Pairwise Associations Research Articles

ConNEcT: An R package to build contingency measure-based networks on binary time series.

Dynamic networks are valuable tools to depict and investigate the concurrent and temporal interdependencies of various variables across time. Although several software packages for computing and drawing dynamic networks have been developed, software that allows investigating the pairwise associations between a set of binary intensive longitudinal variables is still missing. To fill this gap, this paper introduces an R package that yields contingency measure-based networks (ConNEcT). ConNEcT implements different contingency measures: proportion of agreement, corrected and classic Jaccard index, phi correlation coefficient, Cohen's kappa, odds ratio, and log odds ratio. Moreover, users can easily add alternative measures, if needed. Importantly, ConNEcT also allows conducting non-parametric significance tests on the obtained contingency values that correct for the inherent serial dependence in the time series, through a permutation approach or model-based simulation. In this paper, we provide an overview of all available ConNEcT features and showcase their usage. Addressing a major question that users are likely to have, we also discuss similarities and differences of the included contingency measures.

On the efficiency of indicator species for broad-scale monitoring of bird diversity across climate conditions

Multiple constraints (e.g., time, funding, expertise) can impede the broad-scale monitoring of human-induced loss of biodiversity. Non-random species co-occurrence provides an opportunity to estimate local species richness by tracking only a few species that are indicators of more global assemblages. Despite promising fine-scale studies, it remains unclear whether such indicator species remain effective over broad spatial extents. We assessed the value of indicator species to consistently predict species richness despite variation in short-term (5–9 years) climate conditions. Our study involves 22,041 point-count stations distributed over 1.75 M km2 of Canadian boreal forest, where 197,000 birds from 216 species were detected. Using null model analysis of species co-occurrence followed by regression analysis, we identified the minimum set of indicator species that can predict 70% of spatial variation in local bird richness in each of eight short-term climate regions. We found that indicators were non-random subsets of the species pool, as they were species whose presence explained a relatively high percentage of variations in species richness within the short-term climate region, and those sharing a relatively high number of significant pairwise associations. Although only 11 to 26 indicator species were needed to predict species richness at the regional scale, 57 of the 216 species pool (26%) were necessary to make predictions over the entire study area. This large number reflects regional variations in the best indicator species, and those that remained indicators in several regions were representative of different species assemblages. Our observations thus cast doubts on the use of indicator species as an effective and efficient tool for biodiversity monitoring under changing climate conditions. Broad-scale (spatial or temporal) use of indicators thus comes with the colossal challenge of having to determine under which new conditions a given set of indicators must be replaced by another, and by which one.

Genetic Architecture of Stroke of Undetermined Source: Overlap with Known Stroke Etiologies and Associations with Modifiable Risk Factors.

Ischemic stroke etiology remains undetermined in 30% of cases. We explored the genetic architecture of stroke classified as undetermined to test if mechanisms and risk factors underlying large-artery atherosclerotic (LAAS), cardioembolic (CES), and small-vessel stroke (SVS) contribute to its pathogenesis. We analyzed genome-wide data from 16,851 ischemic stroke cases and 32,473 controls. Using polygenic risk scores for LAAS, CES, and SVS, we assessed the genetic overlap with stroke of undetermined source and used pairwise genomewide association study (GWAS-PW) to search for shared loci. We then applied Mendelian randomization (MR) to identify potentially causal risk factors of stroke of undetermined source. Genetic risk for LAS, CES, and SVS was associated with stroke of undetermined source pointing to overlap in their genetic architecture. Pairwise analyses revealed 19 shared loci with LAAS, 2 with CES, and 5 with SVS that have been implicated in atherosclerosis-related phenotypes. Genetic liability to both carotid atherosclerosis and atrial fibrillation was associated with stroke of undetermined source, but the association with atrial fibrillation was attenuated after excluding cases with incomplete diagnostic workup. MR analyses showed effects of genetically determinants of blood pressure, diabetes, waist-to-hip ratio, inflammatory pathways (IL-6 signaling, MCP-1/CCL2 levels), and factor XI levels on stroke of undetermined source. Stroke of undetermined source shares genetic and vascular risk factors with other stroke subtypes, especially LAAS, thus highlighting the diagnostic limitations of current subtyping approaches. The potentially causal associations with carotid atherosclerosis and atherosclerotic risk factors might have implications for prevention strategies targeting these mechanisms. ANN NEUROL 2022;91:640-651.

Genetic Regulation of DNA Methylation Yields Novel Discoveries in GWAS of Colorectal Cancer.

Colorectal cancer has a strong epigenetic component that is accompanied by frequent DNA methylation (DNAm) alterations in addition to heritable genetic risk. It is of interest to understand the interrelationship of germline genetics, DNAm, and colorectal cancer risk. We performed a genome-wide methylation quantitative trait locus (meQTL) analysis in 1,355 people, assessing the pairwise associations between genetic variants and lymphocytes methylation data. In addition, we used penalized regression with cis-genetic variants ± 1 Mb of methylation to identify genome-wide heritable DNAm. We evaluated the association of genetically predicted methylation with colorectal cancer risk based on genome-wide association studies (GWAS) of over 125,000 cases and controls using the multivariate sMiST as well as univariately via examination of marginal association with colorectal cancer risk. Of the 142 known colorectal cancer GWAS loci, 47 were identified as meQTLs. We identified four novel colorectal cancer-associated loci (NID2, ATXN10, KLHDC10, and CEP41) that reside over 1Mb outside of known colorectal cancer loci and 10 secondary signals within 1Mb of known loci. Leveraging information of DNAm regulation into genetic association of colorectal cancer risk reveals novel pathways in colorectal cancer tumorigenesis. Our summary statistics-based framework sMiST provides a powerful approach by combining information from the effect through methylation and residual direct effects of the meQTLs on disease risk. Further validation and functional follow-up of these novel pathways are needed. Using genotype, DNAm, and GWAS, we identified four new colorectal cancer risk loci. We studied the landscape of genetic regulation of DNAm via single-SNP and multi-SNP meQTL analyses.

Pairwise association of key lifestyle factors and risk of colorectal cancer: a prospective pooled multicohort study.

Several lifestyle factors are associated with an increased risk of colorectal cancer (CRC). Although lifestyle factors co-occur, in most previous studies these factors have been studied focusing upon a single risk factor or assuming independent effects between risk factors. To examine the pairwise effects and interactions of smoking, alcohol consumption, physical inactivity, and body mass index (BMI) with risk of subsequent colorectal cancer (CRC). We used METCA cohort data (pooled data from seven population-based Finnish health behavior survey studies during years 1972-2015) consisting of 171 063 women and men. Participants' smoking, alcohol consumption, physical inactivity and BMI measures were gathered, and participants were categorized into those exposed and those not exposed. The incidence of CRC was modeled by Poisson regression with main and interaction effects of key lifestyle factors. The cohort members were followed-up through register linkage to the Finnish Cancer Registry for first primary CRC case until the end of 2015. Follow-up time was 1715, 690 person years. The highest pairwise CRC risk was among male smokers who had overweight (BMI ≥ 25 kg/m2 ) (HR 1.75, 95% CI 1.36-2.26) and women who had overweight and consumed alcohol (HR 1.45, 95% CI 1.14-1.85). Overall, among men the association of lifestyle factors and CRC risk was stronger than among women. In men, both having overweight and being a smoker combined with any other adverse lifestyle factor increased CRC risk. Among women, elevated CRC risks were observed for those who were physically inactive and who consumed alcohol or had overweight. No statistically significant interactions were detected between pairs of lifestyle factors. This study strengthens the evidence of overweight, smoking, and alcohol consumption as CRC risk factors. Substantial protective benefits in CRC risk can be achieved by preventing smoking, maintaining BMI to <25 kg/m2 and not consuming alcohol.

Do Children Use Multi-Word Information in Real-Time Sentence Comprehension?

Meaning in language emerges from multiple words, and children are sensitive to multi-word frequency from infancy. While children successfully use cues from single words to generate linguistic predictions, it is less clear whether and how they use multi-word sequences to guide real-time language processing and whether they form predictions on the basis of multi-word information or pairwise associations. We address these questions in two visual-world eye-tracking experiments with 5- to 8-year-old children. In Experiment 1, we asked whether children generate more robust predictions for the sentence-final object of highly frequent sequences (e.g., "Throw the ball"), compared to less frequent sequences (e.g., "Throw the book"). We further examined if gaze patterns reflect event knowledge or phrasal frequency by comparing the processing of phrases that have the same event structure but differ in multi-word content (e.g., "Brush your teeth" vs. "Brush her teeth"). In the second study, we employed a training paradigm to ask if children are capable of generating predictio.ns from novel multi-word associations while controlling for the overall frequency of the sequences. While the results of Experiment 1 suggested that children primarily relied on event associations to generate real-time predictions, those of Experiment 2 showed that the same children were able to use recurring novel multi-word sequences to generate real-time linguistic predictions. Together, these findings suggest that children can draw on multi-word information to generate linguistic predictions, in a context-dependent fashion, and highlight the need to account for the influence of multi-word sequences in models of language processing.

Dietary correlates of urinary phthalate metabolite concentrations in 6-19 Year old children and adolescents.

Children are vulnerable to adverse health effects associated with phthalates, and food is one source of exposure. A comprehensive analysis investigating urinary phthalate metabolite concentrations in relation to food type and source has yet to be undertaken. We use reduced rank regression, a dimension reduction method, to identify dietary patterns associated with urinary phthalate metabolites in children in a large US study. We used data from 2369 participants 6-19 years old from the 2011-2016 National Health and Nutrition Examination Survey who recalled their diet over the 24h prior to urine collection. We used dietary data to estimate intake and source (i.e., prepared at a restaurant vs. purchased from a grocery store) of 136 food groups. We used reduced rank regression to identify dietary patterns explaining variation in overall urinary concentrations of ∑di-2-ethylhexyl phthalate and seven phthalate metabolites. We also examined pairwise associations between food groups and urinary phthalate metabolites. We identified eight dietary patterns that cumulatively explained 12.1% of variation in urinary phthalate metabolites, including a dietary pattern characterized by certain starchy vegetables (e.g., plantains and lima beans), quick breads, and citrus juice prepared at a restaurant. A one SD increase in this food pattern score was associated with a 37.2% higher monocarboxyoctyl phthalate (MCOP) concentration (95% CI: 30.3, 44.4). We also observed weak associations between certain food groups and urinary phthalate metabolites (e.g., a one SD increase in intake of certain starchy vegetables prepared at a restaurant was associated with a 1.8% [95% CI: 0.7, 2.8] higher MCOP). Children whose diets were characterized by higher consumption of certain starchy vegetables, quick breads, and citrus juices prepared at a restaurant had higher urinary phthalate metabolites. More detailed information on the specific methods of food processing and details on packaging materials is needed.

Posttraumatic stress symptoms and financial toxicity among adolescent and young adult oncology patients and their caregivers at cancer diagnosis.

Adolescent and young adult oncology (AYAO) patients and caregivers may experience significant psychosocial dysfunction and financial toxicity. Understanding early risk factors is critical to improving survivorship trajectories. The authors conducted a cross-sectional study of baseline survey data from a prospective cohort of AYAO patient-caregiver dyads enrolled within 1month of medical oncology treatment initiation. Posttraumatic stress symptoms (PTSS) were measured by the Impacts of Events Scale-Revised, and financial toxicity was measured with the Comprehensive Score (COst). The authors fit models of linear association between PTSS, financial toxicity, and other end points and pairwise associations of PTSS and financial toxicity within dyads. The analytic cohort contained 41 patients, 37 caregivers, and 34 complete dyads. Clinically-concerning PTSS were observed among patients (44%) and caregivers (52%). The median COst scores were 20.0 for patients (quartiles, 12.5-29.5) and 22.0 for caregivers (quartiles, 12.8-26.0), which were consistent with high financial toxicity (patients, 46%; caregivers, 44%). PTSS were positively associated with financial toxicity (P=.013 for patients, P=.039 for caregivers), subjective distress (P<.001 for all), depressive (P<.001 for all) and anxiety symptoms (P=.005 for patients, P=.024 for caregivers), and poorer quality of life (P<.001 for patients, P=.003 for caregivers). A significant paired association was not found in PTSS (Pearson correlation coefficient [PCC], 0.23; 95% confidence interval [CI], -0.15 to 0.56). Financial toxicity was positively associated within dyads (PCC, 0.65; 95% CI, 0.36-0.83). At diagnosis, AYAO patients and caregivers exhibit substantial PTSS, which are associated with greater financial toxicity and other psychosocial distress.

The quantification of Simpson’s paradox and other contributions to contingency table theory

The analysis of contingency tables is a powerful statistical tool used in experiments with categorical variables. This study improves parts of the theory underlying the use of contingency tables. Specifically, the linkage disequilibrium parameter as a measure of two-way interactions applied to three-way tables makes it possible to quantify Simpson's paradox by a simple formula. With tests on three-way interactions, there is only one that determines whether the partial interactions of all variables agree or whether there is at least one variable whose partial interactions disagree. To date, there has been no test available that determines whether the partial interactions of a certain variable agree or disagree, and the presented work closes this gap. This work reveals the relation of the multiplicative and the additive measure of a three-way interaction. Another contribution addresses the question of which cells in a contingency table are fixed when the first- and second-order marginal totals are given. The proposed procedure not only detects fixed zero counts but also fixed positive counts. This impacts the determination of the degrees of freedom. Furthermore, limitations of methods that simulate contingency tables with given pairwise associations are addressed.

Concordance between the schedule for the evaluation of individual quality of life-direct weighting (SEIQoL-DW) and the EuroQoL-5D (EQ-5D) measures of quality of life outcomes in adults with X-linked hypophosphatemia

BackgroundAccurate measurement of any constructs in clinical studies is of critical importance, especially if the adoption of an intervention relies on detecting a significant treatment effect where one exists. Under Neutral theory, the amount of relevant and irrelevant indicators selected to operationalise the construct contribute equally to the accuracy of the observation. The Neutral or accurate observation is achieved by observing all relevant indicators only. Generic QoL instruments such as EQ-5D are increasingly being accepted as imprecise, especially in rare diseases, based on the relevance of their indicators. QoL is a construct that embodies a patient's subjectivity, individuality, and local circumstances at measurement. SEIQoL-DW is an instrument designed to respect these characteristics of QoL through eliciting indicators or cues directly from the subject along with the proportion of the overall QoL they contribute. EQ-5D and SEIQoL can therefore be considered as being at opposing ends of accuracy in QoL measurement. XLH is a hereditary, progressive, rare disease characterised by phosphate wasting, affecting both children and adults and impacting their QoL. The purpose of this study was to observe if any change in QoL of adult XLH patients were detectable using EQ-5D, SEIQoL eliciting new cues at each visit, and SEIQoL administering baseline cues overall visits (thereby silencing its time-dependency) versus baseline over 12 months. In addition, any association between the three sets of observations was explored.ResultsAll quality of life scores were observed to decrease from baseline by 13.36%, 7.32% and 2.7% based on SEIQoLvisit_cues, SEIQoLbaseline_cues, and EQ-5D assessments, respectively. The decrease in the quality of life scores was only statistically significant (p = 0.037) for SEIQoLvisit_cues. Beyond the baseline visit, the only highly positive and statistically significant pairwise association was between SEIQoLvisit_cues and SEIQoLbaseline_cues at M6 (ρ = 0.782, P value < 0.05) and M9 (ρ = 0.879, P value < 0.05).ConclusionsEQ-5D and SEIQoLbaseline_cues failed to detect the same statistically significant decrease in QoL observed by SEIQoLvisit_cues. Both sets of SEIQoL observations were more closely associated with each other than with EQ-5D. Observing constructs such as QoL in rare diseases benefit from a Neutrality in indicator selection and respecting variation in dominance of various indicators over time.

Regularity of prescriptions for intermediate and advanced lung cancer based on latent structure combined with association rules

The present study explored the regularity of prescriptions for the treatment of intermediate and advanced lung cancer to provide references for clinical medication. CNKI, Wanfang, VIP, and CBM were searched for the research papers on the treatment of lung cancer by Chinese medicine published from database inception to May 31, 2021. The relevant information of qualified papers was extracted to establish a database. The Chinese medicines with frequency &gt;3% underwent analysis of the latent structure and association rules by Lantern 5.0 and SPSS Molder 14.1, respectively, and the prescription regularity in the treatment of intermediate and advanced lung cancer was analyzed based on the frequency description. A total of 713 papers were included, involving 327 Chinese medicines with a cumulative frequency of 12 794 and 106 prescriptions with a cumulative frequency of 824. The commonly used Chinese medicines were dominated by deficiency-tonifying, heat-clearing, phlegm-resolving, and cough/dyspnea-relieving drugs, such as Astragali Radix, Atractylodis Macrocephalae Rhizoma, Glycyrrhizae Radix et Rhizoma, Ophiopogonis Radix, Poria, and Hedyotis Diffusa, which are cold, warm, and plain in nature and sweet, bitter, and pungent in flavor, and mainly act on lung, spleen, and stomach meridians. Commonly used prescriptions included Shashen Maidong Decoction, Liujunzi Decoction, and Baihe Gujin Decoction. The latent structure analysis revealed 32 latent variables and 65 hidden classes. Six comprehensive clustering models and 11 core prescriptions were obtained by professional knowledge inference. The common syndromes of intermediate and advanced lung cancer were inferred to be Qi and Yin deficiency in the lung, Qi deficiency in the lung and spleen, Yin deficiency in the liver and kidney, combined phlegm and stasis, phlegm-heat obstructing lung, and Qi stagnation and blood stasis. Forty-four strong associations were screened out by association rules analysis, including four pairwise strong associations(Polygonati Odorati Rhizoma→Ophiopogonis Radix, Polygonati Odorati Rhizoma→Glehniae Radix, Amomi Fructus→Atractylodis Macrocephalae Rhizoma, and Polygonati Rhizoma→Astragali Radix) and 40 triplet strong associations(such as Trichosanthis Radix+Glehniae Radix→Ophiopogonis Radix, Polygonati Odorati Rhizoma+Glehniae Radix→Ophiopogonis Radix, Trichosanthis Radix+Ophiopogonis Radix→Glehniae Radix, and Scutellariae Barbatae Herba+Codonopsis Radix→Hedyotis Diffusa). In the treatment of intermediate and advanced lung cancer, Qi-replenishing and Yin-nourishing drugs are mainly employed, assisted with cancer-resisting, toxin-removing, spleen-invigorating, phlegm/stasis-resolving, and blood-activating drugs based on syndrome differentiation. The roots were treated following the principles of tonifying lungs and replenishing the spleen, and symptoms following the principles of removing the toxin, dispelling stasis, and resolving phlegm.

A method for machine learning generation of realistic synthetic datasets for validating healthcare applications.

Digital health applications can improve quality and effectiveness of healthcare, by offering a number of new tools to users, which are often considered a medical device. Assuring their safe operation requires, amongst others, clinical validation, needing large datasets to test them in realistic clinical scenarios. Access to datasets is challenging, due to patient privacy concerns. Development of synthetic datasets is seen as a potential alternative. The objective of the paper is the development of a method for the generation of realistic synthetic datasets, statistically equivalent to real clinical datasets, and demonstrate that the Generative Adversarial Network (GAN) based approach is fit for purpose. A generative adversarial network was implemented and trained, in a series of six experiments, using numerical and categorical variables, including ICD-9 and laboratory codes, from three clinically relevant datasets. A number of contextual steps provided the success criteria for the synthetic dataset. A synthetic dataset that exhibits very similar statistical characteristics with the real dataset was generated. Pairwise association of variables is very similar. A high degree of Jaccard similarity and a successful K-S test further support this. The proof of concept of generating realistic synthetic datasets was successful, with the approach showing promise for further work.

Comparative evaluation of effectiveness of manual and various rotary retreatment techniques for removal of root canal filling material: An in-vitro study

Aim: This study intended to compare the effectiveness of manual files H-files and ProTaper universal retreatment (PTUR), R-Endo retreatment, and gutta-percha removal (GPR) Mani retreatment rotary files for removing root canal filling material.Materials and Methods: Eighty mandibular premolars with single canal were selected. The canals were enlarged with ProTaper rotary files up to F2 and obturated using the lateral condensation technique. Then, the samples were randomly divided into four groups and retreated using hand files and rotary retreatment files. The area of remaining root canal filling material was evaluated by splitting the tooth using a stereomicroscope with ×10 and auto computer-aided design software and the percentage of remaining filling material on canal walls was calculated with equation: area % of remaining filling material = area of remaining material ×100/area of canal. One-way ANOVA was carried out to compare the groups and post hoc Tukey's test for pairwise association with each of four groups.Results: Group A (H-files) showed 61.87% of remaining filling material, Group B (PTUR) showed 24.94%, Group C (R-Endo) showed 37.12%, and Group D (GPR Mani) showed 45.95% of remaining root canal filling material.Conclusions: All the groups showed residual root canal filling material. Least was seen in group retreated using PTUR files and maximum was seen in group retreated with H-files.

Pictorial Blood Loss Assessment Chart for evaluating heavy menstrual bleeding in Asian women.

Heavy menstrual bleeding is a common gynaecological problem, but some women may prefer not to articulate their menstrual problems. The objective of this study was to evaluate the usefulness and acceptability of the Pictorial Blood Loss Assessment Chart (PBAC) as a selfscreening tool in evaluation of menstrual blood loss among Asian women in Hong Kong. This prospective cohort study recruited 206 women from the general gynaecology ward and out-patient clinic: 118 had self-perceived heavy menstrual bleeding and 88 had self-perceived normal menstrual flow. Participants were asked to fill in the PBAC for one menstrual cycle. Compared with women who had self-perceived normal menstrual flow, women with self-perceived heavy menstrual bleeding had significantly higher total PBAC scores and numbers of flooding episodes, larger clot sizes and numbers, more days of bleeding, and lower haemoglobin levels. Receiver-operating characteristic curve analysis demonstrated good pairwise associations of self-perceived symptoms with PBAC score and haemoglobin level. The PBAC can be used to differentiate self-perceived heavy and normal menstrual bleeding in Asian women in Hong Kong. It can also serve as an additional indicator of possible heavy menstrual bleeding to alert women of the need to seek early medical attention.

The Coexistence Relationship Between Plants and Soil Bacteria Based on Interdomain Ecological Network Analysis.

The relationship between plants and their associated soil microbial communities plays a crucial role in maintaining ecosystem processes and function. However, identifying these complex relationships is challenging. In this study, we constructed an interdomain ecology network (IDEN) of plant–bacteria based on SparCC pairwise associations using synchronous aboveground plant surveys and belowground microbial 16S rRNA sequencing among four different natural forest types along the climate zones in China. The results found that a total of 48 plants were associated with soil bacteria among these four sites, and soil microbial group associations with specific plant species existed within the observed plant–bacteria coexistence network. Only 0.54% of operational taxonomy units (OTUs) was shared by the four sites, and the proportion of unique OTUs for each site ranged from 43.08 to 76.28%, which occupied a large proportion of soil bacterial community composition. The plant–bacteria network had a distinct modular structure (p < 0.001). The tree Acer tetramerum was identified as the network hubs in the warm temperate coniferous and broad-leaved mixed forests coexistence network and indicates that it may play a key role in stabilizing of the community structure of these forest ecosystems. Therefore, IDEN of plant–bacteria provides a novel perspective for exploring the relationships of interdomain species, and this study provides valuable insights into understanding coexistence between above-ground plants and below-ground microorganisms.

1368. Species Diversity and Co-Occurrence Patterning in Bone and Soft Tissue Infections

Abstract Background Polymicrobial results from conventional cultures in bone and soft tissue infections have generally been reported as independent co-occurrences (i.e. assuming no relationship or interaction between species). The objective of this study was to identify non-random (dependent) co-occurrence in bone and soft tissue infections. Methods We used single center Theradoc (Premiere Inc., Charlotte, NC) data on microbiology specimens collected over a decade (4/2010 to 4/2020). We included only tissue, fluid and swab specimens identified as being obtained from bone, skin, soft tissue or a wound during a patient’s first episode of infection. Blood and liver biopsy specimens were excluded. Patterning involving &amp;gt;2 organisms was examined with factor analysis. Analyses were done with R version 3.6.1. Threshold p-values were adjusted for multiple comparisons, often to p&amp;lt; 0.001. Results Polymicrobial culture results were seen much more frequently in the foot (76%) and sacrum (78%) than specimens from other locations (43%). The highest Shannon diversity index was seen in sacral specimens, followed by foot and leg specimens. Pseudomonas aeruginosa and E. coli was found more frequently in sacral bone and soft tissue specimens, while S. aureus was found most often in the spine bone and foot bone and soft tissue. Corynebacterium was found more frequently in foot bone (23%) and foot soft tissue (18%) than in non-foot soft tissue specimens (11%). Analysis of pairwise species co-occurrences in foot specimens (n=765) demonstrated that S. aureus made significantly less frequent the presence of Bacteroides, Enterobacter, and other staphylococcal species (Table 1). Analysis of non-foot soft tissue specimens (n=2,647) confirmed this pairwise associations and suggested three patterns (Table 2): A, polymicrobial without S. aureus; B, Enterococcus + gram negatives; and C, streptococcal. Table 1. Pairwise co-occurrences between species in bone and soft tissue specimens from foot infections (n=765). Factor analysis results demonstrating species associated with three polymicrobial patterns in soft tissue infection specimens. Blue denotes positive association (more often occurring as part of the pattern), red or grey denotes negative associations (less often occurring as part of the pattern). Numbers denote factor weights, a measure of the strength of the association. Conclusion Species incidence and diversity vary by location and tissue type. The many non-random occurrences found suggest complex interrelationships between microbes in soft tissue and bone infections, including organisms often classified as non-pathogens. Further study may further characterize these relationships and aid in antibiotic selection. Disclosures All Authors: No reported disclosures

Predictive Value of Second-Trimester Maternal Lipid Profiling in Early-Onset Pre-eclampsia: A Prospective Cohort Study and Nomogram.

Purpose: Maternal lipid profile in second trimester has rarely been investigated in the risk assessment for pre-eclampsia (PE). Since early-onset PE often companied by much worse clinical outcomes, thus, we aimed to evaluate the predictive value of second-trimester maternal lipid profiling for early-onset PE.Methods: A prospective cohort study was conducted to measure the second-trimester maternal lipid profile of pregnant women from January to December 2019. The pairwise association between maternal lipid profile and PE onset or pregnancy termination time was quantified. Multiple logistic regression was preformed to define risk factors for early-onset PE, and a nomogram for early-onset PE was developed. The net benefit of our model was evaluated by calibration curve and decision curve analyses.Results: We enrolled 5,908 pregnant women and they were divided into healthy (n = 5,789), late-onset PE (n = 64), and early-onset PE (n = 55) groups. Total cholesterol (TC), triglycerides (TG), and low-density lipoprotein cholesterol (LDL-c) were elevated in patients with PE, while high-density lipoprotein cholesterol (HDL-c) was decreased in patients with PE. TC, TG, and LDL-c were negatively correlated with PE onset time or gestational week at delivery. Receiver operating characteristic curves (ROC) defined the cutoff values of TG and HDL-c, and the final regression model included five statistically significant risk predictors for early-onset PE (maternal age of ≥35 years, multipara, pre-pregnancy body mass index (BMI) ≥25 kg/m2, second trimester TG ≥ 2.59 mmol/L and second trimester HDL-c ≤ 2.03 mmol/L. The nomogram had an excellent diagnostic performance (area under the curve = 0.912, sensitivity = 92.7%, and specificity = 76%) and was further validated with good calibration and positive net benefits in a decision curve analysis.Conclusions: An abnormally increased TG concentration and a decreased HDL-c concentration might serve as predictors of early-onset PE. Whether blood lipid-lowering measures can improve severe PE prognosis require further clarification.

Data and Network Analytics for COVID-19 ICU Patients: A Case Study for a Spanish Hospital.

The COVID-19 pandemic presents unprecedented challenges to the healthcare systems around the world. In 2020, Spain was among the countries with the highest Intensive Care Unit (ICU) hospitalization and mortality rates. This work analyzes data of COVID-19 patients admitted to a Spanish ICU during the first wave of the pandemic. The patients in our study either died (deceased patients) or were discharged from the ICU (non-deceased patients) and underwent the following landmarks: beginning of symptoms; arrival at the emergency department; beginning of the hospital stay; and ICU admission. Our goal is to create a graph-based data-science methodology to find associations among patients' comorbidities, previous medication, symptoms, and the COVID-19 treatment, and to analyze their evolution across landmarks. Towards that end, we first perform a hypothesis test based on bootstrap to identify discriminative features among deceased and non-deceased patients. Then, we leverage graph-based representations and network analytics to determine pairwise associations and complex relations among clinical features. The descriptive statistical analysis confirms that deceased patients exhibit multiple comorbidities with stronger levels of association and are treated with a wider range of drugs during the ICU stay. We also observe that the most common treatment was the simultaneous administration of lopinavir/ritonavir with hydroxychloroquine, regardless of the patients' outcome. Our results illustrate how graph tools and representations yield insights on the relations among comorbidities, drug treatments, and patients' evolution. All in all, the approach puts forth a new data-analysis tool for clinicians that can be applied to analyze (post-COVID) symptom/patient evolution.

Psychiatric comorbidity in the Baltimore ECA follow-up study: the matrix approach.

Psychiatric comorbidity is defined as the joint occurrence of two or more mental or substance use disorders. Widespread psychiatric comorbidity has been reported in treatment and population-based studies. The aim of this study was to measure the extent and impact of psychiatric comorbidity in a cohort of the Baltimore Epidemiologic Catchment Area study. We examined the comorbidity burden of 16 mental disorders in a cohort of 847 participants using both established and novel analytical approaches The Comorbidity to Diagnosis Inflation Ratio (CDIR), is a statistical instrument that quantifies impact of pairwise comorbid associations, both on the whole sample, as well as on each specific disorder. Most anxiety disorders had substantial co-occurrence with each other, as well as with Major Depressive Disorder (MDD). In addition, mood disorders had a high degree of comorbidity with Alcohol Dependence (AD). The CDIR for the whole sample was 1.32, indicating a ratio of 132 comorbidities per 100 diagnoses. The conditions with high sample prevalence were relatively less comorbid than the low prevalence conditions. Obsessive Compulsive Disorder had a comorbidity burden that was 89% greater than the overall sample. Anxiety disorders are highly interrelated, as well as highly comorbid with depression. The comorbidity phenomenon is linked to the differential prevalence of the analyzed conditions. Comorbidity frequency (most prevalent comorbid condition) appears mutually exclusive to comorbidity burden (most widely interrelated condition). While AD and MDD were the most frequently diagnosed disorders; low prevalence conditions as OCD and GAD were the most widely interrelated.

Mental illness, physical and sexual abuse associated with HIV risk behaviors among adults evaluated for substance use and treatment planning in the National Addictions Vigilance Intervention and Prevention Program — United States, 2014–2019

BackgroundThe association between sexual and physical abuse history, mental illness, and HIV risk behaviors among persons starting treatment for substance use is not well-understood. MethodsThe study population included 216,877 US residents in the National Addictions Vigilance Intervention and Prevention Program (NAVIPPRO) from January 1, 2014–December 31, 2019. We used logistic regression models to estimate strength of pairwise association between mental illness, sexual or physical abuse histories and each of 3 HIV risk outcomes. ResultsCompared with no history of physical or sexual abuse, a history of sexual and physical abuse combined was associated with greater odds ratios for: (1) injection drug use among persons without a history of mental illness (odds ratio [OR] 2.4; 95% confidence interval [CI]: 2.3–2.6) than among persons with a history of mental illness (OR 2.0; 95% CI: 1.9–2.0); (2) prostitution conviction among persons without mental illness (OR 3.8; 95% CI: 2.8–5.1) than among persons with mental illness (OR 2.8; 95% CI: 2.4–3.4); and (3) and ≥2 sex partners within the past 30 days with a history of mental illness (OR 1.3; 95% CI: 1.2–1.4). ConclusionsThe findings imply that efforts to reduce HIV risk behaviors during and after substance use treatment can be improved by considering the patient's history of physical or sexual abuse and mental illness when providing care.