Local Health Equity Action Teams (LHEATS) as a Novel and Emerging Practice of the Communities Organizing to Promote Equity (COPE) Project in Kansas.

The Communities Organizing to Promote Equity (COPE) Project was implemented in 20 counties across Kansas to build capacity to address health equity by forming local health equity action teams (LHEATS), hiring and training community health workers, facilitating state-wide learning collaboratives, and tailoring communication strategies. We conducted interviews and focus groups with project stakeholders who identified pragmatic recommendations related to LHEAT formation and leadership, establishing trust, nurturing autonomy, and optimizing impact. Insights can improve future community-based health equity efforts. (Am J Public Health. 2024;114(S7):S570-S574. https://doi.org/10.2105/AJPH.2024.307802).

Generative Artificial Intelligence and Large Language Models in Primary Care Medical Education.

Generative artificial intelligence and large language models are the continuation of a technological revolution in information processing that began with the invention of the transistor in 1947. These technologies, driven by transformer architectures for artificial neural networks, are poised to broadly influence society. It is already apparent that these technologies will be adapted to drive innovation in education. Medical education is a high-risk activity: Information that is incorrectly taught to a student may go unrecognized for years until a relevant clinical situation appears in which that error can lead to patient harm. In this article, I discuss the principal limitations to the use of generative artificial intelligence in medical education-hallucination, bias, cost, and security-and suggest some approaches to confronting these problems. Additionally, I identify the potential applications of generative artificial intelligence to medical education, including personalized instruction, simulation, feedback, evaluation, augmentation of qualitative research, and performance of critical assessment of the existing scientific literature.

Communities organizing to promote equity: engaging local communities in public health responses to health inequities exacerbated by COVID-19-protocol paper.

Background: The COVID-19 pandemic has disproportionately impacted rural and under-resourced urban communities in Kansas. The state's response to COVID-19 has relied on a highly decentralized and underfunded public health system, with 100 local health departments in the state, few of which had prior experience engaging local community coalitions in a coordinated response to a public health crisis. Methods: To improve the capacity for local community-driven responses to COVID-19 and other public health needs, the University of Kansas Medical Center, in partnership with the Kansas Department of Health and Environment, will launch Communities Organizing to Promote Equity (COPE) in 20 counties across Kansas. COPE will establish Local Health Equity Action Teams (LHEATs), coalitions comprised of community members and service providers, who work with COPE-hired community health workers (CHWs) recruited to represent the diversity of the communities they serve. CHWs in each county are tasked with addressing unmet social needs of residents and supporting their county's LHEAT. LHEATs are charged with implementing strategies to improve social determinants of health in their county. Monthly, LHEATs and CHWs from all 20 counties will come together as part of a learning collaborative to share strategies, foster innovation, and engage in peer problem-solving. These efforts will be supported by a multilevel communications strategy that will increase awareness of COPE activities and resources at the local level and successes across the state. Our mixed methods evaluation design will assess the processes and impact of COPE activities as well as barriers and facilitators to implementation using aspects of both the Consolidated Framework for Implementation Research (CFIR) and Reach, Effectiveness, Adoption, Implementation and Maintenance (RE-AIM) models. Discussion: This protocol is designed to expand community capacity to strategically partner with local public health and social service partners to prioritize and implement health equity efforts. COPE intentionally engages historically resilient communities and those living in underserved rural areas to inform pragmatic strategies to improve health equity.

Unawareness of partner risk factors thwarts implementation of USPSTF recommendations for HIV pre-exposure prophylaxis in primary care.

The United States Preventive Services Task Force recommends pre-exposure chemoprophylaxis for persons at high risk of HIV exposure. The optimal screening strategy for at-risk individuals in primary care is not known. We evaluated the strategy of universal screening and discussed challenges to the implementation of this recommendation in primary care. Around 430 of 500 (86%) screening surveys were completed. Mutual monogamy was common but monogamous partners with recent negative HIV testing were uncommon. Likewise, among heterosexually active men and women, inconsistent condom use was common. Such individuals would be on guideline for HIV pre-exposure prophylaxis (PrEP) if their partner was at risk for HIV exposure. None of these potentially at-risk individuals met the criteria for PrEP, but 13% lacked knowledge of their partners' sexual and behavioral risk factors, preventing a clear decision on whether PrEP should be used. Our screen identified no individuals who clearly met the guideline criteria for HIV PrEP. We conclude that universal screening for HIV PrEP in primary care is unlikely to be an efficient strategy; targeted screening is likely more appropriate. Moreover, the 2019 guidelines for heterosexually active men and women rely on information that is often unknown to the patient, which makes these guidelines difficult to implement. Future guidelines should address these problems.

Quality, Accuracy, and Bias in ChatGPT-Based Summarization of Medical Abstracts.

PURPOSE: Worldwide clinical knowledge is expanding rapidly, but physicians have sparse time to review scientific literature. Large language models (eg, Chat Generative Pretrained Transformer [ChatGPT]), might help summarize and prioritize research articles to review. However, large language models sometimes "hallucinate" incorrect information. METHODS: We evaluated ChatGPT's ability to summarize 140 peer-reviewed abstracts from 14 journals. Physicians rated the quality, accuracy, and bias of the ChatGPT summaries. We also compared human ratings of relevance to various areas of medicine to ChatGPT relevance ratings. RESULTS: ChatGPT produced summaries that were 70% shorter (mean abstract length of 2,438 characters decreased to 739 characters). Summaries were nevertheless rated as high quality (median score 90, interquartile range [IQR] 87.0-92.5; scale 0-100), high accuracy (median 92.5, IQR 89.0-95.0), and low bias (median 0, IQR 0-7.5). Serious inaccuracies and hallucinations were uncommon. Classification of the relevance of entire journals to various fields of medicine closely mirrored physician classifications (nonlinear standard error of the regression [SER] 8.6 on a scale of 0-100). However, relevance classification for individual articles was much more modest (SER 22.3). CONCLUSIONS: Summaries generated by ChatGPT were 70% shorter than mean abstract length and were characterized by high quality, high accuracy, and low bias. Conversely, ChatGPT had modest ability to classify the relevance of articles to medical specialties. We suggest that ChatGPT can help family physicians accelerate review of the scientific literature and have developed software (pyJournalWatch) to support this application. Life-critical medical decisions should remain based on full, critical, and thoughtful evaluation of the full text of research articles in context with clinical guidelines.

Assessing Social Needs and Engaging Community Health Workers in Underserved Kansas Counties: Insights From Primary Care Providers and Clinic Managers.

INTRODUCTION: Rural and under-resourced urban communities face unique challenges in addressing patients' social determinants of health needs (SDoH). Community health workers (CHWs) can support patients experiencing social needs, yet little is known about how rural and under-resourced primary care clinics are screening for SDoH or utilizing CHWs. METHODS: Interviews were conducted with primary care clinic providers and managers across a geographically large and predominately rural state to assess screening practices for SDoH and related community resources, and perspectives on using CHWs to address SDoH. Interviews were conducted by phone, recorded, and transcribed. Data were analyzed using thematic analysis. We completed interviews with 27 respondents (12 providers and 15 clinic managers) at 26 clinics. RESULTS: Twelve (46.1%) clinics had a standardized process for capturing SDoH, but this was primarily limited to Medicare wellness visits. Staffing and time were identified as barriers to proper SDoH screening. Lack of transportation and affordable medication were the most cited SDoH. While respondents were all aware of CHWs, only 8 (30.8%) included a CHW on their care team. Perceived barriers to engaging CHWs included cost, space, and availability of qualified CHWs. Perceived benefits of engaging CHWs in their practice were: assisting patients with navigating resources and programs, relieving clinical staff of non-medical tasks, and bridging language barriers. CONCLUSIONS: Rural and under-resourced primary care clinics need help in identifying and addressing SDoH. CHWs could play an important part in addressing social needs and promoting preventive care if financial constraints could be addressed and local CHWs could be trained.

Perceived Value of Public Health and Health Policy Information Among Medical Students: A Cross-sectional Study.

Introduction: Health educators have had difficulty introducing health policy and public health training into an already intensive medical school curriculum. Although the COVID-19 pandemic may have changed perspectives on the importance of public health, it may not change educational approaches. Assessment of medical student opinions and perceptions of health policy and public health might influence the weight given to these topics in medical education. Methods: We used a 39-item instrument to cross-sectionally survey medical students, to assess their perceptions of the value of public health and health policy within their professional education. Results: One hundred two students completed the survey (13% response rate). Most students reported an interest in public health (94.1%) and health policy (92.2%). Although interested, most students lacked confidence in their knowledge of public health and health policy on both state (health policy 34.3% confident; public health 43.1%) and national (health policy 41.0%; public health 44.1%) levels. Most students perceived that their institution has not sufficiently prepared them to understand health policy (34% felt prepared) and most reported insufficient information to participate in policy discussions (30.3% sufficiently informed). Conclusions: Medical students reported an interest in public health and health policy while also reporting a lack of confidence in their level of preparedness to understand and participate in these fields, thus demonstrating a need for increased public health and health policy education within medical school curricula.

PYK-SubstitutionOME: an integrated database containing allosteric coupling, ligand affinity and mutational, structural, pathological, bioinformatic and computational information about pyruvate kinase isozymes.

Interpreting changes in patient genomes, understanding how viruses evolve and engineering novel protein function all depend on accurately predicting the functional outcomes that arise from amino acid substitutions. To that end, the development of first-generation prediction algorithms was guided by historic experimental datasets. However, these datasets were heavily biased toward substitutions at positions that have not changed much throughout evolution (i.e. conserved). Although newer datasets include substitutions at positions that span a range of evolutionary conservation scores, these data are largely derived from assays that agglomerate multiple aspects of function. To facilitate predictions from the foundational chemical properties of proteins, large substitution databases with biochemical characterizations of function are needed. We report here a database derived from mutational, biochemical, bioinformatic, structural, pathological and computational studies of a highly studied protein family-pyruvate kinase (PYK). A centerpiece of this database is the biochemical characterization-including quantitative evaluation of allosteric regulation-of the changes that accompany substitutions at positions that sample the full conservation range observed in the PYK family. We have used these data to facilitate critical advances in the foundational studies of allosteric regulation and protein evolution and as rigorous benchmarks for testing protein predictions. We trust that the collected dataset will be useful for the broader scientific community in the further development of prediction algorithms. Database URL https://github.com/djparente/PYK-DB.

More Than Half of Family Medicine Clerkships Do Not Address Systemic Racism: A CERA Study.

BACKGROUND AND OBJECTIVES: The influence of racism in medicine is increasingly acknowledged, and the negative effect of systemic racism on individual and population health is well established. Yet, little is known about how or whether medical students are being educated on this topic. This study investigated the presence and features of curricula related to systemic racism in North American family medicine clerkships. METHODS: We conducted a survey of North American family medicine clerkship directors as part of the 2021 Council of Academic Family Medicine's Educational Research Alliance (CERA) survey. RESULTS: The survey response rate was 49% (78/160). Almost all clerkship directors agreed (n=68; 97.1%) that teaching about racism at all levels of medical education was appropriate. Yet, 60% (n=42) of family medicine clerkship directors reported no formalized curriculum for teaching about racism or bias. Teaching about systemic racism was more likely to be present in the family medicine clerkship at institutions where clerkship directors reported that faculty receive 5 or more hours of training in racism and bias, as compared to institutions where faculty receive 4 or fewer hours of training in racism/bias (odds ratio 2.82, 95% CI 1.05-8.04, P=.045). Programs reported using racism in medicine curricula based in cultural competency (20%); structural competency (10%); both cultural and structural competency (31%); and neither or uncertain (40%). Clerkship directors reported high faculty, student, and institutional engagement in addressing systemic racism. We did not find an association between underrepresented in medicine director identity and racism curricula. CONCLUSIONS: In more than half of family medicine clerkships, systemic racism is not addressed, despite interest from students and institutional support. A higher number of hours of faculty training time on the topic of racism was associated with having a systemic racism module in the clerkship curriculum, but we lacked data to identify a causal relationship. Investments in faculty development to teach systemic racism, including discussion of structural competency, are needed.

Machine Learning Prediction of Urine Cultures in Primary Care.

Context: Antibiotics for suspected urinary tract infection (UTI) is appropriate only when an infection is present. Urine culture is definitive but takes >1 day to result. A machine learning urine culture predictor was recently devised for Emergency Department (ED) patients but requires use of urine microscopy ("NeedMicro" predictor), which is not routinely available in primary care (PC). Objective: To adapt this predictor to use only features available in primary care and determine if predictive accuracy generalizes to the primary care setting. We call this the "NoMicro" predictor. Study Design and Analysis: Multicenter, retrospective, observational, cross-sectional analysis. Machine learning predictors were trained using extreme gradient boosting, artificial neural networks, and random forests. Models were trained on the ED dataset and were evaluated on both the ED dataset (internal validation) and the PC dataset (external validation). Setting: United States (US) academic medical centers emergency department and family medicine clinic. Population Studied: 80387 (ED, previously described) and 472 (PC, newly curated) US adults. Instrument: Physicians performed retrospective chart review. The primary outcome extracted was pathogenic urine culture growing ≥100,000 colony forming units. Predictor variables included age; gender; dipstick urinalysis nitrites, leukocytes, clarity, glucose, protein, and blood; dysuria; abdominal pain; and history of UTI. Outcome Measures: Predictor overall discriminative performance (receiver operating characteristic area under the curve, ROC-AUC), performance statistics (e.g., sensitivity, negative predictive value, etc.), and calibration. Results: The "NoMicro" model performs similarly to the "NeedMicro" model in internal validation on the ED dataset: NoMicro ROC-AUC 0.862 (95% CI: 0.856-0.869) vs. NeedMicro 0.877 (95% CI: 0.871-0.884). External validation on the primary care dataset also yielded high performance (NoMicro ROC-AUC 0.850 [95% CI: 0.808-0.889]), despite being trained on Emergency Department data. Simulation of a hypothetical, retrospective clinical trial suggests the NoMicro model could be used to avoid antibiotic overuse by safely withhold antibiotics in low-risk patients. Conclusions: The hypothesis that the NoMicro predictor generalizes to both PC and ED contexts is supported. Prospective trials to determine the real-world impact of using the NoMicro model to reduce antibiotic overuse are appropriate.

Critical gaps in knowledge and implementation of recommendations by the US Preventive Services Task Force.

Introduction: The United States Preventive Services Task Force (USPSTF) has issued 31 recommendations applicable to non-pregnant adults. We hypothesized variability in knowledge and implementation of these recommendations among US family medicine resident physicians. Methods: We performed two electronic surveys: a local survey, and then a nationally-representative, multicenter, survey. We evaluated self-reported knowledge and implementation of USPSTF recommendations related to non-pregnant adults. Results: 84 family medicine residents from 40 residency programs across 25 states participated. Knowledge and implementation of recommendations varied widely. Most residents lacked knowledge relating to breast cancer chemoprophylaxis (9.9 % "known in detail" or "mostly know"), BRCA-related genetic counseling (BRCA-GC) referral (30 %), tuberculosis (TB) screening (41 %), and sexually transmitted infection (STI) counseling (45 %). There is virtually no implementation of recommendations for breast cancer chemoprophylaxis (90 % never/rarely implement). Many residents never/rarely implement recommendations for BRCA-GC referral (75 %), TB screening (62 %), and HIV pre-exposure prophylaxis (61 %). This remained true even for residents in their final year of training. Relative to their male counterparts, female physicians more frequently implemented recommendations for BRCA-GC referral (11 % vs 0 % always/often implement, p = 0.019), cervical cancer screening (100 % vs 83 %, p = 0.019), and folic acid supplementation (60 % vs 29 %, p = 0.007). Knowledge and implementation of recommendations were strongly related (ß = 0.75, 95 % CI 0.50-1.00, p < 0.001, Spearman R2 = 0.56). Conclusion: Critical gaps exist in resident knowledge and implementation of USPSTF recommendations. We discuss urgent implications for cancer prevention, public health, and health equity.

Adaptation and External Validation of Pathogenic Urine Culture Prediction in Primary Care Using Machine Learning.

BACKGROUND: Urinary tract infection (UTI) symptoms are common in primary care, but antibiotics are appropriate only when an infection is present. Urine culture is the reference standard test for infection, but results take >1 day. A machine learning predictor of urine cultures showed high accuracy for an emergency department (ED) population but required urine microscopy features that are not routinely available in primary care (the NeedMicro classifier). METHODS: We redesigned a classifier (NoMicro) that does not depend on urine microscopy and retrospectively validated it internally (ED data set) and externally (on a newly curated primary care [PC] data set) using a multicenter approach including 80,387 (ED) and 472 (PC) adults. We constructed machine learning models using extreme gradient boosting (XGBoost), artificial neural networks, and random forests (RFs). The primary outcome was pathogenic urine culture growing ≥100,000 colony forming units. Predictor variables included age; gender; dipstick urinalysis nitrites, leukocytes, clarity, glucose, protein, and blood; dysuria; abdominal pain; and history of UTI. RESULTS: Removal of microscopy features did not severely compromise performance under internal validation: NoMicro/XGBoost receiver operating characteristic area under the curve (ROC-AUC) 0.86 (95% CI, 0.86-0.87) vs NeedMicro 0.88 (95% CI, 0.87-0.88). Excellent performance in external (PC) validation was also observed: NoMicro/RF ROC-AUC 0.85 (95% CI, 0.81-0.89). Retrospective simulation suggested that NoMicro/RF can be used to safely withhold antibiotics for low-risk patients, thereby avoiding antibiotic overuse. CONCLUSIONS: The NoMicro classifier appears appropriate for PC. Prospective trials to adjudicate the balance of benefits and harms of using the NoMicro classifier are appropriate.

Impact of the COVID-19 Pandemic on Exercise Habits Among US Primary Care Patients.

BACKGROUND: To explore how the COVID-19 pandemic has affected exercise habits, we hypothesized that participants' physical activity would have increased by at least 30 min/wk after the onset of the pandemic. METHODS: We distributed an anonymous survey to ambulatory patients at the Family Medicine Clinic, University of Kansas Medical Center to analyze changes in exercise habits and weight. RESULTS: Of the 500 adult patients surveyed, 382 were included. Results were stratified by demographics, including employment status before and during COVID-19. The median change in weekly exercise duration was 0.0 minutes, but the mean change was -25.7 minutes; total exercise duration decreased after the pandemic's onset (paired Wilcox signed rank test P < .001). More individuals reported participation in virtual group classes (6.3% before the pandemic vs 13.1% during the pandemic; McNemar's P < .001). Individuals with home exercise equipment before the pandemic were more likely to acquire more than were those who had none before (Chi square test P < .005). Overall, there is a significant trend in the direction of weight gain (Wilcox signed rank test P < .001). CONCLUSIONS: Most participants decreased physical activity during the unprecedented COVID-19 pandemic, expanding our understanding of how exercise habits change during stressful life events.

Association Between Unmet Essential Social Needs and Influenza Vaccination in US Adults.

BACKGROUND: Although social factors influence uptake of preventive services, the association between social needs and influenza vaccination has not been comprehensively evaluated for adults seeking primary care in the USA. OBJECTIVE: To determine the association between unmet social needs and influenza vaccination. DESIGN: Retrospective, cross-sectional, multivariable logistic regression. PARTICIPANTS: Persons completing ambulatory visits in a primary care department at a midwestern, urban, multispecialty, academic medical center between July 2017 and July 2019 (N = 7955 individuals included). MAIN MEASURES: Completion of influenza vaccination in the 2018-2019 influenza season (primary outcome) or any year (secondary outcome) against 11 essential social needs (childcare, companionship, food security, health literacy, home safety, neighborhood safety, housing, health care provider costs, prescription costs, transportation, and utilities). Demographics, diabetic status, COPD, smoking status, office visit frequency, and hierarchical condition category risk scores were included as covariates. KEY RESULTS: Individuals with transportation vulnerability were less likely to be vaccinated against influenza (current-year aOR 0.65, 95% CI: 0.53-0.78, p < 0.001; any-year aOR 0.58, 95% CI: 0.47-0.71, p < 0.001). Poor health literacy promoted any-year, but not current-year, influenza vaccination (any-year aOR 1.30, 95% CI: 1.01-1.69, p = 0.043). Older age, female sex, diabetes, more comorbidities, and more frequent primary care visits were associated with greater influenza vaccination. Persons with Black or other/multiple race and current smokers were less frequently vaccinated. CONCLUSIONS: Transportation vulnerability, health literacy, smoking, age, sex, race, comorbidity, and office visit frequency are associated with influenza vaccination. Primary care-led interventions should consider these factors when designing outreach interventions. TRIAL REGISTRATION: Not applicable.

Rheostat functional outcomes occur when substitutions are introduced at nonconserved positions that diverge with speciation.

When amino acids vary during evolution, the outcome can be functionally neutral or biologically-important. We previously found that substituting a subset of nonconserved positions, "rheostat" positions, can have surprising effects on protein function. Since changes at rheostat positions can facilitate functional evolution or cause disease, more examples are needed to understand their unique biophysical characteristics. Here, we explored whether "phylogenetic" patterns of change in multiple sequence alignments (such as positions with subfamily specific conservation) predict the locations of functional rheostat positions. To that end, we experimentally tested eight phylogenetic positions in human liver pyruvate kinase (hLPYK), using 10-15 substitutions per position and biochemical assays that yielded five functional parameters. Five positions were strongly rheostatic and three were non-neutral. To test the corollary that positions with low phylogenetic scores were not rheostat positions, we combined these phylogenetic positions with previously-identified hLPYK rheostat, "toggle" (most substitution abolished function), and "neutral" (all substitutions were like wild-type) positions. Despite representing 428 variants, this set of 33 positions was poorly statistically powered. Thus, we turned to the in vivo phenotypic dataset for E. coli lactose repressor protein (LacI), which comprised 12-13 substitutions at 329 positions and could be used to identify rheostat, toggle, and neutral positions. Combined hLPYK and LacI results show that positions with strong phylogenetic patterns of change are more likely to exhibit rheostat substitution outcomes than neutral or toggle outcomes. Furthermore, phylogenetic patterns were more successful at identifying rheostat positions than were co-evolutionary or eigenvector centrality measures of evolutionary change.

Acceptance of COVID-19 Vaccination Among Health System Personnel.

INTRODUCTION: One-third of the general public will not accept Coronavirus disease 2019 (COVID-19) vaccination but factors influencing vaccine acceptance among health care personnel (HCP) are not known. We investigated barriers and facilitators to vaccine acceptance within 3 months of regulatory approval (primary outcome) among adult employees and students at a tertiary-care, academic medical center. METHODS: We used a cross-sectional survey design with multivariable logistic regression. Covariates included age, gender, educational attainment, self-reported health status, concern about COVID-19, direct patient interaction, and prior influenza immunization. RESULTS: Of 18,250 eligible persons, 3,347 participated. Two in 5 (40.5%) HCP intend to delay (n = 1020; 30.6%) or forgo (n = 331; 9.9%) vaccination. Male sex (adjusted OR [aOR], 2.43; 95% confidence interval [CI], 2.00-2.95; P < .001), prior influenza vaccination (aOR, 2.35; 95% CI, 1.75-3.18; P < .001), increased concern about COVID-19 (aOR, 2.40; 95% CI, 2.07-2.79; P < .001), and postgraduate education (aOR, 1.41; 95% CI, 1.21-1.65; P < .001) - but not age, direct patient interaction, or self-reported overall health - were associated with vaccine acceptance in multivariable analysis. Barriers to vaccination included concerns about long-term side effects (n = 1197, 57.1%), safety (n = 1152, 55.0%), efficacy (n = 777, 37.1%), risk-to-benefit ratio (n = 650, 31.0%), and cost (n = 255, 12.2%).Subgroup analysis of Black respondents indicates greater hesitancy to accept vaccination (only 24.8% within 3 months; aOR 0.13; 95% CI, 0.08-0.21; P < .001). CONCLUSIONS: Many HCP intend to delay or refuse COVID-19 vaccination. Policymakers should impartially address concerns about safety, efficacy, side effects, risk-to-benefit ratio, and cost. Further research with minority subgroups is urgently needed.

PolyBoost: An enhanced genomic variant classifier using extreme gradient boosting.

PURPOSE: Human exome sequences contain 15,000-20,000 variants but many variants have unknown clinical impact. In silico predictive classifiers are recognized by the American College of Medical Genetics as a resource for interpreting these "variants of uncertain significance." Many in silico classifiers have been developed, of which PolyPhen-2 is highly successful and widely used. PolyPhen-2 uses a naïve Bayes model to synthesize sequence, structural and genomic information. I investigated whether predictive performance could be improved by replacing PolyPhen-2's naïve Bayes model with alternative machine learning methods. EXPERIMENTAL DESIGN: Classifiers using the PolyPhen-2 feature set were retrained using extreme gradient boosting (XGBoost), random forests, artificial neural networks, and support vector machines. Classifiers were externally validated on "pathogenic" and "benign" ClinVar variants absent from the training datasets. Software is implemented in Python and is freely available at https://github.com/djparente/polyboost and the Python Package Index (PyPI) under the BSD license. RESULTS: An XGBoost-based classifier-designated PolyBoost (PolyPhen-2 Booster)-improves discriminative performance and calibration relative to PolyPhen-2 in external validation on ClinVar. CONCLUSIONS AND CLINICAL RELEVANCE: PolyBoost analyzes PolyPhen-2 output and can be incorporated into existing bioinformatics workflows as a post-analysis method to improve interpretation of clinical exome sequences obtained to identify monogenic disease.

BRCA-Related Cancer Genetic Counseling is Indicated in Many Women Seeking Primary Care.

BACKGROUND: Guidelines updated by the United States Preventive Services Task Force (USPSTF) in 2019 recommend referral to genetic counseling for asymptomatic women that have a family history of cancers potentially associated with variants in the breast cancer type 1 and 2 susceptibility genes (BRCA1 and BRCA2). METHODS: I performed a needs assessment for BRCA-related cancer genetic counseling among undifferentiated women seeking primary care at an urban, academic medical center with an underserved population. Adult, English-speaking women with outpatient primary care appointments were surveyed. Questions included personal and family history of potentially BRCA-related malignancies, history of genetic counseling and/or testing, and a version of the USPSTF-recommended 7-Question Family History Screening (FHS-7) tool, modified to promote accessibility among women with low health literacy. RESULTS: Out of 397 women, 97 women (24.4% ± 4.2%, 95% CI) met criteria for referral to genetic counseling. Among women with referral indications, 80 women (82.4% ± 7.6%) had no prior contact with genetic counseling and/or testing services (comprising 20.1% ± 3.9% of all women surveyed). The most common indication for BRCA-related genetic counseling referral was family history of female breast cancer before age 50 years. CONCLUSIONS: The rate that undifferentiated women seeking primary care met 2019 USPSTF criteria for BRCA-related cancer genetic counseling referral (24.4% ± 4.2%) exceeds earlier estimates (4 to 5%) but agrees with later, population-level estimates (24.1%). Health systems will need to appropriately allocate capacity to genetic counseling services and/or reconsider the appropriateness of FHS-7 as a primary care risk-stratification tool.