### B.1 Data Preparation The nuMoM2b study recruited individuals from hospitals affiliated with eight clinical centers and collected data on each participant over the course of four study visits. Visit 1 occurred between 6 and 14 weeks, Visit 2 occurred between 16 and 22 weeks, Visit 3 occurred between 22 and 30 weeks, and Visit 4 occurred at the time of delivery. In the nuMoM2b study, participants’ racial and ethnic status was self-reported during Visit 1, and participants were given the option to select more than one racial category. In our analysis, we used a single feature to represent race and ethnicity. We grouped maternal race and ethnicity as non-Hispanic White, non-Hispanic Black, Hispanic, Asian, multiracial, and other, where the “Other” category includes American Indian, Native Hawaiian, any other race/ethnicity status. No features were missing at more than 20%, so none were removed. We imputed the missing entries using multivariate imputation by chained equations (MICE),32 an algorithm that preserves the distribution of each variable. Specifically, we used predictive mean matching (PMM)32 because this method imputes only values that are observed and therefore preserves categorical variables. #### B.2 Primary Outcomes The outcome of “severe preeclampsia” in our study is indicated by of one of the following diagnoses from the nuMoM2b dataset: severe preeclampsia, superimposed preeclampsia, HELLP syndrome, or eclampsia. These diagnoses were identified through a detailed chart review performed by a site investigator or staff member certified for abstraction of complicated charts. If necessary, difficult cases were adjudicated by the principal investigators and final classification was reached by consensus. The outcome of “blood transfusion” is indicated by any amount of any blood product administered anytime during pregnancy or postpartum (up to 14 days) as recorded in a participant’s medical records. The occurrence of postpartum readmission is indicated by any readmission to the hospital within 14 days of delivery. A detailed description of the nuMoM2b study definitions of hypertensive orders is included as a supplement to the paper by Facco et al. (2017).33 View this table: [Table B.2.1:](http://medrxiv.org/content/early/2022/11/02/2022.11.02.22281812/T6) Table B.2.1: Adverse maternal outcomes and their corresponding indicators in the nuMoM2b dataset View this table: [Table B.2.2:](http://medrxiv.org/content/early/2022/11/02/2022.11.02.22281812/T7) Table B.2.2: Rates of adverse maternal outcomes in the nuMoM2b population overall and by race/ethnicity #### B.3 Study variables from the nuMoM2b dataset View this table: [Table B.3.1:](http://medrxiv.org/content/early/2022/11/02/2022.11.02.22281812/T8) Table B.3.1: Pre-pregnancy conditions and their corresponding indicators in the nuMoM2b dataset In our analysis, we excluded the condition types: Acute Infection and Cancer. Acute infection is recorded as occurring anytime during a participant’s life, which lacks relevance in our analysis because the nuMoM2b study did not collect the timing of the acute infection and because lifetime acute infections are common. Cancer was excluded due to its low prevalence. View this table: [Table B.3.2:](http://medrxiv.org/content/early/2022/11/02/2022.11.02.22281812/T9) Table B.3.2: Prevalence of pre-Pregnancy conditions by race in nuMoM2b dataset Number of condition types is the sum of different condition types, for example if a woman only has 2 cardiovascular conditions this is counted as 1 condition type #### B.3 Model development details The balancing class weight heuristic is inspired by King et al (2001) 34 and identical to scikit learn’s “balanced” class weight function. The weights are given by the formula: ![Formula][1] Where, ![Formula][2] #### B.4 Mediation Analysis and Correlation Firstly, each adverse maternal outcome was regressed individually by race/ethnicity, confounders, and potential mediators (all pre-pregnancy condition types and all their combined effects) (hereafter referred to as the “Outcome Model”), and only combined effects and their components reaching statistical significance (p-value ≤ 0.05) in this model were selected for further analysis. To analyze mediation, the selected potential mediator condition types and their combined effects were individually regressed by race/ethnicity and adjusted for confounders (“Mediator Model”). The Outcome and Mediator models were combined to compute average causal mediation effects (ACME) and total effect (TE) for each participant which was then averaged. Quasi-Bayesian estimation with 1,000 iterations were used for estimating the 95% CI and p-values of the natural indirect effect (NIE) and TE.19 Mediation proportion, which estimates the proportion of the risk factor’s impact on the outcome that is attributable to the mediator, was calculated as ACME / TE.19 This analysis was conducted for each adverse maternal outcome and results were averaged across the five imputed datasets. We assessed whether features were highly correlated and, a priori, planned to remove any that were highly correlated with another measure that was more relevant to our analysis to preserve the interpretability of the risk ratios. High correlation was indicated by a Pearson’s coefficient absolute value of 0.70 or greater. No features, other than confounding factors, were highly correlated. ### Appendix C: Discussion #### C.1 Further Discussion of Interaction Effects In predicting postpartum readmission, many pre-pregnancy condition type combined effects were associated with a decreased risk. A limitation of the nuMoM2b dataset is that postpartum readmission is only collected for 14 days postpartum. Some condition type combined effects may increase risk so that participants may be more likely to remain hospitalized following delivery and are less likely to be readmitted by 14 days postpartum. We found that neurological conditions increased risk, and their interaction with kidney conditions increased risk even further. Migraines have been shown to be associated with chronic kidney disease,35 and the interaction effect may be due to a medication interaction, where individuals with migraines and kidney issues may have limited medication options available for pain control which may exacerbate adverse maternal outcomes.36 In predicting severe preeclampsia, we found that the interaction between cardiovascular and hematologic conditions and the interaction between gastrointestinal and gynecological conditions increased risk. Individually, cardiovascular conditions are highly associated with severe preeclampsia outcomes; however, hematologic conditions are associated with a decreased risk for severe preeclampsia. The current diagnostic criteria for preeclampsia includes the presence of hypertension and organ dysfunction – including hematological dysfunction,37 which is consistent with this interaction. These results indicate that pregnant individuals with only hematological conditions should not be considered high-risk for preeclampsia, and pregnant individuals with both cardiovascular and hematological conditions should be considered higher risk than those with only cardiovascular conditions. The interaction between gastrointestinal and gynecological conditions has a less straightforward relationship with severe preeclampsia. Gastrointestinal conditions, such as irritable bowel disease, Crohn’s disease, and liver disease, are some of the least studied topics in the field of obstetrics.38 However, renal or liver dysfunction can be a diagnostic criteria of preeclampsia.37 Gynecological conditions, such as fibroids and PCOS, have been shown to be associated with a greater risk for adverse pregnancy outcomes, including preeclampsia.39–42 While the gastrointestinal and gynecological systems are physically close, there is a lack of literature on how they interact. Our results may indicate that individually, gynecological conditions do not increase risk for preeclampsia, however their interaction with gastrointestinal conditions increases risk regardless of obesity. * Received November 2, 2022. * Revision received November 2, 2022. * Accepted November 2, 2022. * © 2022, Posted by Cold Spring Harbor Laboratory The copyright holder for this pre-print is the author. All rights reserved. The material may not be redistributed, re-used or adapted without the author's permission. ## References 1. 1.Pregnancy Mortality Surveillance System | Maternal and Infant Health | CDC. Accessed May 18, 2022. [https://www.cdc.gov/reproductivehealth/maternal-mortality/pregnancy-mortality-surveillance-system.htm](https://www.cdc.gov/reproductivehealth/maternal-mortality/pregnancy-mortality-surveillance-system.htm) 2. 2.Hoyert L. D. Maternal Mortality Rates in the United States, 2019.; 2021. doi:10.15620/CDC:103855 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.15620/CDC:103855&link_type=DOI) 3. 3.Petersen EE, Davis NL, Goodman D, et al. Racial/Ethnic Disparities in Pregnancy-Related Deaths — United States, 2007–2016. MMWR Morb Mortal Wkly Rep. 2019;68(35):762–765. doi:10.15585/MMWR.MM6835A3 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.15585/mmwr.mm6835a3&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=31487273&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 4. 4.Liese KL, Mogos M, Abboud S, Decocker K, Koch AR, Geller SE. Racial and Ethnic Disparities in Severe Maternal Morbidity in the United States. J Racial Ethn Health Disparities. 2019;6(4):790–798. doi:10.1007/S40615-019-00577-W [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/S40615-019-00577-W&link_type=DOI) 5. 5.Trost S, Beauregard J, Chandra G, et al. Pregnancy-Related Deaths: Data from Maternal Mortality Review Committees in 36 US States, 2017-2019.; 2017. 6. 6.Geller SE, Rosenberg D, Cox S, Brown M, Simonson L, Kilpatrick S. A scoring system identified near-miss maternal morbidity during pregnancy. J Clin Epidemiol. 2004;57(7):716–720. doi:10.1016/J.JCLINEPI.2004.01.003 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.jclinepi.2004.01.003&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=15358399&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 7. 7.Admon LK, Winkelman TNA, Zivin K, Terplan M, Mhyre JM, Dalton VK. Racial and ethnic disparities in the incidence of severe maternal morbidity in the United States, 2012-2015. Obstetrics and Gynecology. 2018;132(5):1158–1166. doi:10.1097/AOG.0000000000002937 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/AOG.0000000000002937&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 8. 8.Stanhope KK, Worrell N, Jamieson DJ, Geary FH, Boulet SL. Double, Triple, and Quadruple Jeopardy: Entering Pregnancy With Two or More Multimorbid Diagnoses and Increased Risk of Severe Maternal Morbidity and Postpartum Readmission. Women’s Health Issues. Published online July 11, 2022. doi:10.1016/J.WHI.2022.06.005 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/J.WHI.2022.06.005&link_type=DOI) 9. 9.Admon L, Winkelman T, Moniz M, Davis M, Heisler M, Dalton V. Chronic Disease Prevalence Among Mothers Delivering in the United States, 2004-2013 [38N]. Obstetrics & Gynecology. 2017;129(1):S152–S153. doi:10.1097/01.AOG.0000514748.00812.AE [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/01.AOG.0000514748.00812.AE&link_type=DOI) 10. 10.Brown CC, Adams CE, George KE, Moore JE. Associations Between Comorbidities and Severe Maternal Morbidity. Obstetrics and gynecology. 2020;136(5):892–901. doi:10.1097/AOG.0000000000004057 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/AOG.0000000000004057&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 11. 11.Easter SR, Bateman BT, Sweeney VH, et al. A comorbidity-based screening tool to predict severe maternal morbidity at the time of delivery. Am J Obstet Gynecol. 2019;221(3):271.e1-271.e10. doi:10.1016/J.AJOG.2019.06.025 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/J.AJOG.2019.06.025&link_type=DOI) 12. 12.Bateman BT, Mhyre JM, Hernandez-Diaz S, et al. Development of a Comorbidity Index for Use in Obstetric Patients. Obstetrics and gynecology. 2013;122(5):957–965. doi:10.1097/AOG.0B013E3182A603BB [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/AOG.0b013e3182a603bb&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=24104771&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 13. 13.Leonard SA, Kennedy CJ, Carmichael SL, Lyell DJ, Main EK. An Expanded Obstetric Comorbidity Scoring System for Predicting Severe Maternal Morbidity. Obstetrics and gynecology. 2020;136(3):440. doi:10.1097/AOG.0000000000004022 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/AOG.0000000000004022&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 14. 14.Brown HK, McKnight A, Aker A. Association between pre-pregnancy multimorbidity and adverse maternal outcomes: A systematic review. Journal of multimorbidity and comorbidity. 2022;12:26335565221096584. doi:10.1177/26335565221096584 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1177/26335565221096584&link_type=DOI) 15. 15.Goretsky A, Dmitrienko A, Tang I, et al. Data Preparation of the nuMoM2b Dataset. medRxiv. Published online August 26, 2021:2021.08.24.21262142. doi:10.1101/2021.08.24.21262142 [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NzoibWVkcnhpdiI7czo1OiJyZXNpZCI7czoyMToiMjAyMS4wOC4yNC4yMTI2MjE0MnYxIjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjIvMTEvMDIvMjAyMi4xMS4wMi4yMjI4MTgxMi5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 16. 16.Haas DM, Parker CB, Wing DA, et al. A description of the methods of the Nulliparous Pregnancy Outcomes Study: monitoring mothers-to-be (nuMoM2b). Am J Obstet Gynecol. 2015;212(4):539.e1-539.e24. doi:10.1016/J.AJOG.2015.01.019 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.ajog.2015.01.019&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=25648779&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 17. 17.van Buuren S, Boshuizen HC, Knook DL. Multiple imputation of missing blood pressure covariates in survival analysis. STATISTICS IN MEDICINE Statist Med. 1998;18:681–694. doi:10.1002/(SICI)1097-0258(19990330)18:6 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1002/(SICI)1097-0258(19990330)18:6&link_type=DOI) 18. 18.Tang LH, Thygesen LC, Willadsen TG, et al. The association between clusters of chronic conditions and psychological well-being in younger and older people-A cross-sectional, population-based study from the Lolland-Falster Health Study, Denmark. J Comorb. 2020;10:2235042X20981185. doi:10.1177/2235042X20981185 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1177/2235042X20981185&link_type=DOI) 19. 19.Tingley D, Yamamoto T, Hirose K, Keele L, Imai K. mediation: R Package for Causal Mediation Analysis. J Stat Softw. 2014;59(5):1–38. doi:10.18637/JSS.V059.I05 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.18637/JSS.V059.I05&link_type=DOI) 20. 20.Creanga AA, Bateman BT, Kuklina E v., Callaghan WM. Racial and ethnic disparities in severe maternal morbidity: a multistate analysis, 2008-2010. Am J Obstet Gynecol. 2014;210(5):435.e1-435.e8. doi:10.1016/J.AJOG.2013.11.039 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.ajog.2013.11.039&link_type=DOI) 21. 21.Leonard SA, Main EK, Scott KA, Profit J, Carmichael SL. Racial and ethnic disparities in severe maternal morbidity prevalence and trends. Ann Epidemiol. 2019;33:30–36. doi:10.1016/J.ANNEPIDEM.2019.02.007 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/J.ANNEPIDEM.2019.02.007&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 22. 22.Grobman WA, Bailit JL, Rice MM, et al. Racial and Ethnic Disparities in Maternal Morbidity and Obstetric Care. Obstetrics and gynecology. 2015;125(6):1460. doi:10.1097/AOG.0000000000000735 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/AOG.0000000000000735&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 23. 23.Prophet J, Kelly K, Domingo J, et al. Severe pre-eclampsia among pregnant women with sickle cell disease and HIV. Pregnancy Hypertens. 2018;11:87–91. doi:10.1016/J.PREGHY.2018.01.006 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/J.PREGHY.2018.01.006&link_type=DOI) 24. 24.Czerwinski S, Gollero J, Qiu C, Sorensen TK, Williams MA. Migraine-Asthma Comorbidity and Risk of Hypertensive Disorders of Pregnancy. J Pregnancy. 2012;2012. doi:10.1155/2012/858097 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1155/2012/858097&link_type=DOI) 25. 25.Schmidt LJ, Rieger O, Neznansky M, et al. A machine-learning–based algorithm improves prediction of preeclampsia-associated adverse outcomes. Am J Obstet Gynecol. 2022;227(1):77.e1-77.e30. doi:10.1016/j.ajog.2022.01.026 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.ajog.2022.01.026&link_type=DOI) 26. 26.Binder J, Kalafat E, Palmrich P, Pateisky P, Khalil A. Angiogenic markers and their longitudinal change for predicting adverse outcomes in pregnant women with chronic hypertension. Am J Obstet Gynecol. 2021;225(3):305.e1-305.e14. doi:10.1016/j.ajog.2021.03.041 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.ajog.2021.03.041&link_type=DOI) 27. 27.Sheen JJ, Wright JD, Goffman D, et al. Maternal age and risk for adverse outcomes. Am J Obstet Gynecol. 2018;219(4):390.e1-390.e15. doi:10.1016/j.ajog.2018.08.034 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.ajog.2018.08.034&link_type=DOI) 28. 28.Erickson EN, Carlson NS. Maternal Morbidity Predicted by an Intersectional Social Determinants of Health Phenotype: A Secondary Analysis of the NuMoM2b Dataset. Reprod Sci. Published online 2022. doi:10.1007/S43032-022-00913-2 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1007/S43032-022-00913-2&link_type=DOI) 29. 29.Leonard SA, Kennedy CJ, Carmichael SL, Lyell DJ, Main EK. An Expanded Obstetric Comorbidity Scoring System for Predicting Severe Maternal Morbidity. Obstetrics and gynecology. 2020;136(3):440–449. doi:10.1097/AOG.0000000000004022 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/AOG.0000000000004022&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 30. 30.Jean-Francois B, Bailey Lash T, Dagher RK, Green Parker MC, Han SB, Lewis Johnson T. The Potential for Health Information Technology Tools to Reduce Racial Disparities in Maternal Morbidity and Mortality. J Womens Health. 2021;30(2):274. doi:10.1089/JWH.2020.8889 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1089/JWH.2020.8889&link_type=DOI) 31. 31.Declercq ER, Cabral HJ, Cui X, et al. Using Longitudinally Linked Data to Measure Severe Maternal Morbidity. Obstetrics and gynecology. 2022;139(2):165–171. doi:10.1097/AOG.0000000000004641 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/AOG.0000000000004641&link_type=DOI) 32. 32.van Buuren S, Groothuis-Oudshoorn K. Journal of Statistical Software Mice: Multivariate Imputation by Chained Equations in R. Vol 45.; 2011. [http://www.jstatsoft.org/](http://www.jstatsoft.org/) 33. 33.Facco FL, Parker CB, Reddy UM, et al. Association Between Sleep-Disordered Breathing and Hypertensive Disorders of Pregnancy and Gestational Diabetes Mellitus. Obstetrics and gynecology. 2017;129(1):31. doi:10.1097/AOG.0000000000001805 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/AOG.0000000000001805&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=27926645&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 34. 34.King G, Langche Zeng GHE, Alt J, et al. Logistic Regression in Rare Events Data.; 2001. [http://GKing.Harvard.Edu](http://GKing.Harvard.Edu). 35. 35.Weng SC, Wu CL, Kor CT, et al. Migraine and subsequent chronic kidney disease risk: a nationwide population-based cohort study. BMJ Open. 2017;7(12):e018483. doi:10.1136/BMJOPEN-2017-018483 [Abstract/FREE Full Text](http://medrxiv.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NzoiYm1qb3BlbiI7czo1OiJyZXNpZCI7czoxMjoiNy8xMi9lMDE4NDgzIjtzOjQ6ImF0b20iO3M6NTA6Ii9tZWRyeGl2L2Vhcmx5LzIwMjIvMTEvMDIvMjAyMi4xMS4wMi4yMjI4MTgxMi5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 36. 36.Brandes JL. Practical Use of Topiramate for Migraine Prevention. Headache: The Journal of Head and Face Pain. 2005;45(SUPPL. 1):S66–S73. doi:10.1111/J.1526-4610.2005.4501007.X [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1111/J.1526-4610.2005.4501007.X&link_type=DOI) 37. 37.Tanner MS, Davey MA, Mol BW, Rolnik DL. The evolution of the diagnostic criteria of preeclampsia-eclampsia. Am J Obstet Gynecol. 2022;226(2):S835–S843. doi:10.1016/j.ajog.2021.11.1371 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.ajog.2021.11.1371&link_type=DOI) 38. 38.Mikolasevic I, Filipec-Kanizaj T, Jakopcic I, et al. Liver Disease During Pregnancy: A Challenging Clinical Issue. Med Sci Monit. 2018;24:4080. doi:10.12659/MSM.907723 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.12659/MSM.907723&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) 39. 39.Farland L v., Stern JE, Liu C ling, et al. Pregnancy outcomes among women with endometriosis and fibroids: registry linkage study in Massachusetts. Am J Obstet Gynecol. 2022;226(6):829.e1-829.e14. doi:10.1016/j.ajog.2021.12.268 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/j.ajog.2021.12.268&link_type=DOI) 40. 40.Chen Y, Lin M, Guo P, et al. Uterine fibroids increase the risk of hypertensive disorders of pregnancy: a prospective cohort study. J Hypertens. 2021;39(5):1002. doi:10.1097/HJH.0000000000002729 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/HJH.0000000000002729&link_type=DOI) 41. 41.Yu HF, Chen HS, Rao DP, Gong J. Association between polycystic ovary syndrome and the risk of pregnancy complications: A PRISMA-compliant systematic review and meta-analysis. Medicine. 2016;95(51):e4863. doi:10.1097/MD.0000000000004863 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1097/MD.0000000000004863&link_type=DOI) 42. 42.Palomba S, de Wilde MA, Falbo A, Koster MPH, la Sala GB, Fauser BCJM. Pregnancy complications in women with polycystic ovary syndrome. Hum Reprod Update. 2015;21(5):575–592. doi:10.1093/HUMUPD/DMV029 [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/humupd/dmv029&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=26117684&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2022%2F11%2F02%2F2022.11.02.22281812.atom) [1]: /embed/graphic-14.gif [2]: /embed/graphic-15.gif