Atrium Health Research Articles

Using a Quality-Controlled Dataset From ViSi Mobile Monitoring for Analyzing Posture Patterns of Hospitalized Patients: Retrospective Observational Study.

ViSi Mobile has the capability of monitoring a patient's posture continuously during hospitalization. Analysis of ViSi telemetry data enables researchers and health care providers to quantify an individual patient's movement and investigate collective patterns of many patients. However, erroneous values can exist in routinely collected ViSi telemetry data. Data must be scrutinized to remove erroneous records before statistical analysis. The objectives of this study were to (1) develop a data cleaning procedure for a 1-year inpatient ViSi posture dataset, (2) consolidate posture codes into categories, (3) derive concise summary statistics from the continuous monitoring data, and (4) study types of patient posture habits using summary statistics of posture duration and transition frequency. This study examined the 2019 inpatient ViSi posture records from Atrium Health Wake Forest Baptist Medical Center. First, 2 types of errors, record overlap and time inconsistency, were identified. An automated procedure was designed to search all records for these errors. A data cleaning procedure removed erroneous records. Second, data preprocessing was conducted. Each patient's categorical time series was simplified by consolidating the 185 ViSi codes into 5 categories (Lying, Reclined, Upright, Unknown, User-defined). A majority vote process was applied to remove bursts of short duration. Third, statistical analysis was conducted. For each patient, summary statistics were generated to measure average time duration of each posture and rate of posture transitions during the whole day and separately during daytime and nighttime. A k-means clustering analysis was performed to divide the patients into subgroups objectively. The analysis used a sample of 690 patients, with a median of 3 days of extensive ViSi monitoring per patient. The median of posture durations was 10.2 hours/day for Lying, 8.0 hours/day for Reclined, and 2.5 hours/day for Upright. Lying had similar percentages of patients in low and high durations. Reclined showed a decrease in patients for higher durations. Upright had its peak at 0-2 hours, with a decrease for higher durations. Scatter plots showed that patients could be divided into several subgroups with different posture habits. This was reinforced by the k-means analysis, which identified an active subgroup and two sedentary ones with different resting styles. Using a 1-year ViSi dataset from routine inpatient monitoring, we derived summary statistics of posture duration and posture transitions for each patient and analyzed the summary statistics to identify patterns in the patient population. This analysis revealed several types of patient posture habits. Before analysis, we also developed methodology to clean and preprocess routinely collected inpatient ViSi monitoring data, which is a major contribution of this study. The procedure developed for data cleaning and preprocessing can have broad application to other monitoring systems used in hospitals.

Screening, identification, and diagnosis of malnutrition in hospitalized patients with solid tumors: A retrospective cohort study.

Malnutrition is common in hospitalized patients with cancer and adversely affects clinical outcomes. We evaluated the prevalence of malnutrition risk, dietitian-identified malnutrition (DIMN), and physician-diagnosed malnutrition (PDMN) at admission. This retrospective study included adults diagnosed with a stage I-IV solid tumor malignancy and admitted to Atrium Health Carolinas Medical Center from January 2016 to May 2019. Malnutrition risk was determined by a score ≥2 on the Malnutrition Screening Tool (MST) administered by a registered nurse during the intake process. Registered dietitian nutritionist (RDN) assessments were reviewed for DIMN and grade (mild, moderate, or severe). PDMN included malnutrition International Classification of Diseases, Tenth Revision codes in the discharge summary. Univariate models were estimated; multivariate logistic regression models identified associations between clinicodemographic factors and malnutrition prevalence with stepwise selection. A total of 5143 patients were included. Median age was 63 (range 18-102) years, 48% were female, 70% were White, and 24% were Black. Upper gastrointestinal (21%), thoracic (18%), and genitourinary (18%) cancers were most common. A total of 28% had stage IV disease. MST scores were available for 4085 (79%); 1005 of 4085 (25%) were at malnutrition risk. Eleven percent (n = 557) had malnutrition coded by a physician or documented by an RDN; 4% (n = 223) of these were identified by both clinicians, 4% (n = 197) by RDNs only, and 3% (n = 137) by physicians only. Malnutrition appears to be underdiagnosed by both RDNs and physicians. Underdiagnosis of malnutrition may have significant clinical, operational, and financial implications in cancer care.

Predictive Models for Sustained, Uncontrolled Hypertension and Hypertensive Crisis Based on Electronic Health Record Data: Algorithm Development and Validation.

Assessing disease progression among patients with uncontrolled hypertension is important for identifying opportunities for intervention. We aim to develop and validate 2 models, one to predict sustained, uncontrolled hypertension (≥2 blood pressure [BP] readings ≥140/90 mm Hg or ≥1 BP reading ≥180/120 mm Hg) and one to predict hypertensive crisis (≥1 BP reading ≥180/120 mm Hg) within 1 year of an index visit (outpatient or ambulatory encounter in which an uncontrolled BP reading was recorded). Data from 142,897 patients with uncontrolled hypertension within Atrium Health Greater Charlotte in 2018 were used. Electronic health record-based predictors were based on the 1-year period before a patient's index visit. The dataset was randomly split (80:20) into a training set and a validation set. In total, 4 machine learning frameworks were considered: L2-regularized logistic regression, multilayer perceptron, gradient boosting machines, and random forest. Model selection was performed with 10-fold cross-validation. The final models were assessed on discrimination (C-statistic), calibration (eg, integrated calibration index), and net benefit (with decision curve analysis). Additionally, internal-external cross-validation was performed at the county level to assess performance with new populations and summarized using random-effect meta-analyses. In internal validation, the C-statistic and integrated calibration index were 0.72 (95% CI 0.71-0.72) and 0.015 (95% CI 0.012-0.020) for the sustained, uncontrolled hypertension model, and 0.81 (95% CI 0.79-0.82) and 0.009 (95% CI 0.007-0.011) for the hypertensive crisis model. The models had higher net benefit than the default policies (ie, treat-all and treat-none) across different decision thresholds. In internal-external cross-validation, the pooled performance was consistent with internal validation results; in particular, the pooled C-statistics were 0.70 (95% CI 0.69-0.71) and 0.79 (95% CI 0.78-0.81) for the sustained, uncontrolled hypertension model and hypertensive crisis model, respectively. An electronic health record-based model predicted hypertensive crisis reasonably well in internal and internal-external validations. The model can potentially be used to support population health surveillance and hypertension management. Further studies are needed to improve the ability to predict sustained, uncontrolled hypertension.

Navigating the path to equitable rheumatologic care for underserved children with quality improvement.

The aim of this quality improvement project is to identify children with rheumatologic conditions to prevent delayed or missed diagnosis in underserved pediatric populations. Our focus is on prompt and accurate identification and subsequent treatment of rheumatologic symptoms in pediatric patients referred from Atrium Health safety-net primary care clinics that deliver care to families without private insurance, including those lacking insurance entirely. We collaborated with providers at one safety-net clinic to improve the processes of identification and subspecialty referral, resulting in an increase in the number of identified pediatric patients and referrals for these patients with potential rheumatologic disease. We used the Model for Improvement framework with rapid Plan-Do-Study-Act cycles and evaluated improvement with run and statistical process control charts. We achieved improvement, with zero referrals in the previous 5 years for the targeted population increasing to 15 patient referrals within 1 year of project initiation. Despite this increase in referrals, the rheumatology clinic was able to see all priority patients within 20 business days from referral. An awareness of concerning rheumatologic symptoms in safety-net primary care clinics, combined with the use of both visual and decision aids, allows care teams to efficiently recognize and accurately refer patients needing specialty care.

Geographic residence of patients with medication-related osteonecrosis of the jaw.

135 Background: Bone-modifying agents (BMAs) are utilized to reduce the risk of skeletal-related events in patients with bone metastases. Medication-related osteonecrosis of the jaw (MRONJ) is a rare but difficult to treat adverse event of BMAs. There are known patient and treatment related risk factors for development of ONJ, however, little is known on social and geographic risks. In patients with bone metastases and concurrent MRONJ, we seek to describe differences in patient's geographic residence (rural versus urban) with other social factors as secondary outcomes. Methods: We conducted a retrospective chart review of Atrium Health Wake Forest Baptist Comprehensive Cancer Center patients in the date range of 01/01/2013 to 11/01/2023. Included were patients with metastatic breast cancer, metastatic prostate cancer or multiple myeloma who had a diagnosis of MRONJ. We recorded Rural-Urban Commuting Area (RUCA) codes, the Area Deprivation Index (ADI), age, race, ethnicity, smoking status, alcohol use, body mass index, insurance status, type of cancer, type of BMA, and cumulative dosing of BMAs. Patients were defined as rural or urban based on criteria from the health resources and services administration website. Relative socioeconomic advantage versus disadvantage was determined by the patients’ ADIs with a decile rank from 1-10 for state ADIs and a percentile rank of 1-100 for national ADIs. SAS software (version 9.4) was used for data analysis (2023, SAS Institute Inc, Cary, NC). Results: A total of 88 patients met inclusion criteria. 64% of patients lived in urban areas and 36% in rural. Demographic features of the identified patients are listed in table 1. MRONJ patients from both urban and rural areas had an above average national and state ADI, with a trend toward higher ADIs in rural locations compared to urban locations; mean of 6.7 vs 5.3 state ADI and mean 70.6 vs 62.0 national ADI, for rural and urban patients respectively. There were no statistically significant differences between rural and urban patients in any of the demographic or health-related characteristics we examined. Conclusions: Patients with bone metastases and MRONJ in our cohort demonstrate relative socioeconomic disadvantage in comparison to national and state averages. In comparing the rural with urban patients, however, we observed relatively similar profiles. Patient characteristics by rural/urban status. Rural (n=32) Urban (n=56) Age in years (mean, sd) 64 (11.3) 65 (10.9) Sex, N (%) Male 19 (59) 29 (52) Female 13 (41) 27 (48) Race, N (%) Non-Hispanic White 30 (93.8) 40 (71.4) Non-Hispanic Black 1 (3.1) 11 (19.6) Hispanic 1 (3.1) 3 (5.4) Non-Hispanic Asian 0 (0.0) 1 (1.8) Other 0 (0.0) 1 (1.8) State ADI (mean, sd) 6.7 (2.4) 5.5 (2.7) National ADI (mean, sd) 70.6 (16.9) 62.0 (23.8) Former/current smoker, N (%) 18 (56.3) 25 (44.6) Medicaid/Medicare, N (%) 16 (50.0) 16 (28.6) Months on BMA (mean, sd) 37.2 (35.5) 40.3 (38.2)

Pediatric Facial Fractures: A Multi-Institutional Level 1 Trauma Center Analysis of Incidence, Interventions, and Outcomes

Study Design Retrospective chart review. Objective The management of pediatric facial fractures presents distinctive considerations compared to adults. This study aims to provide a unique perspective on the correlations between the mechanism of injury, types of facial fractures, and fracture interventions and management utilized in 2 North Carolina Level 1 Trauma Centers to determine the optimal management options for this patient population. Methods An IRB-approved retrospective chart review was performed of pediatric facial trauma patients ages <18 years old between January 2020 and December 2022 at Atrium Health Wake Forest Baptist Medical Center and Atrium Health Charlotte Medical Center. Data on patient demographics, mechanism of injury, facial fractures, interventions, and outcomes were collected. Results Of 2,977 pediatric facial trauma patients, 582 patients sustained at least 1 facial fracture at the time of injury. Adolescents were significantly less likely to be transferred from outside institutions and to be admitted for further care ( P = 0.002). Adolescents experienced higher levels of residual symptoms following initial discharge ( P = 0.001) and were less likely to have a symptom resolution within 1 year ( P < 0.0001). Neonates and infants were significantly more likely to receive conservative interventions and to sustain calvarium and skull base fractures ( P < 0.0001). Conclusions This study identifies differences in pediatric age groups related to transfers, admittance, fracture type, management, and outcomes. Our data suggests adolescent patients may experience a higher incidence of residual symptoms with lower levels of symptom resolution within 1 year. Further investigation into these differences may elicit optimized methods of fracture management in pediatric age groups and allow for effective, individualized care with improved long-term outcomes.

Incorporating patient, caregiver, and provider perspectives in the co-design of an app to guide Hospital at Home admission decisions: a qualitative analysis.

Hospital at Home (HaH) programs currently lack decision support tools to help efficiently navigate the complex decision-making process surrounding HaH as a care option. We assessed user needs and perspectives to guide early prototyping and co-creation of 4PACS (Partnering Patients and Providers for Personalized Acute Care Selection), a decision support app to help patients make an informed decision when presented with discrete hospitalization options. From December 2021 to January 2022, we conducted semi-structured interviews via telephone with patients and caregivers recruited from Atrium Health's HaH program and physicians and a nurse with experience referring patients to HaH. Interviews were evaluated using thematic analysis. The findings were synthesized to create illustrative user descriptions to aid 4PACS development. In total, 12 stakeholders participated (3 patients, 2 caregivers, 7 providers [physicians/nurse]). We identified 4 primary themes: attitudes about HaH; 4PACS app content and information needs; barriers to 4PACS implementation; and facilitators to 4PACS implementation. We characterized 3 user descriptions (one per stakeholder group) to support 4PACS design decisions. User needs included patient selection criteria, clear program details, and descriptions of HaH components to inform care expectations. Implementation barriers included conflict between app recommendations and clinical judgement, inability to adequately represent patient-risk profile, and provider burden. Implementation facilitators included ease of use, auto-populating features, and appropriate health literacy. The findings indicate important information gaps and user needs to help inform 4PACS design and barriers and facilitators to implementing 4PACS in the decision-making process of choosing between hospital-level care options.

TrajVis: a visual clinical decision support system to translate artificial intelligence trajectory models in the precision management of chronic kidney disease.

Our objective is to develop and validate TrajVis, an interactive tool that assists clinicians in using artificial intelligence (AI) models to leverage patients' longitudinal electronic medical records (EMRs) for personalized precision management of chronic disease progression. We first perform requirement analysis with clinicians and data scientists to determine the visual analytics tasks of the TrajVis system as well as its design and functionalities. A graph AI model for chronic kidney disease (CKD) trajectory inference named DisEase PrOgression Trajectory (DEPOT) is used for system development and demonstration. TrajVis is implemented as a full-stack web application with synthetic EMR data derived from the Atrium Health Wake Forest Baptist Translational Data Warehouse and the Indiana Network for Patient Care research database. A case study with a nephrologist and a user experience survey of clinicians and data scientists are conducted to evaluate the TrajVis system. The TrajVis clinical information system is composed of 4 panels: the Patient View for demographic and clinical information, the Trajectory View to visualize the DEPOT-derived CKD trajectories in latent space, the Clinical Indicator View to elucidate longitudinal patterns of clinical features and interpret DEPOT predictions, and the Analysis View to demonstrate personal CKD progression trajectories. System evaluations suggest that TrajVis supports clinicians in summarizing clinical data, identifying individualized risk predictors, and visualizing patient disease progression trajectories, overcoming the barriers of AI implementation in healthcare. The TrajVis system provides a novel visualization solution which is complimentary to other risk estimators such as the Kidney Failure Risk Equations. TrajVis bridges the gap between the fast-growing AI/ML modeling and the clinical use of such models for personalized and precision management of chronic diseases.

Genomic characteristics of exceptional responders to treatment with PD-1 inhibitors in refractory/metastatic head and neck squamous cell carcinoma.

6062 Background: With the emergence of immune checkpoint inhibitors (ICIs), there are patients with advanced cancers experiencing exceptional treatment response with overall low treatment-associated toxicities. Despite success in refractory/metastatic head and neck squamous cell carcinoma (HNSCC), only a portion of patients experience benefit from the treatment with ICIs. Critical information on predictors of patient’s response to treatment is limited. At our institution we analyzed the genomic profiles of 36 patients with advanced HNSCC who demonstrated an exceptional response to treatment with PD-1 inhibitors. Methods: We reviewed patients with histologically confirmed HNSCC that had been treated with PD-1 inhibitors at Atrium Health Wake Forest Baptist Health. Treatment regimen included pembrolizumab 200mg every 3 weeks or nivolumab 240mg every 2 weeks. Response to therapy and duration of remission were monitored via CT scans and/or MRI scans at 3-month intervals, using RECIST 1.1 criteria. Exceptional responders were defined as achieving complete response (CR) or partial response (PR) for greater than six months, or stable disease (SD) for greater than one year. Progression free survival (PFS) and overall survival (OS) were measured from the date of PD-1 inhibitor initiation. PD-L1 level was measured via FoundationOne immunohistochemistry staining and reported as combined positive score (CPS). Additionally, next generation sequencing was performed on tumor (30/36 patients) and blood (33/36 patients) specimens via FoundationOne and Guardant360 testing platforms. Results: Out of 155 HNSCC patients treated with a PD-1 inhibitor, 36 patients demonstrated an exceptional response as defined above. 20 with CR, 9 with PR and 7 with SD. With an average follow up of 32 mo (95% CI, 26 mo-ongoing), the average OS was 33 mo (95% CI, 28 mo–ongoing) and the average PFS was 25 mo (95% CI, 20 mo-ongoing). These patients underwent an average of 27 treatments (range, 10-64 treatments to date). PD-L1 average score at 36 (95% CI, 22-49) and average TMB at 19 mut/mb (range, 0-90 mut/mb). Of 290 detected gene mutations, the most frequently mutated genes were TP53 (75%), PIK3CA (42%), CDKN2A (36%), EGFR (33%), TERT (25%), BRCA2 (22%), ERBB2 (22%), HNF1A (22%), MLL2 (22%), NOTCH1 (22%), and RAD21 (22%). Conclusions: There is limited data reporting the genomic profile of exceptional responders to PD-1 inhibitors in refractory/metastatic HNSCC. Higher TMB and single gene mutations such as TP53, PIK3CA, CDKN2A, EGFR, and TERT are frequently detected in exceptional responders and may provide future direction into the investigation of predictors of response to ICIs.

Evaluation of a pharmacist-led personal continuous glucose monitor workflow to improve glycemic management in an internal medicine clinic

BackgroundThe use of personal continuous glucose monitors (CGMs) in patients with diabetes has increased substantially and is expected to continue to increase as CGMs become more affordable and insurance plans improve coverage. The utilization of CGMs has improved diabetes management and reduced hypoglycemic events. ObjectivesTo create pharmacist-led personal CGM workflow and evaluate its impact on glycemic management in patients with diabetes. Practice DescriptionThe study took place at an Internal Medicine Clinic. The practice providers include 2 medical doctors, 5 physician assistants, 2 nurse practitioners, and 1 clinical pharmacist. Practice InnovationTo create and implement a sustainable pharmacy led CGM workflow for enhanced CGM use within an internal medicine clinic. Evaluation MethodsThis was a prospective, investigator-initiated pilot study conducted at an Atrium Health Internal Medicine clinic over 28 weeks. In this pilot, 42 patients were qualifying candidates with diabetes and personal CGM use. In addition, 30 patients were followed until study completion and included into final analysis. ResultsThe average baseline A1c was reduced from 8.3% to 7.1% over a 3- to 6-month period. The pharmacist-led CGM workflow revealed a statistically significant reduction in A1c from baseline by an average of 1.2% (95% CI −0.6 to −1.8, P = 0.0006). On average, patients were enrolled for 19.9 weeks and had an average of 5 visits during this time. During the study duration, 100 medications changes were implemented under the existing clinical pharmacist practitioner agreement between the pharmacists and the provider. Overall, 58 Current Procedural Terminology 95251 codes were billed yielding $7052.00 in billed CGM services for the clinic. This project generated 40.6 provider relative value units. ConclusionThe utilization of a pharmacist-led personal CGM workflow can improve diabetes outcomes.

Abstract PO3-19-04: The CROWN STUDY: A multicenter prospective cohort study of cardiovascular outcomes in premenopausal women with hormone receptor positive breast cancer treated with ovarian suppression and an aromatase inhibitor

Abstract Background: Treatment for premenopausal women with high or intermediate risk hormone receptor (HR)+ breast cancer (BC) includes the concurrent use of ovarian function suppression (OFS) and an aromatase inhibitor (AI) to induce near complete estrogen deprivation (NCED). The long-term cardiovascular (CV) sequela for women treated with NCED is unknown. Premature menopause in non-cancer populations is associated with CV disease. The CROWN study (CaRdiac Outcomes With Near complete estrogen deprivation) will use sophisticated imaging assessments of cardiac dysfunction to understand the evolution of CV injury, as well as biomarker and demographic correlates, with the goal of developing tools to assess and mitigate CV risk. Methods: This is a NIH funded prospective cohort study conducted at 3 regional NCI-supported Cancer Centers (Atrium Health Wake Forest Baptist, Virginia Commonwealth and Duke) that will include premenopausal women, age ≤ 55, with Stage I-III BC following completion of planned chemotherapy, surgery and radiation with an ECOG 0-1. HR+ BC patients will receive an AI and OFS. Women with HR- BC are included as comparators. CV imaging and biomarkers will be obtained at baseline, 1 year and 2 years (Table 1). These assessments will include serial cardiac magnetic resonance (CMR) and coronary computed tomography angiography (CCTA) imaging as well as laboratory measurements, including exploratory biomarkers. The primary objective is to determine the 24-month difference in stress myocardial blood flow as measured by adenosine CMR imaging in both groups. We will correlate CMR imaging with CCTA to provide complementary detail of coronary plaque changes. The study will also assess the relevance of pre-existing risk factors on study outcomes, including an emphasis on racial disparities, and dynamic change in modifiable and treatment related risk factors. We plan to enroll 90 women, 67 in the NCED group and 23 in the HR-group, allowing for a 10% drop out rate. There are two primary types of statistical analyses. The first includes testing hypotheses between group (NCED vs HR-) and within group (longitudinal changes within the NCED group). The second analyses, involve developing predictive equations utilizing a stepwise linear regression approach to determine if patient demographics, clinical parameters and serum biomarkers are associated with cardiovascular changes over time. A total of 17 patients are enrolled as of 07/2023. NCT05309655 Study Procedures XX- XX Citation Format: Susan Dent, Emily Douglas, Sarah Hatcher, Katherine Ansley, Carolyn Park, Nathaniel O'Connell, Mary Hackney, Ralph D'Agostino, Rani Bansal, Renee Avecilla, Monica Brown, Kacey Clayton-Stiglbauer, Igor Klem, Jennifer Jordan, Alexandra Thomas. The CROWN STUDY: A multicenter prospective cohort study of cardiovascular outcomes in premenopausal women with hormone receptor positive breast cancer treated with ovarian suppression and an aromatase inhibitor [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO3-19-04.

A regionally tailored epidemiological forecast and monitoring program to guide a healthcare system in the COVID-19 pandemic

BackgroundDuring the COVID-19 pandemic, analytics and predictive models built on regional data provided timely, accurate monitoring of epidemiological behavior, informing critical planning and decision-making for health system leaders. At Atrium Health, a large, integrated healthcare system in the southeastern United States, a team of statisticians and physicians created a comprehensive forecast and monitoring program that leveraged an array of statistical methods. MethodsThe program utilized the following methodological approaches: (i) exploratory graphics, including time plots of epidemiological metrics with smoothers; (ii) infection prevalence forecasting using a Bayesian epidemiological model with time-varying infection rate; (iii) doubling and halving times computed using changepoints in local linear trend; (iv) death monitoring using combination forecasting with an ensemble of models; (v) effective reproduction number estimation with a Bayesian approach; (vi) COVID-19 patients hospital census monitored via time series models; and (vii) quantified forecast performance. ResultsA consolidated forecast and monitoring report was produced weekly and proved to be an effective, vital source of information and guidance as the healthcare system navigated the inherent uncertainty of the pandemic. Forecasts provided accurate and precise information that informed critical decisions on resource planning, bed capacity and staffing management, and infection prevention strategies. ConclusionsIn this paper, we have presented the framework used in our epidemiological forecast and monitoring program at Atrium Health, as well as provided recommendations for implementation by other healthcare systems and institutions to facilitate use in future pandemics.

Abstract 5065: Integrating blood based liquid biopsies with TNM stage improves survival prediction model for non-small cell lung cancer

Abstract Introduction:Stage is the most important prognostic factor for survival in non-small cell lung cancer (NSCLC). The current TNM (Tumor, Node, Metastasis) staging system for NSCLC uses physical exam, histology, imaging, and surgical findings to define the extent and spread of a patient’s cancer. Despite advances in molecular testing which can impact treatment and prognosis for NSCLC patients, the current system does not incorporate molecular data into staging. We proposed a novel predictive model for survival analysis which integrates molecular analysis of liquid biopsies (B) into the TNM staging system (1). In this study, we tested this new staging system and our predictive model shows TNM + B (TNMB) provides a better predictive model for survival compared to TNM alone. Methods:176 patients were identified at Atrium Health Wake Forest Baptist with Guardant 360 liquid biopsies sent at diagnosis for NSCLC. Clinical data was obtained through retrospective chart review. Molecular analysis was performed on 81 genes across all Guardant 360 liquid biopsies. Using Cox proportional hazards model, we identified significant gene mutations or genes with significant magnitude of variant allele frequency (VAF) that improved survival prediction. We then designed a novel staging system incorporating whether a patient had a positive liquid biopsy test (B1), defined by the presence of these impact gene variables, to TNM stage. Results: Three impact genes STK11 (p=0.0253), NFE2L2 (p = 0.0025), TP53 (p=0.0129) and one gene with a magnitude of VAF, ARID1A (p=0.0082) was identified which significantly improved the predictive model for survival. Patients with negative liquid biopsy test, B0, had improved survival compared to patients with a positive liquid biopsy test, B1 (p<0.005). In our cohort, TNM staging alone did not show a significant survival difference between stage II and III (p = 0.19). Whereas TNMB staging showed a significant survival difference between stage II, III and IV. Conclusions: While TNM stage remains a major prognostic factor for NSCLC patients, it fails to incorporate molecular data which has a critical impact on management and prognosis. We designed a novel method to analyze molecular data obtained by liquid biopsy to identify significant gene mutations and gene magnitude of VAF that impacts survival prediction. In our cohort, we showed 4 impact genes which we used to create a new TNMB staging system which led to improved survival prediction compared to TNM alone. Acknowledgments: Research partly supported by NIH T32 Grant (T32CA247819) and Cancer Center Support Grant (P30CA012197) Reference: (1) Yang M, Forbes ME, Bitting RL, O'Neill SS, Chou PC, Topaloglu U, Miller LD, Hawkins GA, Grant SC, DeYoung BR, Petty WJ, Chen K, Pasche BC, Zhang W. Incorporating Blood-based Liquid Biopsy Information into Cancer Staging: Time for a TNMB System? Ann Oncol. 2018 Feb 1;29(2):311-323. Citation Format: Lauren Schmalz, Cetin Urtis, Liang Liu, Elizabeth Forbes, Fang-Chi Hsu, Nury Steuerwald, Alberto de Hoyos, Thomas Lycan, Jimmy Ruiz, William Petty, Wei Zhang. Integrating blood based liquid biopsies with TNM stage improves survival prediction model for non-small cell lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 5065.

Continuous Glucose Monitoring in Older Adults: What We Know and What We Have Yet to Learn.

To assess the growing use of continuous glucose monitoring (CGM) systems by older adults and explore additional areas integration that could benefit adults with frailty. The use of CGM devices has expanded rapidly in the last decade. This has been supported by substantial data showing significant benefit in glycemic metrics: hemoglobin A1c improvements, less hypoglycemia, and improved quality of life. However, sub-populations, such as older persons, exist where available data are limited. Furthermore, frail older adults represent a heterogeneous population with their own unique challenges to the management of diabetes. This group has some of the poorest outcomes related to the sequela of diabetes. For example, hypoglycemia resulting in significant morbidity and mortality is more frequent in older person with diabetes than in younger persons with diabetes. We present a concise literature review on CGM use in the older adult as well as expand upon glycemic and nonglycemic benefits of CGM for patients, caregivers, and providers. Retrospective analysis of inpatient glycemic data of 16,935 older adults with Type 2 diabetes mellitus at Atrium Health Wake Forest Baptist indicated those with fraility managed with insulin or sulfonylurea had the highest rates of delirium (4.8%), hypoglycemia (3.5%), cardiovascular complications (20.2%) and ED visits/hospitalizatoins (49%). In addition, we address special consideration of specific situations including inpatient, palliative and long term care settings. This review article summarizes the available data for CGM use in older adults, discusses the benefits and obstacles with CGM use in this population, and identifies areas of future research needed for improved delivery of care to older persons with diabetes.

Development and validation of an electrocardiographic artificial intelligence model for detection of peripartum cardiomyopathy

This study used electrocardiogram data in conjunction with artificial intelligence methods as a noninvasive tool for detecting peripartum cardiomyopathy. This study aimed to assess the efficacy of an artificial intelligence-based heart failure detection model for peripartum cardiomyopathy detection. We first built a deep-learning model for heart failure detection using retrospective data at the University of Tennessee Health Science Center. Cases were adult and nonpregnant female patients with a heart failure diagnosis; controls were adult nonpregnant female patients without heart failure. The model was then tested on an independent cohort of pregnant women at the University of Tennessee Health Science Center with or without peripartum cardiomyopathy. We also tested the model in an external cohort of pregnant women at Atrium Health Wake Forest Baptist. Key outcomes were assessed using the area under the receiver operating characteristic curve. We also repeated our analysis using only lead I electrocardiogram as an input to assess the feasibility of remote monitoring via wearables that can capture single-lead electrocardiogram data. The University of Tennessee Health Science Center heart failure cohort comprised 346,339 electrocardiograms from 142,601 patients. In this cohort, 60% of participants were Black and 37% were White, with an average age (standard deviation) of 53 (19) years. The heart failure detection model achieved an area under the curve of 0.92 on the holdout set. We then tested the ability of the heart failure model to detect peripartum cardiomyopathy in an independent University of Tennessee Health Science Center cohort of pregnant women and an external Atrium Health Wake Forest Baptist cohort of pregnant women. The independent University of Tennessee Health Science Center cohort included 158 electrocardiograms from 115 patients; our deep-learning model achieved an area under the curve of 0.83 (0.77-0.89) for this data set. The external Atrium Health Wake Forest Baptist cohort involved 80 electrocardiograms from 43 patients; our deep-learning model achieved an area under the curve of 0.94 (0.91-0.98) for this data set. For identifying peripartum cardiomyopathy diagnosed ≥10 days after delivery, the model achieved an area under the curve of 0.88 (0.81-0.94) for the University of Tennessee Health Science Center cohort and of 0.96 (0.93-0.99) for the Atrium Health Wake Forest Baptist cohort. When we repeated our analysis by building a heart failure detection model using only lead-I electrocardiograms, we obtained similarly high detection accuracies, with areas under the curve of 0.73 and 0.93 for the University of Tennessee Health Science Center and Atrium Health Wake Forest Baptist cohorts, respectively. Artificial intelligence can accurately detect peripartum cardiomyopathy from electrocardiograms alone. A simple electrocardiographic artificial intelligence-based peripartum screening could result in a timelier diagnosis. Given that results with 1-lead electrocardiogram data were similar to those obtained using all 12 leads, future studies will focus on remote screening for peripartum cardiomyopathy using smartwatches that can capture single-lead electrocardiogram data.

AI-based preeclampsia detection and prediction with electrocardiogram data.

More than 76,000 women die yearly from preeclampsia and hypertensive disorders of pregnancy. Early diagnosis and management of preeclampsia can improve outcomes for both mother and baby. In this study, we developed artificial intelligence models to detect and predict preeclampsia from electrocardiograms (ECGs) in point-of-care settings. Ten-second 12-lead ECG data was obtained from two large health care settings: University of Tennessee Health Science Center (UTHSC) and Atrium Health Wake Forest Baptist (AHWFB). UTHSC data was split into 80% training and 20% holdout data. The model used a modified ResNet convolutional neural network, taking one-dimensional raw ECG signals comprising 12 channels as an input, to predict risk of preeclampsia. Sub-analyses were performed to assess the predictive accuracy for preeclampsia prediction within 30, 60, or 90 days before diagnosis. The UTHSC cohort included 904 ECGs from 759 females (78.8% African American) with a mean ± sd age of 27.3 ± 5.0 years. The AHWFB cohort included 817 ECGs from 141 females (45.4 African American) with a mean ± sd age of 27.4 ± 5.9 years. The cross-validated ECG-AI model yielded an AUC (95% CI) of 0.85 (0.77-0.93) on UTHSC holdout data, and an AUC (95% CI) of 0.81 (0.77-0.84) on AHWFB data. The sub-analysis of different time windows before preeclampsia prediction resulted in AUCs (95% CI) of 0.92 (0.84-1.00), 0.89 (0.81-0.98) and 0.90 (0.81-0.98) when tested on ECGs 30 days, 60 days and 90 days, respectively, before diagnosis. When assessed on early onset preeclampsia (preeclampsia diagnosed at <34 weeks of pregnancy), the model's AUC (95% CI) was 0.98 (0.89-1.00). We conclude that preeclampsia can be identified with high accuracy via application of AI models to ECG data.

VISUAL OUTCOMES AFTER INTERNAL LIMITING MEMBRANE PEELING VERSUS FLAP IN THE CLOSURE OF FULL-THICKNESS MACULAR HOLES.

To examine postoperative outcomes of internal limiting membrane peeling (ILMP) versus flap (ILMF) in the closure of full-thickness macular holes. Retrospective chart review of patients who underwent pars plana vitrectomy and gas tamponade with ILMP or ILMF to close full-thickness macular hole at the Atrium Health Wake Forest Baptist from January 2012 to October 2022 with at least 3 months follow-up. Main outcome measures were type 1 primary full-thickness macular hole closure and postoperative best-corrected visual acuity in mean logMAR. One hundred thirty and 30 eyes underwent ILMP and ILMF, respectively. There were no significant differences in baseline characteristics between the groups. Ninety-six percent of ILMP eyes and 90% of ILMF eyes achieved primary hole closure ( P = 0.29). Among all eyes with primary hole closure, best-corrected visual acuity at 1 year was not different between the groups, but when stratified by lens status, it was superior in the ILMP versus ILMF group in pseudophakic eyes: the estimated least-squares mean best-corrected visual acuity (Snellen equivalent) (95% confidence interval) was 0.42 (20/50) (0.34, 0.49) in the ILMP group and 0.71 (20/100) (0.50, 0.92) in the ILMF group. Internal limiting membrane peeling and ILMF techniques yielded similarly high full-thickness macular hole closure rates. In pseudophakic eyes with primary hole closure, ILMF eyes had worse best-corrected visual acuity at 1 year.

Guidelines for outpatient administration of naxitamab: Experience from Atrium Health Levine Children's Hospital.

In this publication, we will share our experience of AE management, provide guidance for appropriate staffing, and the discuss the importance of patient education when treating patients with R/R HR neuroblastoma using naxitamab. Approved treatments for patients with refractory and/or relapsed (R/R) high-risk (HR) neuroblastoma are limited, and there is a high unmet need for new treatment combinations. Naxitamab is a disialoganglioside 2 (GD2)-binding antibody that was approved by the United States Food and Drug Administration in 2020 for use in combination with granulocyte-macrophage colony-stimulating factor for the treatment of patients with R/R HR neuroblastoma in the bone and/or bone marrow and who have demonstrated a partial response, minor response, or stable disease with prior therapy. The pediatric oncology team at Atrium Health Levine Children's Hospital has successfully treated several patients with naxitamab both alone and in combination with chemotherapy, with no patients requiring unplanned overnight hospitalization and few severe adverse events (AEs). To accomplish this, the team at Levine Children's Hospital established standard operating procedures for naxitamab, a therapy defined as high acuity due to the potential for acute AEs with rapid onset and that benefits from continuous monitoring by a nursing team and a dedicated provider. This will provide a practical guide for institutions offering naxitamab to their patients, and ensure successful administration of this high acuity treatment in the outpatient setting.

Mitigating Cancer Disparities Through an Innovative Navigation Program.

The Office of Cancer Health Equity at the Atrium Health Wake Forest Baptist Comprehensive Cancer Center used a community-engaged approach to develop an innovative Population Health Navigation Program designed to improve access to cancer care and reduce cancer disparities.

Complete loss of the atrial natriuretic peptide-converting enzyme Corin and CHAF-LA syndrome: Implications to natriuretic peptide physiology and left atrium health.

Complete loss of the atrial natriuretic peptide-converting enzyme Corin and CHAF-LA syndrome: Implications to natriuretic peptide physiology and left atrium health.