Large-scale identification of social and behavioral determinants of health from clinical notes: comparison of Latent Semantic Indexing and Generative Pretrained Transformer (GPT) models.

BACKGROUND: Social and behavioral determinants of health (SBDH) are associated with a variety of health and utilization outcomes, yet these factors are not routinely documented in the structured fields of electronic health records (EHR). The objective of this study was to evaluate different machine learning approaches for detection of SBDH from the unstructured clinical notes in the EHR. METHODS: Latent Semantic Indexing (LSI) was applied to 2,083,180 clinical notes corresponding to 46,146 patients in the MIMIC-III dataset. Using LSI, patients were ranked based on conceptual relevance to a set of keywords (lexicons) pertaining to 15 different SBDH categories. For Generative Pretrained Transformer (GPT) models, API requests were made with a Python script to connect to the OpenAI services in Azure, using gpt-3.5-turbo-1106 and gpt-4-1106-preview models. Prediction of SBDH categories were performed using a logistic regression model that included age, gender, race and SBDH ICD-9 codes. RESULTS: LSI retrieved patients according to 15 SBDH domains, with an overall average PPV ≥ 83%. Using manually curated gold standard (GS) sets for nine SBDH categories, the macro-F1 score of LSI (0.74) was better than ICD-9 (0.71) and GPT-3.5 (0.54), but lower than GPT-4 (0.80). Due to document size limitations, only a subset of the GS cases could be processed by GPT-3.5 (55.8%) and GPT-4 (94.2%), compared to LSI (100%). Using common GS subsets for nine different SBDH categories, the macro-F1 of ICD-9 combined with either LSI (mean 0.88, 95% CI 0.82-0.93), GPT-3.5 (0.86, 0.82-0.91) or GPT-4 (0.88, 0.83-0.94) was not significantly different. After including age, gender, race and ICD-9 in a logistic regression model, the AUC for prediction of six out of the nine SBDH categories was higher for LSI compared to GPT-4.0. CONCLUSIONS: These results demonstrate that the LSI approach performs comparable to more recent large language models, such as GPT-3.5 and GPT-4.0, when using the same set of documents. Importantly, LSI is robust, deterministic, and does not have document-size limitations or cost implications, which make it more amenable to real-world applications in health systems.

The BabySeq Project: A clinical trial of genome sequencing in a diverse cohort of infants.

Efforts to implement and evaluate genome sequencing (GS) as a screening tool for newborns and infants are expanding worldwide. The first iteration of the BabySeq Project (2015-2019), a randomized controlled trial of newborn sequencing, produced novel evidence on medical, behavioral, and economic outcomes. The second iteration of BabySeq, which began participant recruitment in January 2023, examines GS outcomes in a larger, more diverse cohort of more than 500 infants up to one year of age recruited from pediatric clinics at several sites across the United States. The trial aims for families who self-identify as Black/African American or Hispanic/Latino to make up more than 50% of final enrollment, and key aspects of the trial design were co-developed with a community advisory board. All enrolled families receive genetic counseling and a family history report. Half of enrolled infants are randomized to receive GS with comprehensive interpretation of pathogenic and likely pathogenic variants in more than 4,300 genes associated with childhood-onset and actionable adult-onset conditions, as well as larger-scale chromosomal copy number variants classified as pathogenic or likely pathogenic. GS result reports include variants associated with disease (Mendelian disease risks) and carrier status of autosomal-recessive and X-linked disorders. Investigators evaluate the utility and impacts of implementing a GS screening program in a diverse cohort of infants using medical record review and longitudinal parent surveys. In this perspective, we describe the rationale for the second iteration of the BabySeq Project, the outcomes being assessed, and the key decisions collaboratively made by the study team and community advisory board.

Influenza vaccine effectiveness against hospitalized SARS-CoV-2 infection.

Background: Some studies conducted before the Delta and Omicron variant-dominant periods have indicated that influenza vaccination provided protection against COVID-19 infection or hospitalization, but these results were limited by small study cohorts and a lack of comprehensive data on patient characteristics. No studies have examined this question during the Delta and Omicron periods (08/01/2021 to 2/22/2022). Methods: We conducted a retrospective cohort study of influenza-vaccinated and unvaccinated patients in the Corewell Health East(CHE, formerly known as Beaumont Health), Corewell Health West(CHW, formerly known as Spectrum Health) and Michigan Medicine (MM) healthcare system during the Delta-dominant and Omicron-dominant periods. We used a test-negative, case-control analysis to assess the effectiveness of the influenza vaccine against hospitalized SARS-CoV-2 outcome in adults, while controlling for individual characteristics as well as pandameic severity and waning immunity of COVID-19 vaccine. Results: The influenza vaccination has shown to provided some protection against SARS-CoV-2 hospitalized outcome across three main healthcare systems. CHE site (odds ratio [OR]=0.73, vaccine effectiveness [VE]=27%, 95% confidence interval [CI]: [18-35], p<0.001), CHW site (OR=0.85, VE=15%, 95% CI: [6-24], p<0.001), MM (OR=0.50, VE=50%, 95% CI: [40-58], p <0.001) and overall (OR=0.75, VE=25%, 95% CI: [20-30], p <0.001). Conclusion: The influenza vaccine provides a small degree of protection against SARS-CoV-2 infection across our study sites.

Paradoxical experiences of healthcare workers during COVID-19: a qualitative analysis of anonymous, web-based, audio narratives.

PURPOSE: To gain a deeper understanding of healthcare workers experiences during COVID-19 using an anonymous, web-based, audio narrative platform. METHODS: Data were collected from healthcare workers in the midwestern United States using a web-enabled audio diary approach. Participant recordings were analysed using a narrative coding and conceptualization process derived from grounded theory coding techniques. RESULTS: Fifteen healthcare workers, in direct patient care or non-patient care roles, submitted 18 audio narratives. Two paradoxical themes emerged: 1) A paradox of distress and meaningfulness, where a harsh work environment resulted in psychological distress while simultaneously resulting in new rewarding experiences, sense of purpose and positive outlooks. 2) A paradox of social isolation and connection, where despite extreme isolation, healthcare workers formed intense and meaningful interpersonal connections with patients and colleagues in new ways. CONCLUSIONS: A web-enabled audio diary approach provided an opportunity for healthcare workers to reflect deeper on their experiences without investigator influence, which led to some unique findings. Paradoxically, amid social isolation and extreme distress, a sense of value, meaning and rewarding human connections emerged. These findings suggest that interventions addressing healthcare worker burnout and distress might be enhanced by leveraging naturally occurring positive experiences as much as mitigating negative ones.

Risk factors associated with glycated hemoglobin A1c trajectories progressing to type 2 diabetes.

BACKGROUND AND OBJECTIVE: The notion of prediabetes, defined by the ADA as glycated hemoglobin A1c (HbA1c) of 5.7-6.4%, implies increased vascular inflammatory and immunologic processes and higher risk for developing diabetes mellitus and major cardiovascular events. We aimed to determine the risk factors associated with rapid progression of normal and prediabetes patients to type 2 diabetes mellitus (T2DM). METHODS: Retrospective cohort study in a single 8-hospital health system in southeast Michigan, between 2006 and 2020. All patients with HbA1c <6.5% at baseline and at least 2 other HbA1c measurements were clustered in five trajectories encompassing more than 95% of the study population. Multivariate linear regression analysis was performed to examine the association of demographic and comorbidities with HbA1c trajectories progressing to diabetes. RESULTS: A total of 5,347 prediabetic patients were clustered based on their HbA1c progression (C1: 4,853, C2: 253, C66: 102, C12: 85, C68: 54). The largest cluster (C1) had a baseline median HbA1c value of 6.0% and exhibited stable HbA1c levels in prediabetic range across all subsequent years. The smallest cluster (C68) had the lowest median baseline HbA1c value and also remained stable across subsequent years. The proportion of normal HbA1c in each of the pre-diabetic trajectories ranged from 0 to 12.7%, whereas 81.5% of the reference cluster (C68) were normal HbA1c at baseline. The C2 (steady rising) trajectory was significantly associated with BMI (adj OR 1.10, 95%CI 1.03-1.17), and family history of DM (adj OR 2.75, 95%CI 1.32-5.74). With respect to the late rising trajectories, baseline BMI was significantly associated with both C66 and C12 trajectory (adj OR 1.10, 95%CI 1.03-1.18) and (adj OR 1.13, 95%CI 1.05-1.23) respectively, whereas, the C12 trajectory was also significantly associated with age (adj OR 1.62, 95%CI 1.04-2.53) and history of MACE (adj OR 3.20, 95%CI 1.14-8.93). CONCLUSIONS: We suggest that perhaps a more aggressive preventative approach should be considered in patients with a family history of T2DM who have high BMI and year-to-year increase in HbA1c, whether they have normal hemoglobin A1c or they have prediabetes.KEY MESSAGESProgression to diabetes from normal or prediabetic hemoglobin A1c within four years is associated with baseline BMI.A steady rise in HbA1c during a four-year period is associated with age and family history of T2DM, whereas age and personal history of MACE are associated with a rapid rise in HbA1c.A more aggressive preventative approach should be considered in patients with a family history of T2DM who have high BMI and year-to-year increase in HbA1c.

Dried blood spots are a valid alternative to venipuncture for COVID-19 antibody testing.

BACKGROUND: Serologic analysis is an important tool towards assessing the humoral response to COVID-19 infection and vaccination. Numerous serologic tests and platforms are currently available to support this line of testing. Two broad antibody testing categories are point-of-care lateral flow immunoassays and semi-quantitative immunoassays performed in clinical laboratories, which typically require blood collected from a finger-stick and a standard venipuncture blood draw, respectively. This study evaluated the use of dried blood spot (DBS) collections as a sample source for COVID-19 antibody testing using an automated clinical laboratory test system. METHODS: Two hundred and ninety-four participants in the BLAST COVID-19 seroprevalence study (NCT04349202) were recruited at the time of a scheduled blood draw to have an additional sample taken via finger stick as a DBS collection. Using the EUROIMMUN assay to assess SARS-CoV-2 anti-spike IgG status, DBS specimens were tested on 7, 14, 21, and 28 days post- collection and compared to the reference serum sample obtained from a blood draw for the BLAST COVID-19 study. RESULTS: SARS-CoV-2 anti-spike IgG status from DBS collections demonstrated high concordance with serum across all time points (7-28 days). However, the semi-quantitative value from DBS collections was lower on average than that from serum, resulting in increased uncertainty around the equivocal-to-positive analytical decision point. CONCLUSIONS: DBS collections can be substituted for venipuncture when assaying for COVID-19 IgG antibody, with samples being stable for at least 28 days at room temperature. Finger-stick sampling can therefore be advantageous for testing large populations for SARS-CoV-2 antibodies without the need for phlebotomists or immediate processing of samples. We have high confidence in serostaus determination from DBS collections, although the reduced semi-quantitative value may cause some low-level positives to fall into the equivocal or even negative range.

NIPSNAP protein family emerges as a sensor of mitochondrial health.

Since their discovery over two decades ago, the molecular and cellular functions of the NIPSNAP family of proteins (NIPSNAPs) have remained elusive until recently. NIPSNAPs interact with a variety of mitochondrial and cytoplasmic proteins. They have been implicated in multiple cellular processes and associated with different physiologic and pathologic conditions, including pain transmission, Parkinson's disease, and cancer. Recent evidence demonstrated a direct role for NIPSNAP1 and NIPSNAP2 proteins in regulation of mitophagy, a process that is critical for cellular health and maintenance. Importantly, NIPSNAPs contain a 110 amino acid domain that is evolutionary conserved from mammals to bacteria. However, the molecular function of the conserved NIPSNAP domain and its potential role in mitophagy have not been explored. It stands to reason that the highly conserved NIPSNAP domain interacts with a substrate that is ubiquitously present across all species and can perhaps act as a sensor for mitochondrial health.

Coronavirus Disease 2019 (COVID-19) Seropositivity and Asymptomatic Rates in Healthcare Workers Are Associated with Job Function and Masking.

BACKGROUND: Although the risk of exposure to Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) is higher for frontline healthcare workers, not all personnel have similar risks. Determining infection rate is difficult due to the limits on testing and the high rate of asymptomatic individuals. Detection of antibodies against SARS-CoV-2 may be useful for determining prior exposure to the virus and assessing mitigation strategies, such as isolation, masks, and other protective equipment. METHODS: An online assessment that included demographic, clinical, and exposure information and a blood sample was collected from 20 614 participants out of ~43 000 total employees at Beaumont Health, which includes 8 hospitals distributed across the Detroit metropolitan area in southeast Michigan. The presence of anti-SARS-CoV-2 IgG was determined using the EUROIMMUN assay. RESULTS: A total of 1818 (8.8%) participants were seropositive between April 13 and May 28, 2020. Among the seropositive individuals, 44% reported that they were asymptomatic during the month prior to blood collection. Healthcare roles such as phlebotomy, respiratory therapy, and nursing/nursing support exhibited significantly higher seropositivity. Among participants reporting direct exposure to a Coronavirus Disease 2019 (COVID-19) positive individual, those wearing an N95/PAPR mask had a significantly lower seropositivity rate (10.2%) compared to surgical/other masks (13.1%) or no mask (17.5%). CONCLUSIONS: Direct contact with COVID-19 patients increased the likelihood of seropositivity among employees but study participants who wore a mask during COVID-19 exposures were less likely to be seropositive. Additionally, a large proportion of seropositive employees self-reported as asymptomatic. (Funded by Beaumont Health and by major donors through the Beaumont Health Foundation). CLINICALTRIALS.GOV NUMBER: NCT04349202.

Coordinated regulation of Rel expression by MAP3K4, CBP, and HDAC6 controls phenotypic switching.

Coordinated gene expression is required for phenotypic switching between epithelial and mesenchymal phenotypes during normal development and in disease states. Trophoblast stem (TS) cells undergo epithelial-mesenchymal transition (EMT) during implantation and placentation. Mechanisms coordinating gene expression during these processes are poorly understood. We have previously demonstrated that MAP3K4-regulated chromatin modifiers CBP and HDAC6 each regulate thousands of genes during EMT in TS cells. Here we show that CBP and HDAC6 coordinate expression of only 183 genes predicted to be critical regulators of phenotypic switching. The highest-ranking co-regulated gene is the NF-κB family member Rel. Although NF-κB is primarily regulated post-transcriptionally, CBP and HDAC6 control Rel transcript levels by binding Rel regulatory regions and controlling histone acetylation. REL re-expression in mesenchymal-like TS cells induces a mesenchymal-epithelial transition. Importantly, REL forms a feedback loop, blocking HDAC6 expression and nuclear localization. Together, our work defines a developmental program coordinating phenotypic switching.

Development of the InCharge Health Mobile App to Improve Adherence to Hydroxyurea in Patients With Sickle Cell Disease: User-Centered Design Approach.

BACKGROUND: Sickle cell disease (SCD) is an inherited blood disorder causing acute complications and chronic progressive end organ damage. SCD is associated with significant morbidity, early mortality, impaired health-related quality of life, and increased acute health care utilization. Hydroxyurea is a US Food and Drug Administration-approved medication that reduces disease complications, acute health care utilization, and costs. However, adherence to hydroxyurea is suboptimal. Mobile health (mHealth) interventions have the potential to improve hydroxyurea adherence, but few examples exist that are specific to the SCD population. OBJECTIVE: This study aimed to design a mHealth intervention for individuals with SCD to improve adherence to hydroxyurea, using a user-centered design that was informed by specific barriers to hydroxyurea adherence and utilization in this population. METHODS: This study consisted of 4 phases. In phase 1, individuals with SCD and health care providers participated in an optimization digital workshop. In phase 2, patients completed surveys pertaining to their interest in mHealth use, barriers and facilitators to hydroxyurea use, and health literacy. Phases 3 and 4 involved semistructured interviews and focus groups, respectively, and used the Health Belief Model (HBM) as the framework to investigate drivers of poor hydroxyurea adherence and to inform the development of an app prototype. In addition, in phase 4, we have incorporated the patients' feedback on the preliminary app prototype and its features. RESULTS: Barriers to hydroxyurea adherence were consistent with the literature and included forgetfulness and several specific thoughts and emotions associated with hydroxyurea use (eg, fear of side effects, depression, stigma, and hopelessness). In addition, more than half of the participants reported potentially low health literacy. Preferred patient app features included 7 key components, namely (1) medication reminders and tracker, (2) disease education, (3) communication, (4) personalization, (5) motivation, (6) support during pain episodes, and (7) social support. Utilizing a user-centered design approach, data obtained from patients and providers were translated into features within the app, mapping to components of the HBM and the specific drivers of hydroxyurea adherence and matching the literacy level of the population, resulting in the development of a novel mobile app called InCharge Health. CONCLUSIONS: The InCharge Health app is an mHealth intervention developed with substantial input from users and by mapping the HBM as the framework that guided the choice for its components. InCharge Health is a customized product for the SCD population aimed at optimizing medication adherence, with the end goal of improving quality of life and health outcomes among patients with SCD. The efficacy and implementation of the InCharge Health app as an mHealth intervention to promote hydroxyurea adherence will be tested in a future stepped-wedge multicenter trial for adolescents and adults with SCD.

Evaluation of Sirtuin-3 probe quality and co-expressed genes using literature cohesion.

BACKGROUND: Gene co-expression studies can provide important insights into molecular and cellular signaling pathways. The GeneNetwork database is a unique resource for co-expression analysis using data from a variety of tissues across genetically distinct inbred mice. However, extraction of biologically meaningful co-expressed gene sets is challenging due to variability in microarray platforms, probe quality, normalization methods, and confounding biological factors. In this study, we tested whether literature derived functional cohesion could be used as an objective metric in lieu of 'ground truth' to evaluate the quality of probes and microarray datasets. RESULTS: We examined Sirtuin-3 (Sirt3) co-expressed gene sets extracted from either liver or brain tissues of BXD recombinant inbred mice in the GeneNetwork database. Depending on the microarray platform, there were as many as 26 probes that targeted different regions of Sirt3 primary transcript. Co-expressed gene sets (ranging from 100-1000 genes) associated with each Sirt3 probe were evaluated using the previously developed literature-derived cohesion p-value (LPv) and benchmarked against 'gold standards' derived from proteomic studies or Gene Ontology classifications. We found that the maximal F-measure was obtained at an average window size of 535 genes. Using set size of 500 genes, the Pearson correlations between LPv and F-measure as well as between LPv and mitochondrial gene enrichment p-values were 0.90 and 0.93, respectively. Importantly, we found that the LPv approach can distinguish high quality Sirt3 probes. Analysis of the most functionally cohesive Sirt3 co-expressed gene set revealed core metabolic pathways that were shared between hippocampus and liver as well as distinct pathways which were unique to each tissue. These results are consistent with other studies that suggest Sirt3 is a key metabolic regulator and has distinct functions in energy-producing vs. energy-demanding tissues. CONCLUSIONS: Our results provide proof-of-concept that literature cohesion analysis is useful for evaluating the quality of probes and microarray datasets, particularly when experimentally derived gold standards are unavailable. Our approach would enable researchers to rapidly identify biologically meaningful co-expressed gene sets and facilitate discovery from high throughput genomic data.

Memphis FitKids: implementing a mobile-friendly web-based application to enhance parents' participation in improving child health.

BACKGROUND: Child obesity is a major public health challenge, increasing the risk of chronic medical conditions such as type 2 diabetes, metabolic syndrome, and hypertension. Among U.S. states, Tennessee has one of the highest rates of child obesity. Emerging communication technologies can help to deliver highly disseminable population-level interventions to improve health behavior. The aim of this paper is to report the implementation and the evaluation of the reach of Memphis FitKids, a web-based application, intended to promote healthy behaviors for families and children. METHODS: A community-level demonstration project, Memphis FitKids, was developed and implemented in Tennessee's Greater Memphis Area. This application ( www.memphisfitkids.org ) was designed for parents to assess their children's obesity risk through determinants such as weight, diet, physical activity, screen time, and sleep adequacy. A built-in "FitCheck" tool used this collected information to create a report with tailored recommendations on how to make healthy changes. A Geographic Information Systems component was implemented to suggest low-cost neighborhood resources that support a healthy lifestyle. A social marketing framework was used to develop and implement FitKids, and a Community Advisory Board with representatives from community partners (e.g., the YMCA of Memphis, the Pink Palace Family of Museums, and the Memphis Public Library) supported the implementation of the project. Five kiosks distributed in the community served as public access points to provide a broad reach across socioeconomic strata. Presentations at community events and the use of Facebook facilitated the promotion of FitKids. Website traffic and Facebook usage were evaluated with Google Analytics and Facebook Insights, respectively. RESULTS: In Tennessee, 33,505 users completed 38,429 FitCheck sessions between July 2014 and December 2016. Among these, 6763 sessions were completed at the five kiosks in the community. FitKids was presented at 112 community events and the social media posts reached 23,767 unique Facebook users. CONCLUSIONS: The Memphis FitKids demonstration project showed that web-based health tools may be a viable strategy to increase access to information about healthy weight and lifestyle options for families. Mobile-friendly web-based applications like Memphis FitKids may also serve health professionals in their efforts to support their clients in adopting healthy behaviors.

Navigating the Functional Landscape of Transcription Factors via Non-Negative Tensor Factorization Analysis of MEDLINE Abstracts.

In this study, we developed and evaluated a novel text-mining approach, using non-negative tensor factorization (NTF), to simultaneously extract and functionally annotate transcriptional modules consisting of sets of genes, transcription factors (TFs), and terms from MEDLINE abstracts. A sparse 3-mode term × gene × TF tensor was constructed that contained weighted frequencies of 106,895 terms in 26,781 abstracts shared among 7,695 genes and 994 TFs. The tensor was decomposed into sub-tensors using non-negative tensor factorization (NTF) across 16 different approximation ranks. Dominant entries of each of 2,861 sub-tensors were extracted to form term-gene-TF annotated transcriptional modules (ATMs). More than 94% of the ATMs were found to be enriched in at least one KEGG pathway or GO category, suggesting that the ATMs are functionally relevant. One advantage of this method is that it can discover potentially new gene-TF associations from the literature. Using a set of microarray and ChIP-Seq datasets as gold standard, we show that the precision of our method for predicting gene-TF associations is significantly higher than chance. In addition, we demonstrate that the terms in each ATM can be used to suggest new GO classifications to genes and TFs. Taken together, our results indicate that NTF is useful for simultaneous extraction and functional annotation of transcriptional regulatory networks from unstructured text, as well as for literature based discovery. A web tool called Transcriptional Regulatory Modules Extracted from Literature (TREMEL), available at http://binf1.memphis.edu/tremel, was built to enable browsing and searching of ATMs.

Loss of C-5 Sterol Desaturase Activity Results in Increased Resistance to Azole and Echinocandin Antifungals in a Clinical Isolate of Candida parapsilosis.

Among emerging non-albicans Candida species, Candida parapsilosis is of particular concern as a cause of nosocomial bloodstream infections in neonatal and intensive care unit patients. While fluconazole and echinocandins are considered effective treatments for such infections, recent reports of fluconazole and echinocandin resistance in C. parapsilosis indicate a growing problem. The present study describes a novel mechanism of antifungal resistance in this organism affecting susceptibility to azole and echinocandin antifungals in a clinical isolate obtained from a patient with prosthetic valve endocarditis. Transcriptome analysis indicated differential expression of several genes in the resistant isolate, including upregulation of ergosterol biosynthesis pathway genes ERG2, ERG5, ERG6, ERG11, ERG24, ERG25, and UPC2 Whole-genome sequencing revealed that the resistant isolate possessed an ERG3 mutation resulting in a G111R amino acid substitution. Sterol profiles indicated a reduction in sterol desaturase activity as a result of this mutation. Replacement of both mutant alleles in the resistant isolate with the susceptible isolate's allele restored wild-type susceptibility to all azoles and echinocandins tested. Disruption of ERG3 in the susceptible and resistant isolates resulted in a loss of sterol desaturase activity, high-level azole resistance, and an echinocandin-intermediate to -resistant phenotype. While disruption of ERG3 in C. albicans resulted in azole resistance, echinocandin MICs, while elevated, remained within the susceptible range. This work demonstrates that the G111R substitution in Erg3 is wholly responsible for the altered azole and echinocandin susceptibilities observed in this C. parapsilosis isolate and is the first report of an ERG3 mutation influencing susceptibility to the echinocandins.