Precision Health Research Articles

Protocol for a randomized controlled trial of the Enhanced Smoking Cessation Approach to Promote Empowerment (ESCAPE) digitalized intervention to promote lung health in high-risk individuals who smoke

Low dose computed tomography (LDCT) is an effective screening test to decrease lung cancer deaths. Lung cancer screening may be a teachable moment helping people who smoke to quit, which may result in increased benefit of screening. Innovative strategies are needed to engage high-risk individuals in learning about LDCT screening. More precise methods such as polygenic risk scores quantify genetic predisposition to tobacco use, and optimize lung health interventions. We present the ESCAPE (Enhanced Smoking Cessation Approach to Promote Empowerment) protocol. This study will test a smoking cessation intervention using personal stories and a lung cancer screening decision-aide compared to standard care (brief advice, referral to a quit line, and a lung cancer screening decision-aide), examine the relationship between a polygenic risk score and smoking abstinence, and describe perceptions about integration of genomic information into smoking cessation treatment. A randomized controlled trial followed by a sequential explanatory mixed methods approach will compare the efficacy of the interventions. Interviews will add insight into the use of genomic information and risk perceptions to tailor smoking cessation treatment. Two-hundred and fifty individuals will be recruited from primary care, community-based organizations, mailing lists and through social media. Data will be collected at baseline, 1, 3 and 6-months. The primary outcomes are 7-day point prevalence smoking abstinence and stage of lung cancer screening at 6-months. The results from this study will provide information to refine the ESCAPE intervention and facilitate integration of precision health into future lung health interventions.Clinical trial registration number: NCT0469129T.

Managing the impact of inter-method bias of prostate specific antigen assays on biopsy referral: the key to move towards precision health in prostate cancer management.

We assessed the inter-method bias of total (tPSA) and free (fPSA) prostate-specific antigen (PSA) immunoassays to establish if tPSA-based risk thresholds for advanced prostate cancer (PCa), obtained from one method (Roche) can be converted into the corresponding concentrations assayed by other methods. Then we evaluated the impact of the bias of tPSA and fPSA on theestimation of the %f/tPSA ratio and performed a re-calibration of the proposed thresholds for the %f/tPSA ratio according to the assay used. tPSA and fPSA were measured in 135 and 137 serum samples, respectively by Abbott Alinity i, Beckman Access Dxl, Roche Cobas e801, and Siemens Atellica IManalytical platforms. Scatterplots, Bland-Altman diagrams, Passing-Bablok (PB) were used to inspect and estimate the systematic and proportional bias between the methods. The linear equations with confidence intervals of the parameter estimates were used to transform the tPSA risk thresholds for advanced PCa into the corresponding concentrations measurable by the other analytical methods. To construct a correction coefficient for converting the %f/tPSA ratio from one method to the other, PB and non-parametric boostrapping were used. The inter-method bias is not constant but strictly linear allowing the conversion of PSA results obtained from Roche into the other assays, which underestimate tPSA vs. Roche. Siemens and Abbott vs. Roche and Beckman assays, being characterized by a positive and a negative proportional bias for tPSA and fPSA measurements, tend to overestimate the %f/tPSA ratio. There is a consistent risk to miss advanced PCa, if appropriate conversion factors are not applied.

When the clock ticks wrong with COVID-19.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a member of the coronavirus family that causes the novel coronavirus disease first diagnosed in 2019 (COVID-19). Although many studies have been carried out in recent months to determine why the disease clinical presentations and outcomes can vary significantly from asymptomatic to severe or lethal, the underlying mechanisms are not fully understood. It is likely that unique individual characteristics can strongly influence the broad disease variability; thus, tailored diagnostic and therapeutic approaches are needed to improve clinical outcomes. The circadian clock is a critical regulatory mechanism orchestrating major physiological and pathological processes. It is generally accepted that more than half of the cell-specific genes in any given organ are under circadian control. Although it is known that a specific role of the circadian clock is to coordinate the immune system's steady-state function and response to infectious threats, the links between the circadian clock and SARS-CoV-2 infection are only now emerging. How inter-individual variability of the circadian profile and its dysregulation may play a role in the differences noted in the COVID-19-related disease presentations, and outcome remains largely underinvestigated. This review summarizes the current evidence on the potential links between circadian clock dysregulation and SARS-CoV-2 infection susceptibility, disease presentation and progression, and clinical outcomes. Further research in this area may contribute towards novel circadian-centred prognostic, diagnostic and therapeutic approaches for COVID-19 in the era of precision health.

Interview: insights from a career researching the protein corona and combating academic bullying.

M Mahmoudi is an assistant professor in the Department of Radiology and Precision Health Program at Michigan State University (MI, USA). His research group has three broad lines of enquiry: nanomedicine, regenerative medicine and academic bullying and harassment. In the field of nanomedicine, the lab focuses on understanding the protein corona - the mix of biomolecules that binds to the surface of a nanoparticle when it interacts with biological fluids - and the complications this brings to reproducibility and data interpretation in the field of nanomedicine. In regenerative medicine, his lab works on cardiac regeneration and wound healing. His lab is also very active in the social sciences, specifically in the fields of gender disparity in the sciences and academic harassment. In addition to his academic postings, MMahmoudi is a cofounder and director of the Academic Parity Movement (a nonprofit organization); a cofounder of NanoServ, Targets' Tip and Partners in Global Wound Care; and a member of the Nanomedicine editorial board.

The metabolomics of human aging: Advances, challenges, and opportunities.

As the global population becomes older, understanding the impact of aging on health and disease becomes paramount. Recent advancements in multiomic technology have allowed for the high-throughput molecular characterization of aging at the population level. Metabolomics studies that analyze the small molecules in the body can provide biological information across a diversity of aging processes. Here, we review the growing body of population-scale metabolomics research on aging in humans, identifying the major trends in the field, implicated biological pathways, and how these pathways relate to health and aging. We conclude by assessing the main challenges in the research to date, opportunities for advancing the field, and the outlook for precision health applications.

Forging a path to precision health.

Forging a path to precision health.

Helping the science speak for itself: lessons learned developing video summaries for a precision medicine and health disparities collaborative.

Skills to effectively communicate research findings are important to expand meaningful inclusion of the public in research, but attempts to summarize findings may be challenging given increasing reliance on succinct communications. Led by our Ethics Advisory Board and within the context of this work with the Precision Medicine and Health Disparities Collaborative, the authors decided to engage in the iterative process of developing video summaries. Our stakeholders taught us to incorporate novel strategies to engage broader audiences, leading to the production of two video summaries, a public summary toolbox and an adapted process for developing video summaries. The authors refreshed the static concept of 'research summaries'and found ways of producing summaries which emphasized a dynamic reframing of the shared benefits of science.

Fine-tuned Sentiment Analysis of COVID-19 Vaccine-Related Social Media Data: Comparative Study.

BackgroundThe emergence of the novel coronavirus (COVID-19) and the necessary separation of populations have led to an unprecedented number of new social media users seeking information related to the pandemic. Currently, with an estimated 4.5 billion users worldwide, social media data offer an opportunity for near real-time analysis of large bodies of text related to disease outbreaks and vaccination. These analyses can be used by officials to develop appropriate public health messaging, digital interventions, educational materials, and policies.ObjectiveOur study investigated and compared public sentiment related to COVID-19 vaccines expressed on 2 popular social media platforms—Reddit and Twitter—harvested from January 1, 2020, to March 1, 2022.MethodsTo accomplish this task, we created a fine-tuned DistilRoBERTa model to predict the sentiments of approximately 9.5 million tweets and 70 thousand Reddit comments. To fine-tune our model, our team manually labeled the sentiment of 3600 tweets and then augmented our data set through back-translation. Text sentiment for each social media platform was then classified with our fine-tuned model using Python programming language and the Hugging Face sentiment analysis pipeline.ResultsOur results determined that the average sentiment expressed on Twitter was more negative (5,215,830/9,518,270, 54.8%) than positive, and the sentiment expressed on Reddit was more positive (42,316/67,962, 62.3%) than negative. Although the average sentiment was found to vary between these social media platforms, both platforms displayed similar behavior related to the sentiment shared at key vaccine-related developments during the pandemic.ConclusionsConsidering this similar trend in shared sentiment demonstrated across social media platforms, Twitter and Reddit continue to be valuable data sources that public health officials can use to strengthen vaccine confidence and combat misinformation. As the spread of misinformation poses a range of psychological and psychosocial risks (anxiety and fear, etc), there is an urgency in understanding the public perspective and attitude toward shared falsities. Comprehensive educational delivery systems tailored to a population’s expressed sentiments that facilitate digital literacy, health information–seeking behavior, and precision health promotion could aid in clarifying such misinformation.

Editorial: The path towards precision health: Prospects and challenges

Editorial: The path towards precision health: Prospects and challenges

Construction of Intelligent Building Integrated Evaluation System Based on BIM Technology

In order to alleviate the impact of the old bridge by the environment and its own life and the problem of increasing maintenance cost, the intelligent building integrated evaluation system based on BIM technology is proposed. Firstly, based on BIM technology, an all-weather hardware monitoring system is constructed by using multiple sensors laid on the bridge. The software communicates with the monitoring system through 5G public network and applies the unique advantages of deep neural network in classification to the assessment of the health status of old bridges with the help of the multiclassification convolutional neural network embedded in the software. The test results show that the preprocessed data is imported into the health diagnosis module for training, and the training accuracy reaches about 70%, and the loss curve is stable at about 0.5. Conclusion. The system meets the design requirements and solves the problem of difficult health diagnosis caused by long monitoring period and low efficiency of old bridges. The system has advantages of low difficulty in obtaining parameters, high precision, and accurate health assessment.

Predictors of ventricular arrhythmias in arrhythmic mitral valve syndrome – a study using implantable loop recorders

Abstract Background Arrhythmic mitral valve syndrome is linked to life-threatening ventricular arrhythmias, but the incidence and methods for risk stratification of ventricular arrhythmias are not well known. Purpose In this prospective study, we aimed to describe the incidence of ventricular arrhythmias by use of continuous rhythm monitoring and propose risk stratification in patients with arrhythmic mitral valve syndrome. Methods We included consecutive patients with arrhythmic mitral valve prolapse, defined as mitral valve prolapse and/or mitral annulus disjunction (MAD) with arrhythmic symptoms or documented complex premature ventricular complexes (PVC). We implanted loop recorders (ILR) in patients with arrhythmic mitral valve syndrome with no previous severe ventricular arrhythmias and all were followed by remote monitoring. We monitored for non-sustained ventricular tachycardias (NSVT) and severe ventricular arrhythmia, defined as aborted cardiac arrest, sustained ventricular tachycardia and NSVT with syncope/presyncope. At baseline, patients underwent echocardiography, 24-hour Holter monitoring, stress ECG and cardiac magnetic resonance imaging. Results We included 60 patients (73% female, 49 years [interquartile range [IQR] 37–60]). At baseline, median PVC burden was 232 per 24-hours (IQR 33–1329), and 8 (13%) patients had NSVT by Holter monitoring or stress ECG. During 3.0±0.5 years of follow-up, severe ventricular arrhythmia occurred in 7 (12%) patients (1 aborted cardiac arrest, 6 NSVT with syncope/pre-syncope). Annual incidence rate of severe events was 4% per person-year (95% confidence interval [CI] 2–9)). Predictors of severe ventricular arrhythmia were frequent PVCs (6683 PVCs per 24h [IQR 612–10861] vs. 154 PVCs per 24h [IQR 25–562], p=0.01), more NSVTs by ILR (4 NSVTs [IQR 2–7] vs. 0 NSVTs [0–1], p<0.001), greater left ventricular diameter (57±6 mm vs. 51±6 mm, p=0.01) and greater posterolateral MAD distance (9 mm [IQR 8–12] vs. 4 mm [IQR 0–6], p=0.02). Adjustment for age and sex did not change significant risk markers (all p<0.02). During follow-up 102 NSVTs were recorded in 24 (40%) patients. In multivariate Poisson regression adjusting for age, mitral regurgitation grade and bileaflet prolapse, predictors of high NSVT burden were frequent PVCs (incidence rate ratio [IRR] 1.14 [95% CI 1.04–1.27], p=0.01), PVCs from inferior left ventricle (IRR 5.19 [95% CI 2.80–9.63], p<0.001) and late gadolinium enhancement (IRR 3.16 [95% CI 1.68–5.97], p<0.001). Conclusion Yearly incidence of first severe ventricular arrhythmia was 4% in patients with arrhythmic mitral valve syndrome, emphasizing the high risk of life-threatening events in these patients. Frequent PVCs, NSVTs, greater left ventricular diameter and greater posterolateral MAD distance were predictors of first severe ventricular arrhythmia and should be used in risk stratification. Funding Acknowledgement Type of funding sources: Public grant(s) – National budget only. Main funding source(s): This study was supported by a public grant from the Norwegian Research Council (#288438 and #309762), Precision Health Center for optimized cardiac care (ProCardio).

Translation and cross-cultural validation of a precision health tool, the Suboptimal Health Status Questionnaire-25, in Korean.

BackgroundSuboptimal health status (SHS) is a reversible stage between health and illness that is characterized by health complaints, low energy, general weakness, and chronic fatigue. The Suboptimal Health Status Questionnaire-25 (SHSQ-25) has been validated in three major populations (African, Asian, and Caucasian) and is internationally recognized as a reliable and robust tool for health estimation in general populations. This study focused on the development of K-SHSQ-25, a Korean version of the SHSQ-25, from its English version.MethodsThe SHSQ-25 was translated from English to Korean according to international guidelines set forth by the World Health Organization (WHO) for health instrument translation between different languages. A subsequent cross-sectional survey involved 460 healthy South Korean participants (aged 18-83 years; 65.4% females) to answer the 25 questions focusing on the health perspectives of 5 domains, 1) fatigue, 2) cardiovascular health, 3) digestive tract, 4) immune system and 5) mental health. The K-SHSQ-25 was further validated using tests for reliability, internal consistency, exploratory factor analysis (EFA), and confirmatory factor analysis (CFA).ResultsThe version of K-SHSQ-25 achieved linguistic, cultural, and conceptual equivalence to the English version. The intraclass correlation coefficient (ICC) of test-retest reliability for individual items ranged from 0.88 to 0.99. Reliability estimates based on internal consistency reached a Cronbach’s α of 0.953; the Cronbach’s α for each domain ranged from 0.76 to 0.94. Regarding construct validity, the EFA of the K-SHSQ-25 generally replicated the multidimensional structure (fatigue, cardiovascular, digestive, immune system, and mental health) and 25 questions. The CFA revealed that the root mean square error of approximation (RMSEA), goodness-of-fit index (GFI) and adjusted goodness of fit index (AGFI) were excellent (RMSEA = 0.069<0.08, GFI = 0.929>0.90, AGFI = 0.907>0.90). The five domains of the K-SHSQ-25 showed significant correlations with each other (r = 0.59-0.81, P<0.001). The cut-off point of K-SHSQ-25 for SHS was determined as an SHS score of 25. The prevalence of SHS in this study was 60.0% (276/460), with 47.8% (76/159) for males and 58.5% for females (176/301).ConclusionsOur results indicate that the Korean version of SHSQ-25, K-SHSQ-25, is a transcultural equivalent, robust, valid, and reliable assessment tool for evaluating SHS in the Korean-speaking population.

Inclusion of Social and Structural Determinants of Health to Advance Understanding of their Influence on the Biology of Chronic Disease.

Social Determinants of Health (SDOH) consider social, political, and economic factors that contribute to health disparities in patients and populations. The most common health-related SDOH exposures are food and housing insecurity, financial instability, transportation needs, low levels of education, and psychosocial stress. These domains describe risks that can impact health outcomes more than health care. Epidemiologic and translational research demonstrates that SDOH factors represent exposures that predict harm and impact the health of individuals. International and national guidelines urge health professionals to address SDOH in clinical practice and public health. The further implementation of these recommendations into basic and translational research, however, is lagging. Herein, we consider a precision health framework to describe how SDOH contributes to the exposome and exacerbates physiologic pathways that lead to chronic disease. SDOH factors are associated with various forms of stressors that impact physiological processes through epigenetic, inflammatory, and redox regulation. Many SDOH exposures may add to or potentiate the pathologic effects of additional environmental exposures. This overview aims to inform basic life science and translational researchers about SDOH exposures that can confound associations between classic biomedical determinants of disease and health outcomes. To advance the study of toxicology through either qualitative or quantitative assessment of exposures to chemical and biological substances, a more complete environmental evaluation should include SDOH exposures. We discuss common approaches to measure SDOH factors at individual and population levels and review the associations between SDOH risk factors and physiologic mechanisms that influence chronic disease. We provide clinical and policy-based motivation to encourage researchers to consider the impact of SDOH exposures on study results and data interpretation. With valid measures of SDOH factors incorporated into study design and analyses, future toxicological research may contribute to an evidence base that can better inform prevention and treatment options, to improve equitable clinical care and population health. © 2022 Wiley Periodicals LLC.

Mobile Health Study Incorporating Novel Fitness Test.

Mobile health (mHealth) is a rapidly expanding field within precision medicine and precision health that provides healthcare support and interventions using mobile technologies, such as smartphones and smartwatches. The growing ubiquity of commercial wireless signals and smartphones allows mHealth technologies to have a substantially broader reach than traditional healthcare networks. My Fitness Counts, a cross-platform My Heart Counts spinout study, is a pioneer cross-platform mHealth study for measuring cardiovascular fitness levels. The study uses Real-World Insights, a platform designed to host mHealth studies. In this paper, we present insights gained through the quality control process undertaken prior to the release of the cross-platform mHealth study My Fitness Counts. Through extensive testing of the 21 iOS and 11 Android builds of the application, over 70 bugs were identified and corrected during the 5-month development process of My Fitness Counts.

Effect of Precision Health Education on Compliance Behavior after Extracorporeal Shock Wave Lithotripsy.

Objective The aim of this study was to investigate the effects of precision health education on patients' compliance behavior and lithotripsy efficacy after extracorporeal shock wave lithotripsy (ESWL). Methods From June 2018 to September 2018, 104 patients with upper urinary tract calculi who underwent ESWL at Anhui No.2 Provincial People's Hospital were randomly divided into two groups. The observation group (n = 53) was given precision health education guidance after ESWL treatment, while the control group (n = 51) was only informed of precautions after lithotripsy. Postoperative compliance behavior, stone expulsion time, and complications were recorded in both groups. Results The clinical baseline data of the two groups were similar in age, gender, number of stones, size of stones, location of stones, complications, number of lithotripsies, and voltage of lithotripsy between the two groups (P > 0.05). However, compared with the control group, the observation group had better compliance behavior of patients. Furthermore, we divided the patients into three groups based on whether they complied with medical advice. Compared with the noncompliance group (n = 17) and the partial compliance group (n = 23), the stone expulsion time in the complete compliance group (n = 64) was significantly shorter. Also, the number of patients with nausea and vomiting, lacrimation, and cold sweat in the complete compliance group was significantly lower than that in the other groups. Conclusion Precision health education for patients treated with ESWL can significantly enhance the compliance behavior of patients, thereby accelerating the expulsion of stones, improving the efficacy and relieving pain. Precision health education and good compliance behavior complement and reinforce each other.

Building Capacity for Implementation Science in Precision Health and Society: Development of a Course for Professional and Graduate Students in Pharmacy.

Training in the field of implementation science is critical for future pharmacists and pharmaceutical scientists to successfully implement precision health interventions in pharmacy practice. We developed an elective course for second- and third-year students at the UNC Eshelman School of Pharmacy to develop foundational knowledge in implementation science with a focus on precision health implementation. The eight-week course used a flipped classroom format featuring lecture videos, suggested readings, quizzes, guest lectures from experts, case studies, and a group project. We evaluated course quality through class participation, a pre- and post-test on course content, and a mixed-methods survey completed by the students. Overall engagement in in the course was high and students demonstrated significant improvement in understanding of implementation science and precision health. Strengths of the course as identified by students were the use of expert guest speakers, pre-class lectures, and case study exercises, while the ordering of content and improved connection between content and guest lectures were identified as areas for improvement. In conclusion, the elective course was well-received and meets a critical need in the field of pharmacy to build implementation science capacity. Future work is needed to expand and refine education for the implementation of precision health for pharmacy professionals.

Validating and automating learning of cardiometabolic polygenic risk scores from direct-to-consumer genetic and phenotypic data: implications for scaling precision health research

IntroductionA major challenge to enabling precision health at a global scale is the bias between those who enroll in state sponsored genomic research and those suffering from chronic disease. More than 30 million people have been genotyped by direct-to-consumer (DTC) companies such as 23andMe, Ancestry DNA, and MyHeritage, providing a potential mechanism for democratizing access to medical interventions and thus catalyzing improvements in patient outcomes as the cost of data acquisition drops. However, much of these data are sequestered in the initial provider network, without the ability for the scientific community to either access or validate. Here, we present a novel geno-pheno platform that integrates heterogeneous data sources and applies learnings to common chronic disease conditions including Type 2 diabetes (T2D) and hypertension.MethodsWe collected genotyped data from a novel DTC platform where participants upload their genotype data files and were invited to answer general health questionnaires regarding cardiometabolic traits over a period of 6 months. Quality control, imputation, and genome-wide association studies were performed on this dataset, and polygenic risk scores were built in a case–control setting using the BASIL algorithm.ResultsWe collected data on N = 4,550 (389 cases / 4,161 controls) who reported being affected or previously affected for T2D and N = 4,528 (1,027 cases / 3,501 controls) for hypertension. We identified 164 out of 272 variants showing identical effect direction to previously reported genome-significant findings in Europeans. Performance metric of the PRS models was AUC = 0.68, which is comparable to previously published PRS models obtained with larger datasets including clinical biomarkers.DiscussionDTC platforms have the potential of inverting research models of genome sequencing and phenotypic data acquisition. Quality control (QC) mechanisms proved to successfully enable traditional GWAS and PRS analyses. The direct participation of individuals has shown the potential to generate rich datasets enabling the creation of PRS cardiometabolic models. More importantly, federated learning of PRS from reuse of DTC data provides a mechanism for scaling precision health care delivery beyond the small number of countries who can afford to finance these efforts directly.ConclusionsThe genetics of T2D and hypertension have been studied extensively in controlled datasets, and various polygenic risk scores (PRS) have been developed. We developed predictive tools for both phenotypes trained with heterogeneous genotypic and phenotypic data generated outside of the clinical environment and show that our methods can recapitulate prior findings with fidelity. From these observations, we conclude that it is possible to leverage DTC genetic repositories to identify individuals at risk of debilitating diseases based on their unique genetic landscape so that informed, timely clinical interventions can be incorporated.

Noncoding RNome as Enabling Biomarkers for Precision Health.

Noncoding RNAs (ncRNAs), in the form of structural, catalytic or regulatory RNAs, have emerged to be critical effectors of many biological processes. With the advent of new technologies, we have begun to appreciate how intracellular and circulatory ncRNAs elegantly choreograph the regulation of gene expression and protein function(s) in the cell. Armed with this knowledge, the clinical utility of ncRNAs as biomarkers has been recently tested in a wide range of human diseases. In this review, we examine how critical factors govern the success of interrogating ncRNA biomarker expression in liquid biopsies and tissues to enhance our current clinical management of human diseases, particularly in the context of cancer. We also discuss strategies to overcome key challenges that preclude ncRNAs from becoming standard-of-care clinical biomarkers, including sample pre-analytics standardization, data cross-validation with closer attention to discordant findings, as well as correlation with clinical outcomes. Although harnessing multi-modal information from disease-associated noncoding RNome (ncRNome) in biofluids or in tissues using artificial intelligence or machine learning is at the nascent stage, it will undoubtedly fuel the community adoption of precision population health.

The New Precision Stewards?

The precision health era is likely to reduce and respond to antimicrobial resistance (AMR). Our stewardship and precision efforts share terminology, seeking to deliver the “right drug, at the right dose, at the right time.” Already, rapid diagnostic testing, phylogenetic surveillance, and real-time outbreak response provide just a few examples of molecular advances we dub “precision stewardship.” However, the AMR causal factors range from the molecular to that of global health policy. Mirroring the cross-sectoral nature of AMR science, the research addressing the ethical, legal and social implications (ELSI) of AMR ranges across academic scholarship. As the rise of AMR is accompanied by an escalating sense of its moral and social significance, what is needed is a parallel field of study. In this paper, we offer a gap analysis of this terrain, or an agenda for “the ELSI of precision stewardship.” In the first section, we discuss the accomplishments of a multi-decade U.S. national investment in ELSI research attending to the advances in human genetics. In the next section, we provide an overview of distinct ELSI topics pertinent to AMR. The distinctiveness of an ELSI agenda for precision stewardship suggests new opportunities for collaboration to build the stewardship teams of the future.

Predicting Hospice Transitions in Dementia Caregiving Dyads: An Exploratory Machine Learning Approach.

Hospice programs assist people with serious illness and their caregivers with aging in place, avoiding unnecessary hospitalizations, and remaining at home through the end-of-life. While evidence is emerging of the myriad of factors influencing end-of-life care transitions among persons living with dementia, current research is primarily cross- sectional and does not account for the effect that changes over time have on hospice care uptake, access, and equity within dyads. Secondary data analysis linking the National Health and Aging Trends Study to the National Study of Caregiving investigating important social determinants of health and quality-of-life factors of persons living with dementia and their primary caregivers (n = 117) on hospice utilization over 3 years (2015-2018). We employ cutting-edge machine learning approaches (correlation matrix analysis, principal component analysis, random forest [RF], and information gain ratio [IGR]). IGR indicators of hospice use include persons living with dementia having diabetes, a regular physician, a good memory rating, not relying on food stamps, not having chewing or swallowing problems, and whether health prevents them from enjoying life (accuracy = 0.685; sensitivity = 0.824; specificity = 0.537; area under the curve (AUC) = 0.743). RF indicates primary caregivers' age, and the person living with dementia's income, census division, number of days help provided by caregiver per month, and whether health prevents them from enjoying life predicts hospice use (accuracy = 0.624; sensitivity = 0.713; specificity = 0.557; AUC = 0.703). Our exploratory models create a starting point for the future development of precision health approaches that may be integrated into learning health systems that prompt providers with actionable information about who may benefit from discussions around serious illness goals-for-care. Future work is necessary to investigate those not considered in this study-that is, persons living with dementia who do not use hospice care so additional insights can be gathered around barriers to care.