Applying Automated Machine Learning to Predict Mode of Delivery Using Ongoing Intrapartum Data in Laboring Patients.

OBJECTIVE: This study aimed to develop and validate a machine learning (ML) model to predict the probability of a vaginal delivery (Partometer) using data iteratively obtained during labor from the electronic health record. STUDY DESIGN: A retrospective cohort study of deliveries at an academic, tertiary care hospital was conducted from 2013 to 2019 who had at least two cervical examinations. The population was divided into those delivered by physicians with nulliparous term singleton vertex (NTSV) cesarean delivery rates <23.9% (Partometer cohort) and the remainder (control cohort). The cesarean rate among this population of lower risk patients is a standard metric by which to compare provider rates; <23.9% was the Healthy People 2020 goal. A supervised automated ML approach was applied to generate a model for each population. The primary outcome was accuracy of the model developed on the Partometer cohort at 4 hours from admission to labor and delivery. Secondary outcomes included discrimination ability (receiver operating characteristics-area under the curve [ROC-AUC]), precision-recall AUC, and calibration of the Partometer. To assess generalizability, we compared the performance and clinical predictors identified by the Partometer to the control model. RESULTS: There were 37,932 deliveries during the study period; after exclusions, 9,385 deliveries were included in the Partometer cohort and 19,683 in the control cohort. Accuracy of predicting vaginal delivery at 4 hours was 87.1% for the Partometer (ROC-AUC: 0.82). Clinical predictors of greatest importance in the stacked Intrapartum Partometer Model included the Admission Model prediction and ongoing measures of dilatation and station which mirrored those found in the control population. CONCLUSION: Using automated ML and intrapartum factors improved the accuracy of prediction of probability of a vaginal delivery over both previously published models based on logistic regression. Harnessing real-time data and ML could represent the bridge to generating a truly prescriptive tool to augment clinical decision-making, predict labor outcomes, and reduce maternal and neonatal morbidity. KEY POINTS: · Our ML-based model yielded accurate predictions of mode of delivery early in labor.. · Predictors for models created on populations with high and low cesarean rates were the same.. · A ML-based model may provide meaningful guidance to clinicians managing labor..

COVID-19 and the Drug Overdose Crisis: Uncovering the Deadliest Months in the United States, January‒July 2020.

Objectives. To determine the magnitude of increases in monthly drug-related overdose mortality during the COVID-19 pandemic in the United States. Methods. We leveraged provisional records from the Centers for Disease Control and Prevention provided as rolling 12-month sums, which are helpful for smoothing, yet may mask pandemic-related spikes in overdose mortality. We cross-referenced these rolling aggregates with previous monthly data to estimate monthly drug-related overdose mortality for January through July 2020. We quantified historical errors stemming from reporting delays and estimated empirically derived 95% prediction intervals (PIs). Results. We found that 9192 (95% PI = 8988, 9397) people died from drug overdose in May 2020-making it the deadliest month on record-representing a 57.7% (95% PI = 54.2%, 61.2%) increase over May 2019. Most states saw large-magnitude increases, with the highest in West Virginia, Kentucky, and Tennessee. We observed low concordance between rolling 12-month aggregates and monthly pandemic-related shocks. Conclusions. Unprecedented increases in overdose mortality occurred during the pandemic, highlighting the value of presenting monthly values alongside smoothed aggregates for detecting shocks. Public Health Implications. Drastic exacerbations of the US overdose crisis warrant renewed investments in overdose surveillance and prevention during the pandemic response and postpandemic recovery efforts.

International COVID-19 mortality forecast visualization: covidcompare.io.

COVID-19 mortality forecasting models provide critical information about the trajectory of the pandemic, which is used by policymakers and public health officials to guide decision-making. However, thousands of published COVID-19 mortality forecasts now exist, many with their own unique methods, assumptions, format, and visualization. As a result, it is difficult to compare models and understand under which circumstances a model performs best. Here, we describe the construction and usability of covidcompare.io, a web tool built to compare numerous forecasts and offer insight into how each has performed over the course of the pandemic. From its launch in December 2020 to June 2021, we have seen 4600 unique visitors from 85 countries. A study conducted with public health professionals showed high usability overall as formally assessed using a Post-Study System Usability Questionnaire. We find that covidcompare.io is an impactful tool for the comparison of international COVID-19 mortality forecasting models.

Deciphering intratumoral heterogeneity using integrated clonal tracking and single-cell transcriptome analyses.

Cellular heterogeneity is a major cause of treatment resistance in cancer. Despite recent advances in single-cell genomic and transcriptomic sequencing, it remains difficult to relate measured molecular profiles to the cellular activities underlying cancer. Here, we present an integrated experimental system that connects single cell gene expression to heterogeneous cancer cell growth, metastasis, and treatment response. Our system integrates single cell transcriptome profiling with DNA barcode based clonal tracking in patient-derived xenograft models. We show that leukemia cells exhibiting unique gene expression respond to different chemotherapies in distinct but consistent manners across multiple mice. In addition, we uncover a form of leukemia expansion that is spatially confined to the bone marrow of single anatomical sites and driven by cells with distinct gene expression. Our integrated experimental system can interrogate the molecular and cellular basis of the intratumoral heterogeneity underlying disease progression and treatment resistance.

Resident microbes of lactation rooms and daycares.

Dedicated lactation rooms are a modern development as mothers return to work while still providing breastmilk to their absent infants. This study describes the built environment microbiome of lactation rooms and daycares, and explores the influence of temperature and humidity on the microbiome of lactation rooms. Sterile swabs were used to collect samples from five different sites in lactation rooms at University of California, Davis and from five different sites in daycares located in Davis, California. DNA from the swabs was extracted and the V4 region of the 16S rRNA gene was sequenced using Illumina MiSeq. Temperature and relative humidity data were collected on a subset of the lactation rooms. Sampled lactation rooms could be either dedicated lactation rooms or could also serve other functions (e.g., combined lactation room and restroom lounge). The majority of sequence reads were identified as belonging to family Moraxellaceae, with 73% of all reads included in analysis identified as an unknown species of Acinetobacter. Alpha diversity was analyzed using the Shannon index, while beta diversity was analyzed using unweighted and weighted UniFrac distance. The Jaccard distance was used to measure amount of change at sampling locations between time points for analysis of the impact of temperature and humidity on the microbiome. There were significant differences in the beta diversity of the microbiome of lactation rooms by room type. There were also significant differences in the beta diversity of the microbiome by sample collection location. There were no significant differences in either alpha or beta diversity associated with room temperature or humidity. Additional studies are needed to understand if the differences in lactation room type may result in differences in the breastmilk microbiome of milk collected in those rooms, and to what extent any such differences may influence the infant microbiome.

Oxylipins in triglyceride-rich lipoproteins of dyslipidemic subjects promote endothelial inflammation following a high fat meal.

Elevated triglyceride-rich lipoproteins (TGRL) in circulation is a risk factor for atherosclerosis. TGRL from subjects consuming a high saturated fat test meal elicited a variable inflammatory response in TNFα-stimulated endothelial cells (EC) that correlated strongly with the polyunsaturated fatty acid (PUFA) content. This study investigates how the relative abundance of oxygenated metabolites of PUFA, oxylipins, is altered in TGRL postprandially, and how these changes promote endothelial inflammation. Human aortic EC were stimulated with TNFα and treated with TGRL, isolated from subjects' plasma at fasting and 3.5 hrs postprandial to a test meal high in saturated fat. Endothelial VCAM-1 surface expression stimulated by TNFα provided a readout for atherogenic inflammation. Concentrations of esterified and non-esterified fatty acids and oxylipins in TGRL were quantified by mass spectrometry. Dyslipidemic subjects produced TGRL that increased endothelial VCAM-1 expression by ≥35%, and exhibited impaired fasting lipogenesis activity and a shift in soluble epoxide hydrolase and lipoxygenase activity. Pro-atherogenic TGRL were enriched in eicosapentaenoic acid metabolites and depleted in esterified C18-PUFA-derived diols. Abundance of these metabolites was strongly predictive of VCAM-1 expression. We conclude the altered metabolism in dyslipidemic subjects produces TGRL with a unique oxylipin signature that promotes a pro-atherogenic endothelial phenotype.

Bifidobacterial Dominance of the Gut in Early Life and Acquisition of Antimicrobial Resistance.

Bifidobacterium species are important commensals capable of dominating the infant gut microbiome, in part by producing acids that suppress growth of other taxa. Bifidobacterium species are less prone to possessing antimicrobial resistance (AMR) genes (ARGs) than other taxa that may colonize infants. Given that AMR is a growing public health crisis and ARGs are present in the gut microbiome of humans from early life, this study examines the correlation between a Bifidobacterium-dominated infant gut microbiome and AMR levels, measured by a culture-independent metagenomic approach both in early life and as infants become toddlers. In general, Bifidobacterium dominance is associated with a significant reduction in AMR in a Bangladeshi cohort, both in the number of acquired AMR genes present and in the abundance of AMR genes. However, by year 2, Bangladeshi infants had no significant differences in AMR related to their early-life Bifidobacterium levels. A generalized linear model including all infants in a previously published Swedish cohort found a significant negative association between log-transformed total AMR and Bifidobacterium levels, thus confirming the relationship between Bifidobacterium levels and AMR. In both cohorts, there was no change between early-life and later-life AMR abundance in high-Bifidobacterium infants but a significant reduction in AMR abundance in low-Bifidobacterium infants. These results support the hypothesis that early Bifidobacterium dominance of the infant gut microbiome may help reduce colonization by taxa containing ARGs.IMPORTANCE Infants are vulnerable to an array of infectious diseases, and as the gut microbiome may serve as a reservoir of AMR for pathogens, reducing the levels of AMR in infants is important to infant health. This study demonstrates that high levels of Bifidobacterium are associated with reduced levels of AMR in early life and suggests that probiotic interventions to increase infant Bifidobacterium levels have the potential to reduce AMR in infants. However, this effect is not sustained at year 2 of age in Bangladeshi infants, underscoring the need for more detailed studies of the biogeography and timing of infant AMR acquisition.