Faculty Bibliography

BACKGROUND:Hepatitis C virus (HCV) reinfection rates are substantially higher than primary infection rates among men who have sex with men (MSM) with human immunodeficiency virus (HIV) in European cohorts. The behaviors mediating this high rate of transmission among MSM are poorly characterized. METHODS:We performed a prospective cohort study in New York City (NYC) of MSM with HIV who cleared HCV to determine the incidence of and risk factors for HCV reinfection. We assessed the risk behaviors for primary HCV in NYC: receipt of semen in the rectum, and sexualized methamphetamine use, along with route of use. Multivariable analysis was performed with Andersen-Gill extension of the Cox proportional hazards model. RESULTS:From 2000 through 2018, among 304 MSM with HIV who cleared HCV, 42 reinfections occurred over 898 person-years, for an incidence rate of 4.7 per 100 person-years. Assessing 1245 postclearance visits, only receipt of semen into the rectum was associated with reinfection (hazard ratio, 9.7 [95% confidence interval: 3.3-28.3], P < .001); methamphetamine use was not. CONCLUSIONS:The high HCV reinfection rate over almost 2 decades demonstrates that sexual transmission of HCV is not inefficient or unusual and that direct-acting antiviral treatment is not sufficient for HCV elimination among MSM in NYC. The contrasts between both the rates of and risk factors for primary and HCV reinfection suggest that HCV prevalence is highly heterogenous among sexual networks and that sexualized methamphetamine use, rather than mediating transmission, is instead a surrogate marker for the highest HCV prevalence networks. As neither condoms nor treatment have been successful strategies for HCV prevention in NYC, novel interventions are needed to stem this sexually transmitted HCV epidemic.

BACKGROUND:Antibiotics are a cornerstone in management of intra-abdominal abscesses in Crohn's disease (CD). Yet, the optimal route of antibiotic administration is poorly studied. We aimed to compare surgical and nonsurgical readmission outcomes for patients hospitalized for intra-abdominal abscesses from CD discharged on oral (PO) or intravenous (IV) antibiotics. METHODS:Data for patients with CD hospitalized for an intra-abdominal abscess were obtained from 3 institutions from January 2010 to December 2020. Baseline patient characteristics were obtained. Primary outcomes of interest included need for surgery and hospital readmission within 1 year from hospital discharge. We used multivariable logistic regression models and Cox regression analysis to adjust for abscess size, history of prior surgery, history of penetrating disease, and age. RESULTS:We identified 99 patients discharged on antibiotics (PO = 74, IV = 25). Readmissions related to CD at 12 months were less likely in the IV group (40% vs 77% PO, P = .01), with the IV group demonstrating a decreased risk for nonsurgical readmissions over time (hazard ratio, 0.376; 95% confidence interval, 0.176-0.802). Requirement for surgery was similar between the groups. There were no differences in time to surgery between groups. CONCLUSIONS:In this retrospective, multicenter cohort of CD patients with intra-abdominal abscess, surgical outcomes were similar between patients receiving PO vs IV antibiotics at discharge. Patients treated with IV antibiotics demonstrated a decreased risk for nonsurgical readmission. Further prospective trials are needed to better delineate optimal route of antibiotic administration in patients with penetrating CD.

Discrimination toward trainees is a pervasive problem, with surveys showing it is often perpetrated by patients. For several years, residents and faculty in an internal medicine residency have participated in a workshop offering a framework for responding to discriminatory behavior by patients. As part of a larger effort to reinforce this teaching and promote an inclusive environment, the authors pursued a multi-pronged simulation curriculum that could be incorporated into graduate medical education programs across their institution. First, the authors conducted trainee and faculty focus groups to better understand their experiences. Qualitative data was collected, including recommendations for responding to discrimination, characteristics of the most common experiences, and trainees' own ideas for worthwhile simulation scenarios. Trainees and faculty were then brought together in a collaborative process to co-develop simulation cases that were later implemented in curricula across multiple learner levels, specialties, and contexts. Participants in these simulations reported improved comfort in responding to discrimination in the moment and/or in the wake of such incidents. Through trainee-faculty collaboration, the project yielded authentic and impactful simulation experiences for learners, while also giving trainees an opportunity to turn previous trauma into constructive learning opportunities that promote an inclusive environment.

Black and Hispanic adolescents in New York City experience high rates of pregnancy and sexually transmitted infections. A comprehensive understanding of the complexity of adolescent sexual health and family planning decisions and experiences can provide insights into the sustained disparities and inform practice, policy, and future research. The goal of this study was to explore and analyze this complexity, centering Black and Hispanic adolescent mothers as the experts on sexual and reproductive experiences from pre-pregnancy through parenting. As part of formative research for a human-centered design study, we interviewed 16 Black and Hispanic adolescent mothers living in New York City. Using situational analysis, we mapped relationships, discourse, and social structures to explore the various factors that inform adolescent sexual health decisions, in particular choices about contraception. Situational analysis found that, besides interpersonal factors, organizations and non-human elements like social media and physical birth control devices affected adolescent family planning in three social arenas: home, healthcare, and school. Within and across these arenas, adolescents lacked consistent sexual health education and contraceptive counseling and faced gendered expectations of their behaviors. Participants described parents and healthcare providers as most responsible for providing sexual health counseling yet described parents as uncomfortable or overreactive and healthcare workers as paternalistic and biased. A consideration of the many factors that inform sexual health decision-making and recognition of adolescents"™ desire for parents and healthcare providers to be a source of education can address health disparities and promote adolescent sexual health and wellbeing.

Electronic health record proficiency is critical for health professionals to deliver and document patient care. There is scarce research on this topic within undergraduate nursing student populations. The purpose of this study is to describe the psychometric evaluation of the Digital Literacy, Usability, and Acceptability of Technology Instrument for Healthcare. A cross-sectional pilot study for psychometric evaluation of the instrument was conducted using data collected through an emailed survey. Exploratory factor analysis, inter-item and adjusted item-total correlations, and Cronbach's α calculated subscale reliability. A total of 297 nursing students completed the survey. A seven-factor structure best fit the data: technology use-engagement, technology use-confidence, technology use-history, electronic health record-ease of use, electronic health record-comparability, and electronic health record-burden. Cronbach's α indicated good to very good internal consistency (α = .68 to .89). The instrument effectively measured digital literacy, acceptance, and usability of an electronic health record and may be implemented with good to very good reliability across varied healthcare simulation and training experiences.

Ensuring reliability of Large Language Models (LLMs) in clinical tasks is crucial. Our study assesses two state-of-the-art LLMs (ChatGPT and LlaMA-2) for extracting clinical information, focusing on cognitive tests like MMSE and CDR. Our data consisted of 135,307 clinical notes (Jan 12th, 2010 to May 24th, 2023) mentioning MMSE, CDR, or MoCA. After applying inclusion criteria 34,465 notes remained, of which 765 underwent ChatGPT (GPT-4) and LlaMA-2, and 22 experts reviewed the responses. ChatGPT successfully extracted MMSE and CDR instances with dates from 742 notes. We used 20 notes for fine-tuning and training the reviewers. The remaining 722 were assigned to reviewers, with 309 each assigned to two reviewers simultaneously. Inter-rater-agreement (Fleiss' Kappa), precision, recall, true/false negative rates, and accuracy were calculated. Our study follows TRIPOD reporting guidelines for model validation. For MMSE information extraction, ChatGPT (vs. LlaMA-2) achieved accuracy of 83% (vs. 66.4%), sensitivity of 89.7% (vs. 69.9%), true-negative rates of 96% (vs 60.0%), and precision of 82.7% (vs 62.2%). For CDR the results were lower overall, with accuracy of 87.1% (vs. 74.5%), sensitivity of 84.3% (vs. 39.7%), true-negative rates of 99.8% (98.4%), and precision of 48.3% (vs. 16.1%). We qualitatively evaluated the MMSE errors of ChatGPT and LlaMA-2 on double-reviewed notes. LlaMA-2 errors included 27 cases of total hallucination, 19 cases of reporting other scores instead of MMSE, 25 missed scores, and 23 cases of reporting only the wrong date. In comparison, ChatGPT's errors included only 3 cases of total hallucination, 17 cases of wrong test reported instead of MMSE, and 19 cases of reporting a wrong date. In this diagnostic/prognostic study of ChatGPT and LlaMA-2 for extracting cognitive exam dates and scores from clinical notes, ChatGPT exhibited high accuracy, with better performance compared to LlaMA-2. The use of LLMs could benefit dementia research and clinical care, by identifying eligible patients for treatments initialization or clinical trial enrollments. Rigorous evaluation of LLMs is crucial to understanding their capabilities and limitations.

There is a growing literature exploring the placebo response within specific mental disorders, but no overarching quantitative synthesis of this research has analyzed evidence across mental disorders. We carried out an umbrella review of meta-analyses of randomized controlled trials (RCTs) of biological treatments (pharmacotherapy or neurostimulation) for mental disorders. We explored whether placebo effect size differs across distinct disorders, and the correlates of increased placebo effects. Based on a pre-registered protocol, we searched Medline, PsycInfo, EMBASE, and Web of Knowledge up to 23.10.2022 for systematic reviews and/or meta-analyses reporting placebo effect sizes in psychopharmacological or neurostimulation RCTs. Twenty meta-analyses, summarising 1,691 RCTs involving 261,730 patients, were included. Placebo effect size varied, and was large in alcohol use disorder (g = 0.90, 95% CI [0.70, 1.09]), depression (g = 1.10, 95% CI [1.06, 1.15]), restless legs syndrome (g = 1.41, 95% CI [1.25, 1.56]), and generalized anxiety disorder (d = 1.85, 95% CI [1.61, 2.09]). Placebo effect size was small-to-medium in obsessive-compulsive disorder (d = 0.32, 95% CI [0.22, 0.41]), primary insomnia (g = 0.35, 95% CI [0.28, 0.42]), and schizophrenia spectrum disorders (standardized mean change = 0.33, 95% CI [0.22, 0.44]). Correlates of larger placebo response in multiple mental disorders included later publication year (opposite finding for ADHD), younger age, more trial sites, larger sample size, increased baseline severity, and larger active treatment effect size. Most (18 of 20) meta-analyses were judged 'low' quality as per AMSTAR-2. Placebo effect sizes varied substantially across mental disorders. Future research should explore the sources of this variation. We identified important gaps in the literature, with no eligible systematic reviews/meta-analyses of placebo response in stress-related disorders, eating disorders, behavioural addictions, or bipolar mania.