Academic Health System Research Articles

Abstract 4136592: Comparative performance of PREVENT and Pooled Cohort Equations to predict cardiovascular disease in multisite healthcare populations

Introduction: The American Heart Association’s race-agnostic Predicting Risk of cardiovascular EVENTs (PREVENT) framework has emerged to better capture multilevel determinants of health in contemporary population. Whether PREVENT better predicts incident atherosclerotic cardiovascular disease (ASCVD) than the preceding Pooled Cohort Equations (PCE) is unknown in healthcare populations. We evaluated calibration and discrimination of the two equations across three major U.S. healthcare networks. Methods: We retrospectively evaluated electronic health records of 209,879 individuals (Mass General Brigham [MGB]: 118,801; Mount Sinai Healthcare System: 38,834; Penn Medicine: 52,424) aged 40-79 years and without prior ASCVD between 2010-2012. Calibration was assessed using discordance ratio between 10-year observed and predicted risk. Discrimination was evaluated by Harrell’s C-index. Results: Based on PREVENT, the mean [SD] estimated 10-year ASCVD risk ranged from 5.5 [4.7] % in MGB (mean age 56y, 42% female), 6.6 [5.6]% in Mount Sinai (59y, 54% female), and 6.4 [5.2]% in Penn (60y, 55% female). PREVENT substantially underestimated empirical 10-year ASCVD incidence in MGB (16.9%) and Mount Sinai (10.9%), whereas it closely mirrored the observed rate in Penn (5.5%). Both PCE and PREVENT had comparable discrimination C-index [95% CI] across MGB (0.69 [0.69-0.70] vs. 0.69 [0.68-0.69]), Mount Sinai (0.72 [0.71-0.73] vs. 0.70 [0.70-0.71]), and Penn (0.68 [0.67-0.69] vs. 0.67 [0.66-0.68]). Nevertheless, compared to PREVENT, PCE demonstrated better overall calibration in MGB (-70% vs. -42%) and Mount Sinai (-36% vs. +5%), whereas largely overestimated risk in Penn (+3% vs +94%). Calibration varied across demographics with PREVENT consistently underestimating risk in Hispanics, Asians, Black and African Americans across all sites, whereas PCE largely overestimated risk for White or multiracial Penn healthcare patients. Conclusions: The race-agnostic PREVENT model discriminated ASCVD incidence comparably or better than the race-informed PCE across three geographically distinct academic health systems in the Northeast. However, calibration metrics between the two models varied widely across health systems and demographics.

Physical development of infants born to patients with COVID-19 during pregnancy: 2 years of age

Background SARS-CoV-2 infection during pregnancy and pandemic circumstances could negatively impact infant development. This study aimed to investigate the physical development, common pediatric illness incidence, and healthcare utilization over the first 2 years of life of infants born to COVID+ and COVID- patients. Comparisons were also made with infants born pre-pandemic. Methods This is a retrospective observational study at a major academic health system in New York City. Participants include all infants born to birthing persons with SARS-CoV-2 infection during pregnancy (N = 758) and without (N = 9,345) from 03/01/2020 to 08/17/2022, infants born pre-pandemic (N = 3,221) from 03/01/2017 to 08/17/2019, and birthing persons of all infants. Results There were no differences in weight, length, or head circumference curves between pandemic infants born to COVID+ and COVID- patients over the first 2 years of life (p > 0.05, repeated ANOVA). Annualized incidence of illness occurrence and healthcare utilization were similar between groups. Compared to pre-pandemic infants, the length of pandemic (COVID-) infants was lower from birth to 9 months (p < 0.0001). Pandemic infants additionally had more adverse perinatal outcomes including increased stillbirth (0.75% vs. 0.12%, p = 0.0001) and decreased gestational age (38.41 ± 2.71 vs. 38.68 ± 2.55 weeks, Cohen’s d = −0.10, p < 0.0001), birthweight (2,597 ± 335 vs. 3,142 ± 643 g, Cohen’s d = −1.06, p < 0.0001), and birth length (48.08 ± 4.61 vs. 49.09 ± 3.93 cm, Cohen’s d = −0.24, p < 0.0001). Conclusions Birthing persons’ SARS-CoV-2 infection status, birthing persons’ profiles, and pandemic circumstances negatively affected perinatal outcomes, newborn physical development, and healthcare utilization. These findings draw clinical attention to the need to follow infants closely and implement enrichment to ensure optimal developmental outcomes.

Navigating the Path to Inclusion: Understanding Barriers and Facilitators to Clinical Trial Participation Among Chinese Older Adults in the United States with Multimorbidity.

Older adults with multimorbidity are underrepresented in clinical trials, with enrollment of Asians particularly low. Understand perspectives of US Chinese older adults regarding clinical trial participation. Focus group interviews analyzed using thematic analysis. Community/senior centers, academic health systems in Northern and Southern California, and a nationwide registry of Asian Americans/Pacific Islanders. Mandarin- and English-speaking Chinese adults aged ≥ 65years with multimorbidity. Themes related to barriers and facilitators of enrollment in clinical trials of medications. We conducted 12 focus groups: 7 with non-US-born and 5 with US-born Chinese older adults (n = 83 total). Mean age was 74years (SD = 5.9), 43 (51.8%) were female, and 47 (56.6%) Mandarin-speaking. US-born participants had greater educational attainment than non-US-born participants. Participants took a mean of 6.1 prescriptions (SD = 1.5). Barriers to participation in clinical trials of medications included lack of awareness of/exposure for patients and community-based Chinese physicians, preference for natural/traditional medicine, risk aversion and safety concerns, desire for privacy, and inconvenience. Trusted influences included physicians, hospitals/health systems, Asian/Chinese community centers, and family (for non-US-born participants). Suggestions to enhance participation included using language and culturally concordant materials/personnel, educating community-based Chinese physicians about clinical trials, involving patient-trusted physicians in recruitment, promoting trials on conditions common in Chinese people or for an existing condition, and financial incentives. US-born participants expressed greater understanding and willingness to join trials. All groups attributed low clinical trial enrollment to non-US-born Chinese adults. Chinese older adults perceived obstacles to clinical trial participation that could be mitigated by involving trusted physicians in recruitment, using language and culturally concordant materials/staff, and educating patients and community-based physicians. Recognition of differences in attitudes among US- and non-US-born Chinese people may be important to tailoring recruitment strategies.

Reflections on a Dobutamine Shortage in an Academic Health System: A Roadmap for Risk Reduction

Reflections on a Dobutamine Shortage in an Academic Health System: A Roadmap for Risk Reduction

Individual and community socioeconomic status and receipt of influenza vaccines among adult primary care patients in a large academic health system: 2017-2019

Individual and community socioeconomic status and receipt of influenza vaccines among adult primary care patients in a large academic health system: 2017-2019

Patient experiences of cancer genetic testing by non-genetics providers in the surgical setting.

As indications for hereditary cancer genetic testing (GT) for patients with breast cancer (BC) expand, breast surgery teams offer GT to newly diagnosed patients to inform surgical plans. There is, however, limited data on the experiences of patients undergoing cancer GT by non-genetic providers. This study used in-depth interviews with 21 women recently diagnosed with BC at a large academic health system to capture their experiences. Post-positivist codebook thematic analysis was used to identify major themes from the interviews. Participants reported an overall positive experience of this GT process, stating that they prefer GT at an existing appointment shortly after their diagnosis, even though they described the conversation as brief. Many participants indicated thinking about or desiring GT before the offer was made. Interestingly, most participants did not see surgical decision-making as the main reason for GT and were instead motivated by concern for relatives and to have complete information. Interview data indicated areas for improvement in patient-provider communication, and most participants agreed that additional reference information on GT in the form of written or video materials would be helpful. Offering GT at an initial breast surgery appointment is acceptable and desired by patients with a new BC diagnosis and should be considered as a way to increase access to GT for these patients. However, additional information for patients is needed to close gaps in communication and provide a trustworthy reference following a busy medical appointment.

Telephone Survey-Reported Perceptions of Telehealth Visits Among Black and Non-Black Patients Diagnosed With Cancer.

Empirical evidence underscores both benefits of telehealth visits and persistent disparities in use for Black adults. Guided by the Technology Acceptance Model, we evaluated perceptions regarding telehealth visits among Black and Non-Black adults receiving cancer care from two academic health systems. Between April 2022 and October 2023, a survey was conducted among adult patients treated for cancer in the past 3 years. Survey recruitment was stratified by visit type (telehealth or in-person) and race. Before a scheduled oncology visit, respondents completed a telephone survey to assess their perceptions of usefulness, ease of use, and attitudes toward telehealth visits. Frequency differences were estimated using chi-square tests, and prevalence differences using logistic regression. The survey was completed by 773 respondents (42% Black) with a 15% response rate. Black respondents were younger and less likely to report male sex, being married, having a college education, comfortable income, or confidence in medical and e-literacy. Compared with Non-Black respondents, Black respondents perceived telehealth visits as less useful for determining health needs (37% v 48%) and asking questions (9% v 15%). They thought that telehealth visits were less easy with regard to understanding their health care provider (14% v 21%) and the ability to connect (23% v 30%). Black respondents expressed more concern about internet access (26% v 15%), access to electronic devices (17% v 9%), and finding assistance for connecting (24% v 12%). They also found telehealth visits to be less private (19% v 34%). Racial differences in telehealth ease of use and usefulness perceptions highlight the need for oncology practices to couple connectivity support with effective communication strategies to avoid disparities in oncology telehealth services.

The Impact of Next-Generation Sequencing Workflows on Outcomes in Advanced Lung Cancer: A Retrospective Analysis at One Academic Health System.

Broad-based molecular testing with next-generation sequencing (NGS) is now the standard of care in advanced non-small cell lung cancer (NSCLC). Two approaches to molecular testing are (1) reflexive testing at pathologic NSCLC confirmation, often using an in-house molecular panel, and (2) send-out testing to private vendors, ordered by a clinician. This study explored the outcomes with reflex versus send-out testing. A retrospective chart review was conducted of patients diagnosed with de novo stage IV NSCLC in 2019 and 2020 at three hospitals in the same system, one academic hospital (Northwestern Memorial Hospital, or NMH) utilizing reflex, in-house NGS, and two community-based hospitals (Central DuPage Hospital, or CDH, and Delnor, or D) sending out tissue samples for testing. The outcomes assessed were the time from biopsy to results, biopsy to treatment, the incidence of first-line targetable mutations and the use of first-line targeted therapies, and overall survival. In total, 191 patients met the inclusion criteria, 85 at NMH, 106 at CDH + D, and in total, 131 in 2019 and 60 in 2020. The time to results was significantly shorter with reflexive NGS when compared with send-out testing; the time to treatment was also shorter but not statistically significant. At CDH + D, the time to results was significantly shorter with a limited panel than with comprehensive testing, but the time to treatment was similar. NGS testing rates were 95% at NMH and 84.5% at CDH + D (p = 0.009), with 31.0% at NMH receiving 1L targeted therapies versus 20.8% at CDH + D (p = 0.08). In 2019, the median time from biopsy to treatment was 35 days at NMH and 38 days at CDH and Delnor; in 2020, time to treatment was 26 days and 37 days, respectively. Overall survival trended longer in 2020 relative to 2019 independent of site. Reflexive NGS testing is associated with a shorter time to actionable results and higher rates of first-line targetable mutations than send-out testing. In practices with send-out testing, limited panels had slightly faster turnaround times but no difference in time to treatment. If resources allow, reflexive NGS should be considered in healthcare systems for patients with NSCLC.

Electronic Health Record Phenotyping of Pediatric Suicide-Related Emergency Department Visits

Suicide is a leading cause of death among young people. Accurate detection of self-injurious thoughts and behaviors (SITB) underpins equity in youth suicide prevention. To compare methods of detecting SITB using structured electronic health information and measure algorithmic performance across demographics. This cross-sectional study used medical records among youths aged 6 to 17 years with at least 1 mental health-related emergency department (ED) visit in 2017 to 2019 to an academic health system in Southern California serving 787 000 unique individuals each year. Analyses were conducted between January and September 2023. Multiexpert electronic health record review ascertained the presence of SITB using the Columbia Classification Algorithm of Suicide Assessment. Random forest classifiers with nested cross-validation were developed using (1) International Statistical Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes for nonfatal suicide attempt and self-harm and chief concern and (2) all available structured data, including diagnoses, medications, and laboratory tests. Detection performance was assessed overall and stratified by age group, sex, and race and ethnicity. The sample comprised 2702 unique youths with an MH-related ED visit (1384 youths who identified as female [51.2%]; 131 Asian [4.8%], 266 Black [9.8%], 719 Hispanic [26.6%], 1319 White [48.8%], and 233 other race [8.6%]; median [IQR] age, 14 [12-16] years), including 898 children and 1804 adolescents. Approximately half of visits were related to SITB (1286 visits [47.6%]). Sensitivity of SITB detection using only codes and chief concern varied by age group and increased until age 15 years (6-9 years: 59.3% [95% CI, 48.5%-69.5%]; 10-12 years: 69.0% [95% CI, 63.8%-73.9%]; 13-15 years: 88.4% [95% CI, 85.1%-91.2%]; 16-17 years: 83.1% [95% CI, 79.1%-86.6%]), while specificity remained constant. The area under the receiver operating characteristic curve (AUROC) was lower among preadolescents (0.841 [95% CI, 0.815-0.867]) and male (0.869 [95% CI, 0.848-0.890]), Black (0.859 [95% CI, 0.813-0.905]), and Hispanic (0.861 [95% CI, 0.831-0.891]) youths compared with adolescents (0.925 [95% CI, 0.912-0.938]), female youths (0.923 [95% CI, 0.909-0.937]), and youths of other races and ethnicities (eg, White: 0.901 [95% CI, 0.884-0.918]). Augmented classification (ie, using all available structured data) outperformed classification with codes and chief concern alone (AUROC, 0.975 [95% CI, 0.968-0.980] vs 0.894 [95% CI, 0.882-0.905]; P < .001). In this study, diagnostic codes and chief concern underestimated SITB prevalence, particularly among minoritized youths. These results suggest that priority on algorithmic fairness in suicide prevention strategies must extend to accurate detection of youths with suicide-related emergencies.

Intra-individual Variability in Lipoprotein(a) Levels: Findings from a Large Academic Health System Population.

Lipoprotein(a) [Lp(a)] levels are known to be mainly genetically determined. However, only scarce data are available on the intra-individual variability of Lp(a) levels across time. We included adult patients (≥18 years old) who had baseline and follow-up Lp(a) measurements (between 1997 and 2024) with a minimum of one year apart. Patients were categorized into three groups as follows: normal (<30 mg/dL), borderline (30 to 50 mg/dL) and high Lp(a) (≥50 mg/dL). Multivariable logistic regression was conducted to assess the predictors of the intra-individual changes in Lp(a) ≥10 mg/dL. A total of 11,669 individuals (median age: 54 years, 60% males) were included in our analysis, with median time between measurements of 4.5 years (IQR: 2.2, 10.6). The median Lp(a) was 16 mg/dL (IQR: 7, 52) at baseline, compared with 15 mg/dL (IQR: 7, 52) at follow-up. At follow-up, 96.4% of individuals with normal Lp(a) and 89.9% with high Lp(a) remained in their categories, while 51.2% with borderline Lp(a) changed their category. Of the included population, 24.9% had an intra-individual Lp(a) change ≥10 mg/dL. Female sex (p <0.001), history of ASCVD (p=0.003), statin therapy (p=0.003) and elevated low-density lipoprotein cholesterol (LDL-C) levels ≥100 mg/dL (p <0.001) were significantly associated with higher odds of intra-individual Lp(a) changes ≥10 mg/dL. Lipoprotein(a) levels were generally stable over time; however, patients with borderline levels may require more than one Lp(a) measurement, especially if they are females, have a history of ASCVD, have elevated LDL-C levels or are on statins therapy.

Referral, enrollment, and health care use in a comprehensive patient-centered management program for osteoarthritis of the hip and knee

ObjectiveOsteoarthritis management programs (OAMPs) have become a more common way to deliver patient-centered care. However, there is limited information on real-world use of these programs to guide implementation, payment policy, accessibility, and scaling in the United States. This paper describes 5-year use metrics for the Duke Joint Health Program, an OAMP embedded within a US academic health system. MethodThis analysis includes patients referred into the Program between October 2017 and April 2022. We generated descriptive statistics of referral and enrollment totals, demographics and patient-reported measures of enrollees, retention and healthcare use metrics (e.g., office visit frequency), and data capture rates for patient-reported outcomes. ResultsDuring the study period, 6863 patients were referred to the program and 4162 (61 ​%) enrolled. We observed statistically significant differences between those who did and did not enroll by age (mean difference ​± ​SE: 2.49 ​± ​2.8 years), sex (70.0 ​% vs 67.7 ​% female), race (65.1 ​% vs 55.3 ​% Caucasian/White), employment status (50.0 ​% vs 40.2 ​% retired), and insurance type (53.5 ​% vs 47.0 ​% Medicare). The median (Q1, Q3) number of visits was 2 (1, 4) and ranged from 1 to 67. The median (Q1, Q3) number of days from first to last program visit was 23 (0, 84) days. Questionnaire completion rates were 72 ​% at baseline, 46 ​% at 6 weeks, 39 ​% at 3 months, and 40 ​% at 6 and 12 months. ConclusionFindings can guide the planning, development, and implementation of future OAMPs and inform policies to ensure programs are accessible and equitable.

Improving the Accuracy and Precision of Disease Identification When Utilizing Ehr Data for Research: the Case for Hepatocellular Carcinoma.

Objective We assessed the performance of ICD codes to identify patients with hepatocellular carcinoma (HCC) in a large academic health system and determined whether employing an algorithm using a combination of ICD codes could deliver higher accuracy and precision than single ICD codes in identifying HCC cases using electronic health record (EHR) data. Results The use of a single ICD code entry for HCC (ICD-9-CM 155.0 or ICD-10-CM C22.0) in our cohort of 1,007 established ambulatory care patients with potential HCC yielded 58% false positives (not true HCC cases) based on chart reviews. We developed an ICD code-based algorithm that prioritized positive predictive value (PPV), F-score, and accuracy to minimize false positives and negatives. The highest performing algorithm required at least 10 ICD code entries for HCC and the sum of ICD code entries for HCC to exceed the sum of ICD code entries for non-HCC malignancies. The algorithm demonstrated high performance (PPV 97.4%, F-score 0.92, accuracy 94%), which was internally validated (PPV 92.3%, F-score 0.90, accuracy 91%) using a separate sample of potential HCC cases. Our findings support the need to assess the accuracy and precision of ICD codes before using EHR data to study HCC more broadly.

Special Issue on Informatics Education: Vanderbilt Clinical Informatics Center Education Strategy: To Infinity and Beyond!

The Vanderbilt Clinical Informatics Center (VCLIC) is based in the Department of Biomedical Informatics (DBMI) and operates across Vanderbilt University Medical Center (VUMC) and Vanderbilt University (VU) with a goal of enabling and supporting clinical informatics research and practice. VCLIC supports several types of applied clinical informatics teaching, including teaching of students in courses, professional education for staff and faculty throughout VUMC, and workshops and conferences that are open to the public. In this paper, we provide a detailed accounting of our center and institution's methods of educating and training faculty, staff, students, and trainees from across the academic institution and health system on clinical informatics topics, including formal training programs and informal applied learning sessions. Through a host of informal learning events, such as workshops, seminars, conference-style events, bite-size instructive videos, and hackathons, as well as several formal education programs, such as the Clinical Informatics Graduate Course, Master's in Applied Clinical Informatics, Medical Student Integrated Science Course, Graduate Medical Education Elective, and Fellowship in Clinical Informatics, VCLIC and VUMC provide opportunities for faculty, students, trainees, and even staff to engage with Clinical Informatics topics and learn related skills. The described programs have trained hundreds of participants from across the academic and clinical enterprises. Of the VCLIC-held events, the majority of attendees indicated through surveys that they were satisfied, with the average satisfaction score being 4.63/5, and all events averaging a satisfaction score of greater than 4. Across the 20 events VCLIC has held, our largest audiences are DBMI, HealthIT operational staff, and students from the medical and nursing schools. VCLIC has created and delivered a successful suite of formal and informal educational events and programs to disseminate clinical informatics knowledge and skills to learners across the academic institution and healthcare system.

Associations Between Patients' Primary Language and Outcomes and Patient Engagement Process Metrics for Total Joint Arthroplasty.

Previous research has indicated language-based disparities in outcomes after total hip and knee arthroplasty. In this study, we examined the relationship between primary spoken language and outcomes and patient engagement process metrics in a multihospital academic health system. This retrospective cohort study included patients who underwent elective primary total hip or knee arthroplasty in 2018 to 2022. Primary language was categorized as English, Spanish, and Other. Associations were examined between language and perioperative outcomes (same-day discharge, extended hospital length of stay, nonhome discharge, 30-day hospital returns, 90-day readmissions, and 90-day combined complications), and engagement-related metrics (preoperative joint class attendance and patient-reported outcome measure [PROM] completion). We report adjusted odds ratios (ORs) and 95% confidence intervals (CIs). Among 8,220 patients (7.9% Spanish and 5.1% Other non-English), we found no notable associations between language and same-day discharge, nonhome discharge, 30-day hospital returns, 90-day readmissions, or 90-day complications. Other (versus English) language was significantly associated with decreased odds of extended length of stay (OR 0.49, 95% CI, 0.32 to 0.76, P = 0.001). In patient engagement, Other (versus English) language was associated with significantly decreased odds of preoperative class attendance (OR 0.72, 95% CI, 0.54 to 0.95, P = 0.02). Spanish (versus English) primary language was significantly associated with decreased odds of preoperative and 1-year PROM completion (OR 0.34 to 0.75, P ≤ 0.001), and Other (versus English) primary language was significantly associated with decreased odds of preoperative PROM completion (OR 0.59, 95% CI, 0.47 to 0.73, P < 0.001). Our findings highlight language-based disparities in patient engagement metrics but comparable outcomes. Differences in findings compared with past literature may be related to institution-specific language-related support systems. There is a need for additional resources to support patients' active participation in their care, regardless of their primary spoken language.

Race and Ethnicity and Clinician Linguistic Expressions of Doubt in Hospital Admission Notes

Stigmatizing language in electronic health records (EHRs) can generate or reinforce negative stereotypes about patients from minoritized groups and erode their trust and experience. However, less is known about the use of such language to cast doubt on patient clinical history in hospital settings. To compare the prevalence of language expressing doubt about patient clinical history (eg, experiences and symptoms) in hospital admission notes by patient race and ethnicity. In this cohort study, natural language processing tools were applied to analyze 54 936 admission notes from 1249 clinicians about patients aged 18 years or older at admission and hospitalized at an academic health system in the Northeast US between January 1, 2018, and February 28, 2023. Data were analyzed from September 1, 2022, to July 31, 2023. Patient race and ethnicity (non-Hispanic Black, non-Hispanic White, and racial and ethnic minoritized groups excluding non-Hispanic Black [includes multiple racial and ethnic groups, such as Hispanic and Asian]) as recorded in the EHR based on self-report or registrar determination. Binary indicator for at least 1 term casting doubt on patient clinical history was defined using epistemic stance, a linguistic construct expressing a writer's degree of certainty in information. Terms were manually validated via iterative review of notes by the study team. Among 56 325 admission notes (mean [SD] age of patients, 55.9 [19.0] years; 30 913 notes among female patients [54.9%]; 25 649 notes among non-Hispanic Black patients [45.5%], 26 442 notes among non-Hispanic White patients [46.9%], and 2985 notes among members of racial and ethnic minoritized groups excluding non-Hispanic Black patients [5.3%]), we analyzed 54 936 admission notes that had no missing data. Among all analyzed admission notes, 39 023 notes (71.0%) contained doubt language. Notes written about non-Hispanic Black patients had increased odds of containing at least 1 word or phrase of doubt (adjusted odds ratio, 1.21; 95% CI, 1.14-1.28; P < .001) compared with notes among non-Hispanic White patients. Compared with notes among non-Hispanic White patients, notes written about members of racial and ethnic minoritized groups excluding non-Hispanic Black patients had similar odds of containing at least 1 term of doubt. In this study, language casting doubt on patient clinical history was more likely to be found in notes of non-Hispanic Black patients. These results suggest that with the implementation of policies allowing patients full access to their health records, examining clinical documentation standards may be associated with improved patient experience.

Assessing for race, ethnicity, and socioeconomic disparities in central line-associated bloodstream infection risk in a large academic health system.

To examine the relationship between race and ethnicity and central line-associated bloodstream infections (CLABSI) while accounting for inherent differences in CLABSI risk related to central venous catheter (CVC) type. Retrospective cohort analysis. Acute care facilities within an academic healthcare system. Adult inpatients from January 2012 through December 2017 with CVC present for ≥2 contiguous days. We describe variability in demographics, comorbidities, CVC type/configuration, and CLABSI rate by patient's race and ethnicity. We estimated the unadjusted risk of CLABSI for each demographic and clinical characteristic and then modelled the effect of race on time to CLABSI, adjusting for total parenteral nutrition use and CVC type. We also performed exploratory analysis replacing race and ethnicity with social vulnerability index (SVI) metrics. 32,925 patients with 57,642 CVC episodes met inclusion criteria, most of which (51,348, 89%) were among non-Hispanic White or non-Hispanic Black patients. CVC types differed between race/ethnicity groups. However, after adjusting for CVC type, configuration, and indication in an adjusted cox regression, the risk of CLABSI among non-Hispanic Black patients did not significantly differ from non-Hispanic White patients (adjusted hazard ratio [aHR] 1.19; 95% confidence interval [CI]: 0.94, 1.51). The odds of having a CLABSI among the most vulnerable SVI subset compared to the less vulnerable was no different (odds ratio [OR] 0.95; 95% CI: 0.75-1.2). We did not find a difference in CLABSI risk between non-Hispanic White and non-Hispanic Black patients when adjusting for CLABSI risk inherent in type and configuration of CVC.

Variation in Overactive Bladder Treatment Type by Social Determinants of Health

Importance There is limited understanding of the relationship between social determinants of health (SDOH) and types of overactive bladder/urgency urinary incontinence (OAB/UUI) treatments. Objectives Our objective was to determine if OAB/UUI treatment type differs by SDOH, including insurance and estimated median household income (EMHI). Study Design This was a cross-sectional study of adult patients assigned female at birth with OAB/UUI, identified from 2017 to 2022 within a tertiary academic health system. The primary outcome was any OAB/UUI treatment provision. Secondary outcomes included specific treatments and specialty care. We examined associations between SDOH and OAB/UUI treatments using univariable analysis and multivariable logistic regression. Results Among 33,002 patients with OAB/UUI, 15.4% received treatment. Treatment provision was associated with American Indian/Alaska Native or Black/African American race, Latina ethnicity, Medicaid, low-income Medicare, or Medicare Advantage (vs private insurance) and EMHI ranging from $26,500 to $100,000 (compared with EMHI &gt;$100,000). Patients with Black/African American race, Medicaid or no insurance, and EMHI ≤$80,000 had higher odds of anticholinergic prescription and lower odds of β3-agonist prescription. Advanced therapy was associated with traditional Medicare, Medicare Advantage, and EMHI $26,500–$100,000 and negatively associated with Asian race. Specialty care was associated with Black/African American race, Medicaid, Medicare Advantage, no insurance, and EMHI ≤$80,000. Conclusions Overactive bladder/urgency urinary incontinence treatment provision was associated with American Indian/Alaska Native and Black/African American race, Latina ethnicity, Medicaid, low-income Medicare, or Medicare Advantage insurance. Patients with Black/African American race and with lower EMHI had higher odds of receiving anticholinergic therapy and lower odds of receiving β3-agonist medication, despite higher likelihood of specialty care.

Assessing the impact of social determinants of health on diabetes severity and management.

Adverse Social Determinants of Health (SDoH) are considered major obstacles to effective management of type-2 diabetes. This study aims to quantify the impact of SDoH factors on diabetes management outcomes. We quantified the joint impact of multiple SDoH by applying a self-control case series method-which accounts for confounding by using individuals as their own control-to electronic health record data from an academic health system in Maryland. We found a consistent increase in HbA1c levels associated with SDoH across alternative study designs. The estimated total contributions of SDoH ranged 0.014-0.065 across the alternative designs. Transportation issues demonstrated particularly significant contributions, with estimates of 0.077-0.144. When assuming SDoH's risk window to be ±45 days, for example, the total contribution was estimated to be 0.065 (95% CI [0.010, 0.120]) increase in HbA1c and the transportation issues' contribution 0.134 (95% CI [0.020, 0.249]). Our result suggests that reducing transportation barriers may be an effective SDoH intervention strategy for diabetes management; however, the clinical impact of such interventions warrants further investigation.

Leveraging machine learning to enhance appointment adherence at a novel post-discharge care transition clinic.

This study applies predictive analytics to identify patients at risk of missing appointments at a novel post-discharge clinic (PDC) in a large academic health system. Recognizing the critical role of appointment adherence in the success of new clinical ventures, this research aims to inform future targeted interventions to increase appointment adherence. We analyzed electronic health records (EHRs) capturing a wide array of demographic, socio-economic, and clinical variables from 2168 patients with scheduled appointments at the PDC from September 2022 to August 2023. Logistic regression, decision trees, and eXtreme Gradient Boosting (XGBoost) algorithms were employed to construct predictive models for appointment adherence. The XGBoost machine learning model outperformed logistic regression and decision trees with an area under the curve (AUC) of 72% vs 65% and 67%, respectively, in predicting missed appointments, despite limited availability of historical data. Key predictors included patient age, number of days between appointment scheduling and occurrence, insurance status, marital status, and mental health and cardiac disease conditions. Findings underscore the potential of machine learning predictive analytics to significantly enhance patient engagement and operational efficiency in emerging healthcare settings. Optimizing predictive models can help balance the early identification of patients at risk of non-adherence with the efficient allocation of resources. The study highlights the potential value of employing machine learning techniques to inform interventions aimed at improving appointment adherence in a post-discharge transition clinic environment.

Association between telehealth use in oncology and downstream utilization at a large academic health system.

While telemedicine has been beneficial in oncology by reducing infectious exposure and improving access for patients with poor functional status, it also has intrinsic limitations, including the inability to perform a physical exam, which could lead to increased downstream utilization in this population at high risk of medical decompensation. We conducted a retrospective cohort study investigating the relationship between telemedicine use in oncology and subsequent outpatient oncology encounters, emergency department (ED) visits, and hospitalizations. We included outpatient oncology encounters, including telemedicine and in-person visits, occurring between 1 January 2018 and 31 December 2022 at a large academic health system. Unadjusted descriptive statistics and multiple linear regressions were used to estimate subsequent outpatient oncology visits, ED visits, and hospitalizations within 30 days of an index visit based on modality (telemedicine versus in-person). The multiple regressions were adjusted for various demographic and clinical characteristics, including palliative care visits, baseline utilization, recent chemotherapy, and comorbidities. Our cohort included 63,722 patients with 689,356 outpatient encounters, of which 639,217 (92.7%) were in-person and 50,139 (7.3%) were telemedicine visits. Patients on average had 0.91 outpatient oncology visits, 0.04 ED visits, and 0.05 hospitalizations within 30 days following an index encounter. In our adjusted analyses, telemedicine was associated with 13.7 fewer downstream outpatient oncology visits (95% CI 12.5-14.9; p < 0.001) per 100 index encounters, 0.7 fewer ED visits (95% CI 0.4-1.0; p < 0.001) per 100 index encounters and 0.9 fewer hospitalizations (95% CI 0.6-1.3; p < 0.001) per 100 index encounters compared to in-person visits. Contrary to our hypothesis, oncology patients who had a telemedicine visit had fewer follow-up outpatient oncology encounters, ED visits and hospitalizations after 30 days than those with in-person visits. Future studies should further investigate the efficacy of telemedicine in oncology and outline specific scenarios for appropriate use in this and other populations.