Enhancing healthcare accessibility through telehealth for justice impacted individuals.

Telehealth is a great tool that makes accessing healthcare easier for those incarcerated and can help with reentry into the the community. Justice impacted individuals face many hardships including adverse health outcomes which can be mitigated through access to telehealth services and providers. During the federally recognized COVID-19 pandemic the need for accessible healthcare was exacerbated and telehealth use surged. While access to telehealth should be considered a necessity, there are many challenges and barriers for justice impacted individuals to be able to utilize this service. This perspective examines aspects of accessibility, pandemic, policy, digital tools, and ethical and social considerations of telehealth in correctional facilities. Carceral facilities should continue to innovate and invest in telehealth to revolutionize healthcare delivery, and improve health outcomes for justice impacted individuals.

Recommendations to promote fairness and inclusion in biomedical AI research and clinical use.

OBJECTIVE: Understanding and quantifying biases when designing and implementing actionable approaches to increase fairness and inclusion is critical for artificial intelligence (AI) in biomedical applications. METHODS: In this Special Communication, we discuss how bias is introduced at different stages of the development and use of AI applications in biomedical sciences and health care. We describe various AI applications and their implications for fairness and inclusion in sections on 1) Bias in Data Source Landscapes, 2) Algorithmic Fairness, 3) Uncertainty in AI Predictions, 4) Explainable AI for Fairness and Equity, and 5) Sociological/Ethnographic Issues in Data and Results Representation. RESULTS: We provide recommendations to address biases when developing and using AI in clinical applications. CONCLUSION: These recommendations can be applied to informatics research and practice to foster more equitable and inclusive health care systems and research discoveries.

Leveraging Electronic Health Records to Assess Residential Mobility Among Veterans in the Veterans Health Administration.

BACKGROUND: Residential mobility, or a change in residence, can influence health care utilization and outcomes. Health systems can leverage their patients' residential addresses stored in their electronic health records (EHRs) to better understand the relationships among patients' residences, mobility, and health. The Veteran Health Administration (VHA), with a unique nationwide network of health care systems and integrated EHR, holds greater potential for examining these relationships. METHODS: We conducted a cross-sectional analysis to examine the association of sociodemographics, clinical conditions, and residential mobility. We defined residential mobility by the number of VHA EHR residential addresses identified for each patient in a 1-year period (1/1-12/31/2018), with 2 different addresses indicating one move. We used generalized logistic regression to model the relationship between a priori selected correlates and residential mobility as a multinomial outcome (0, 1, ≥2 moves). RESULTS: In our sample, 84.4% (n=3,803,475) veterans had no move, 13.0% (n=587,765) had 1 move, and 2.6% (n=117,680) had ≥2 moves. In the multivariable analyses, women had greater odds of moving [aOR=1.11 (95% CI: 1.10,1.12) 1 move; 1.27 (1.25,1.30) ≥2 moves] than men. Veterans with substance use disorders also had greater odds of moving [aOR=1.26 (1.24,1.28) 1 move; 1.77 (1.72,1.81) ≥2 moves]. DISCUSSION: Our study suggests about 16% of veterans seen at VHA had at least 1 residential move in 2018. VHA data can be a resource to examine relationships between place, residential mobility, and health.

Large Language Models for Social Determinants of Health Information Extraction from Clinical Notes - A Generalizable Approach across Institutions.

The consistent and persuasive evidence illustrating the influence of social determinants on health has prompted a growing realization throughout the health care sector that enhancing health and health equity will likely depend, at least to some extent, on addressing detrimental social determinants. However, detailed social determinants of health (SDoH) information is often buried within clinical narrative text in electronic health records (EHRs), necessitating natural language processing (NLP) methods to automatically extract these details. Most current NLP efforts for SDoH extraction have been limited, investigating on limited types of SDoH elements, deriving data from a single institution, focusing on specific patient cohorts or note types, with reduced focus on generalizability. This study aims to address these issues by creating cross-institutional corpora spanning different note types and healthcare systems, and developing and evaluating the generalizability of classification models, including novel large language models (LLMs), for detecting SDoH factors from diverse types of notes from four institutions: Harris County Psychiatric Center, University of Texas Physician Practice, Beth Israel Deaconess Medical Center, and Mayo Clinic. Four corpora of deidentified clinical notes were annotated with 21 SDoH factors at two levels: level 1 with SDoH factor types only and level 2 with SDoH factors along with associated values. Three traditional classification algorithms (XGBoost, TextCNN, Sentence BERT) and an instruction tuned LLM-based approach (LLaMA) were developed to identify multiple SDoH factors. Substantial variation was noted in SDoH documentation practices and label distributions based on patient cohorts, note types, and hospitals. The LLM achieved top performance with micro-averaged F1 scores over 0.9 on level 1 annotated corpora and an F1 over 0.84 on level 2 annotated corpora. While models performed well when trained and tested on individual datasets, cross-dataset generalization highlighted remaining obstacles. To foster collaboration, access to partial annotated corpora and models trained by merging all annotated datasets will be made available on the PhysioNet repository.

Cutting-edge care: unleashing artificial intelligence's potential in gynecologic surgery.

PURPOSE OF REVIEW: Artificial intelligence (AI) is now integrated in our daily life. It has also been incorporated in medicine with algorithms to diagnose, recommend treatment options, and estimate prognosis. RECENT FINDINGS: AI in surgery differs from virtual AI used for clinical application. Physical AI in the form of computer vision and augmented reality is used to improve surgeon's skills, performance, and patient outcomes. SUMMARY: Several applications of AI and augmented reality are utilized in gynecologic surgery. AI's potential use can be found in all phases of surgery: preoperatively, intra-operatively, and postoperatively. Its current benefits are for improving accuracy, surgeon's precision, and reducing complications.

Group 2i Isochrysidales thrive in marine and lacustrine systems with ice cover.

Global warming is causing rapid changes to the cryosphere. Predicting the future trajectory of the cryosphere requires quantitative reconstruction of its past variations. A recently identified sea-ice-associated haptophyte, known as Group 2i Isochrysidales, has given rise to a new sea-ice proxy with its characteristic alkenone distributions. However, apart from the occurrence of Group 2i Isochrysidales in regions with sea ice, and the empirical relationship between C37:4 alkenone abundance and sea-ice concentration, little is known about the ecology of these haptophyte species. Here, we systematically mapped the spatial and temporal occurrence of known Group 2i Isochrysidales based on environmental DNA in both marine and lacustrine environments. Our results indicate Group 2i is widely distributed in icy marine and lacustrine environments in both Northern and Southern Hemisphere, but is absent in warm environments. Temporally, Group 2i is part of the sea-ice algae bloom during the cold seasons, in contrast to other Isochrysidales that bloom in open waters during warm seasons. Our results indicate that ice is a prerequisite for the occurrence of the psychrophilic Group 2i haptophytes in marine and lacustrine ecosystems and further affirms its value for past ice reconstructions.

Feasibility and Perioperative Outcomes of Minimally Invasive Higher Order Myomectomy.

STUDY OBJECTIVE: To investigate perioperative outcomes of minimally invasive higher order myomectomy as defined by removal of 10 or more fibroids. DESIGN: A retrospective cohort study between January 2018 and December 2022. SETTING: A tertiary academic medical center. PATIENTS: Women who underwent minimally invasive myomectomy via laparoscopic or robotic approach. INTERVENTIONS: Surgical intervention in the form of minimally invasive myomectomy. MEASUREMENTS AND MAIN RESULTS: A total of 735 women met inclusion criteria of whom 578 had fewer than 10 fibroids removed, and 157 patients had 10 or more removed (average number of fibroids removed 3.8 vs 14.7, p <.001; specimen's weight 317.4 g vs 371.0 g, p = .07). Body mass index was similar in both groups (p = .66) and patients with higher order myomectomy were more likely to have a history of myomectomy (12.0% vs 26.8%, p <.001). The average estimated blood loss (EBL) was 246 mL vs 470 mL in each group (p <.001). There were no significant differences in packed red blood cell transfusion (1.0% vs 0.6%, p = .65), conversion to laparotomy (0.5% vs 0.6%, p = .86), or complications including visceral injury, wound complication, venous thromboembolism, ileus, or readmission (5.9% vs 4.5%, p = .49). The hospital length of stay was similar in both groups (0.5 days vs 0.5 days, p = .63). On linear regression analysis, after adjusting for specimen's weight, operative time, and history of myomectomy, EBL remained significantly higher in patients with 10 or more fibroids removed (p = .02). CONCLUSION: EBL is increased in higher order myomectomy; however, blood transfusions, conversion to laparotomy, complication rates, and length of hospital stay did not differ compared with patients with fewer than 10 fibroids removed, highlighting the feasibility of minimally invasive higher order myomectomy.

Identifying incarceration status in the electronic health record using large language models in emergency department settings.

Background: Incarceration is a significant social determinant of health, contributing to high morbidity, mortality, and racialized health inequities. However, incarceration status is largely invisible to health services research due to inadequate clinical electronic health record (EHR) capture. This study aims to develop, train, and validate natural language processing (NLP) techniques to more effectively identify incarceration status in the EHR. Methods: The study population consisted of adult patients (≥ 18 y.o.) who presented to the emergency department between June 2013 and August 2021. The EHR database was filtered for notes for specific incarceration-related terms, and then a random selection of 1,000 notes was annotated for incarceration and further stratified into specific statuses of prior history, recent, and current incarceration. For NLP model development, 80% of the notes were used to train the Longformer-based and RoBERTa algorithms. The remaining 20% of the notes underwent analysis with GPT-4. Results: There were 849 unique patients across 989 visits in the 1000 annotated notes. Manual annotation revealed that 559 of 1000 notes (55.9%) contained evidence of incarceration history. ICD-10 code (sensitivity: 4.8%, specificity: 99.1%, F1-score: 0.09) demonstrated inferior performance to RoBERTa NLP (sensitivity: 78.6%, specificity: 73.3%, F1-score: 0.79), Longformer NLP (sensitivity: 94.6%, specificity: 87.5%, F1-score: 0.93), and GPT-4 (sensitivity: 100%, specificity: 61.1%, F1-score: 0.86). Conclusions: Our advanced NLP models demonstrate a high degree of accuracy in identifying incarceration status from clinical notes. Further research is needed to explore their scaled implementation in population health initiatives and assess their potential to mitigate health disparities through tailored system interventions.

Structure and function of Semaphorin-5A glycosaminoglycan interactions.

Integration of extracellular signals by neurons is pivotal for brain development, plasticity, and repair. Axon guidance relies on receptor-ligand interactions crosstalking with extracellular matrix components. Semaphorin-5A (Sema5A) is a bifunctional guidance cue exerting attractive and inhibitory effects on neuronal growth through the interaction with heparan sulfate (HS) and chondroitin sulfate (CS) glycosaminoglycans (GAGs), respectively. Sema5A harbors seven thrombospondin type-1 repeats (TSR1-7) important for GAG binding, however the underlying molecular basis and functions in vivo remain enigmatic. Here we dissect the structural basis for Sema5A:GAG specificity and demonstrate the functional significance of this interaction in vivo. Using x-ray crystallography, we reveal a dimeric fold variation for TSR4 that accommodates GAG interactions. TSR4 co-crystal structures identify binding residues validated by site-directed mutagenesis. In vitro and cell-based assays uncover specific GAG epitopes necessary for TSR association. We demonstrate that HS-GAG binding is preferred over CS-GAG and mediates Sema5A oligomerization. In vivo, Sema5A:GAG interactions are necessary for Sema5A function and regulate Plexin-A2 dependent dentate progenitor cell migration. Our study rationalizes Sema5A associated developmental and neurological disorders and provides mechanistic insights into how multifaceted guidance functions of a single transmembrane cue are regulated by proteoglycans.

Equity and use of telehealth modalities among people living with HIV during the COVID-19 pandemic.

Background: COVID-19 forced a rapid transition to telehealth. Little is known about the use of telephone versus video visits among people living with or at risk for HIV (PWH). Setting: We studied electronic health record data from an urban HIV clinic. Our sample included visit- and person-level data. Visit-level data came from appointments scheduled from 30 March 2020 to 31 May 2020. Person-level data came from patients 18+ years of age who completed at least one telephone or video visit during the period of interest. Methods: We performed a cross-sectional analysis. Our primary outcome was telehealth modality (telephone or video). We compared visit completion status by telehealth modality. We evaluated associations between patient characteristics and telehealth modality using logistic regression. Results: In total, 1742 visits included information on telehealth modality: 1432 telephone (82%) and 310 (18%) video visits. 77% of telephone visits were completed compared to 75% of video visits (p = 0.449). The clinic recorded 643 completed telehealth visits in April and 623 in May 2020. The proportion of telephone visits decreased from 84% in April to 79% in May (p = 0.031). Most patients participated in telephone versus video visits (415 vs. 88 patients). Older age (adjusted odds ratio [AOR] 3.28; 95% confidence interval [CI], 1.37-7.82) and Black race (AOR 2.42; 95% CI, 1.20-4.49) were positively associated with telephone visits. Patient portal enrollment (AOR 0.06; 95% CI, 0.02-0.16) was negatively associated with telephone visits. Conclusion: PWH used telephone more than video visits, suggesting that telephone visits are a vital healthcare resource for this population.

Multiple switching between the biosimilar adalimumab PF-06410293 and reference adalimumab in patients with active rheumatoid arthritis: a phase 3, open-label, randomised, parallel-group study.

BACKGROUND: An adalimumab biosimilar with an interchangeability designation could increase access to effective treatment for more patients. We aimed to assess the interchangeability of adalimumab biosimilar PF-06410293 (adalimumab-afzb) and reference adalimumab using a multi-switch study design. METHODS: We did an open-label, randomised, parallel-group study at 61 community (n=29), hospital (n=12), and academic (n=20) sites in ten countries (Bulgaria, Bosnia and Herzegovina, Czech Republic, Lithuania, Poland, Russia, Serbia, South Africa, Ukraine, and USA). Eligible patients were aged 18-70 years and met the 2010 American College of Rheumatology-European League Against Rheumatism classification criteria for rheumatoid arthritis for at least 4 months with moderately to severely active rheumatoid arthritis, based on their physician's evaluation. Eligible patients had been receiving methotrexate for at least 12 weeks and been on a stable dose for at least 4 weeks before the first dose of study medication. All patients received subcutaneous reference adalimumab (40 mg/0·4 mL [100 mg/mL] every 2 weeks) for 10 weeks before randomisation. At week 10, patients were randomly assigned (1:1) to either three switches between subcutaneous reference adalimumab (40 mg/0·4 mL [100 mg/mL] every 2 weeks) and adalimumab-afzb (40 mg/0·8 mL [50 mg/mL] every 2 weeks; switching group), or continuous dosing with subcutanous reference adalimumab (40 mg/0·4 mL [100 mg/mL] every 2 weeks; non-switching group) with stratification by bodyweight groups. Patients, investigators, and site personnel were not masked to treatment allocation. Primary endpoints were maximum observed serum concentration (Cmax) and area under plasma concentration-time curve (AUCτ) during weeks 30-32 in the pharmacokinetic population. Interchangeability was based on geometric mean ratios and corresponding 90% CIs within prespecified equivalence margins of 80-125% for both primary endpoints. Safety was analysed in all patients who received at least one dose of adalimumab-afzb or reference adalimumab. This trial is registered with ClinicalTrials.gov, NCT04230213. FINDINGS: Of the 569 patients assessed for eligibility between Jan 13, 2020, and June 22, 2021, 445 were enrolled, and 427 completed the first 10 weeks and were randomly assigned (213 to the switching group and 214 to the non-switching group). Participants had a median age of 56 years (IQR 46-63), 354 (83%) of 427 patients were women and 73 (17%) were men, and 422 (99%) were White. In the pharmacokinetic population (n=380), no clinically meaningful differences were observed in mean steady-state pharmacokinetic parameters between the switching and non-switching groups (geometric mean AUC 2237 µg × h/mL in the switching group and 2125 µg × h/mL in the non-switching group; Cmax 8·21 µg/mL in the switching group and 8·00 µg/mL in the non-switching group). Geometric mean ratios and 90% CIs for AUCτ (105·31, 89·16-124·39) and Cmax (102·56, 89·78-117·17) were within prespecified equivalence margins. No meaningful differences were observed in the proportion of patients who had serious adverse events (three [1%] of 213 patients in the switching group vs eight [4%] of 214 patients in the non-switching group), grade 3 or higher adverse events of special interest, discontinuations due to adverse events (eight [4%] vs nine [4%]), or immunogenic reactions in antidrug antibody-positive patients. No deaths were reported during the study. INTERPRETATION: The risk of multiple switches between reference adalimumab and adalimumab-afzb with respect to diminished efficacy (using pharmacokinetics as a surrogate) or safety is not greater than the risk of using reference adalimumab alone. FUNDING: Pfizer. VIDEO ABSTRACT.