Nevin Manimala

Clin Neurol Neurosurg. 2026 Jun 10;269:109534. doi: 10.1016/j.clineuro.2026.109534. Online ahead of print.

ABSTRACT

BACKGROUND: Post-stroke vascular dementia (VaD) affects 20-30% of ischemic stroke survivors within the first year, yet existing prediction models lack comprehensive integration of novel blood biomarkers and validated risk stratification strategies. This study aimed to develop and temporally validate a risk-stratified prediction model incorporating clinical, neuroimaging, and serum biomarker variables.

METHODS: This retrospective cohort study comprised a development cohort (n = 998, 2020-2022) and temporal validation cohort (n = 249, 2023-2024). Consecutive acute ischemic stroke patients aged ≥ 18 years with available baseline magnetic resonance imaging (MRI) and 12-month follow-up were included. Candidate predictors encompassed 25 variables: demographics, vascular risk factors, stroke severity assessed by the National Institutes of Health Stroke Scale (NIHSS), neuroimaging markers (Fazekas white matter hyperintensity score, brain atrophy index [BAI]), and serum biomarkers including neurofilament light chain (NFL) and glial fibrillary acidic protein (GFAP) measured by single-molecule array (Simoa). The primary outcome was incident VaD diagnosed by NINDS-AIREN criteria at 12 months.

RESULTS: Of 1247 included patients, 354 (28.4%) developed VaD. Least absolute shrinkage and selection operator (LASSO) selected 8 predictors: age, education level, NIHSS score, Fazekas score ≥ 2, BAI, previous stroke, plasma NFL, and plasma GFAP. In the development cohort, the model demonstrated excellent discrimination (C-statistic 0.89, 95% CI 0.86-0.92) and good calibration. Bootstrap validation yielded optimism-corrected C-statistic 0.88. In temporal validation, performance remained robust (C-statistic 0.85, 95% CI 0.81-0.89). Risk stratification revealed distinct cognitive trajectories: high-risk patients (25% of cohort) exhibited 67.3% VaD incidence and steep cognitive decline (mean Montreal Cognitive Assessment [MoCA] change -6.8 points), capturing 59.3% of all VaD cases.

CONCLUSIONS: This biomarker-enhanced prediction model demonstrates excellent discrimination and calibration for post-stroke VaD. Risk stratification effectively identifies high-risk patients for targeted interventions, providing a practical tool for precision-based clinical management.

PMID:42302346 | DOI:10.1016/j.clineuro.2026.109534

Lung Cancer. 2026 Jun 14;218:109499. doi: 10.1016/j.lungcan.2026.109499. Online ahead of print.

ABSTRACT

BACKGROUND: Immune checkpoint inhibitors (ICIs) have transformed cancer therapy but can cause rare, severe cardiotoxicity. The real-world incidence and impact of a broader spectrum of cardiovascular events remains poorly defined. This study aims to evaluate the real-world incidence of cardiovascular events following ICI therapy, explore predictive biomarkers, and assess the impact of treatment interruptions.

METHODS: A retrospective observational study using Optum’s de-identified Market Clarity Data was conducted in 29,503 patients receiving ICI therapy for any cancer type with a minimum follow-up of 6 months. Cardiovascular events including myocarditis, arrhythmias, and reduced left ventricular ejection fraction (LVEF < 50%), were analyzed. Kaplan-Meier survival curves were used to evaluate the timing of these events. Biomarkers such as NT-proBNP and troponin were evaluated for their predictive value.

RESULTS: Out of 29,503 ICI-treated patients, 27.6% experienced a cardiac event during the follow-up period (2 years). Patients with pre-existing cardiovascular conditions who were receiving cardioprotective treatment prior to ICI therapy had an increased risk of cardiovascular events (35% vs 20%, p < 0.001). Patients who experienced a cardiac event had a significantly higher mortality rate (39% vs 25.4%, p < 0.001). Elevated troponin and NT-proBNP levels were associated with increased mortality (p < 0.001).

CONCLUSIONS: Cardiovascular events are frequent in ICI-treated patients, particularly in those with pre-existing cardiac conditions. Elevated troponin and NT-proBNP levels may serve as useful biomarkers for predicting post-ICI cardiovascular events. These findings demonstrate that these events significantly interrupt ICI therapy and increase mortality. They support biomarker-guided risk stratification and prompt collaboration between oncologists and cardio-oncologists to preserve treatment integrity.

PMID:42302339 | DOI:10.1016/j.lungcan.2026.109499

Epilepsy Behav. 2026 Jun 16;183:111164. doi: 10.1016/j.yebeh.2026.111164. Online ahead of print.

ABSTRACT

BACKGROUND AND OBJECTIVES: Functional/dissociative seizures (FDS) are frequently treated with one-to-one therapeutic interventions. However, data describing patient-reported outcome measures (PROMs) following group-based therapy remain limited. This study evaluated changes in PROMs among a large cohort of individuals receiving virtual, group and individual therapy within a multidisciplinary FDS treatment program.

METHODS: We conducted a retrospective cohort study of 823 individuals referred to a specialty FDS program between April 2020 and January 2025. Baseline PROMs were completed by 436 patients and assessed quality of life (QoL), PTSD, anxiety, depression, and dissociative experiences. Follow up PROMs were available for 165 individuals who participated in one of three treatment pathways: 6-Week psychoeducational group therapy only, 6-Week and 12-Week psychodynamic group therapies combined, or individualized therapy. Mixed-effects linear regression examined changes in PROMs between baseline and follow-up, adjusting for relevant characteristics.

RESULTS: Across all cohorts, significant improvements were observed in depression (PHQ-9: -1.32, p < 0.01), quality of life (QOLIE-10P: -2.20, p < 0.01), and PTSD symptoms (SPRINT: -2.14, p < 0.01). Within-cohort analyses showed significant reductions in PHQ-9 (-2.16, p < 0.01) and SPRINT (-3.14, p < 0.01) for the 6-Week and 12-Week group therapies combined and improved QOLIE-10P (-3.37, p < 0.01) for the 6-Week group only. No significant changes were observed in anxiety or dissociation, and between-cohort differences were not statistically significant.

DISCUSSION: Analyses identified significant improvements in PROMs across treatment cohorts suggesting that group-based and individual interventions can enhance outcomes for individuals with FDS, though absence of a control group limits causal interpretation. This finding is important given limited treatment availability and opportunity for group therapy to increase FDS access.

PMID:42302323 | DOI:10.1016/j.yebeh.2026.111164

J Cataract Refract Surg. 2026 Jun 16. doi: 10.1097/j.jcrs.0000000000002005. Online ahead of print.

ABSTRACT

PURPOSE: To compare tolerance to induced astigmatism (TIA) across four intraocular lenses (IOLs).

SETTING: Medical University of South Carolina, Charleston, South Carolina, USA.

DESIGN: Prospective observational study.

METHODS: Patients underwent cataract surgery with a standard monofocal (ZCB00), enhanced monofocal (DIB00), diffractive violet light filter (VLF) extended depth of field (DOFi) (ZXR00V), or VLF full DOFi (DRN00V) IOL. At 1-3 months postoperatively, astigmatic defocus (0.5-2.0 D) was simulated in WTR, oblique, and ATR orientations.

RESULTS: 100 patients were included. DCIVA was statistically significantly better with VLF extended DOFi compared with the monofocal (P≤0.001) and enhanced monofocal IOLs (P=0.002). DCNVA was significantly better with VLF Full-DOFi IOL than with all other groups (P<0.001). No statistically significant differences were observed among the four IOL groups with induction of WTR astigmatism ≤1.0 D. At 0.5D oblique, the enhanced monofocal group outperformed the extended DOFi group (P=0.014), and at 2.0 D induced oblique astigmatism, it surpassed all other IOL groups. The enhanced monofocal IOL group showed the greatest tolerance to induced ATR astigmatism across all magnitudes. The VLF Full-DOFi IOL group maintained astigmatic tolerance like standard monofocal, enhanced monofocal, and extended DOFi IOLs up to 2.0 D in WTR and ATR, and 1.5 D in oblique orientations.

CONCLUSION: TIA was not statistically significantly different among the standard monofocal, enhanced monofocal, VLF extended DOFi, and VLF Full-DOFi IOL groups with low to moderate induced WTR astigmatism. The enhanced monofocal IOL provided the greatest tolerance to induced astigmatism, especially under oblique and ATR astigmatic defocus.

PMID:42302314 | DOI:10.1097/j.jcrs.0000000000002005

JMIR Nurs. 2026 Jun 16;9:e76682. doi: 10.2196/76682.

ABSTRACT

BACKGROUND: Pressure injuries (PIs) are a common complication in people with reduced mobility or sensation and can be burdensome for individuals with PIs and their caregivers. Valuable insights and real-world challenges faced by individuals living with PIs can be captured through candid accounts posted on social media. Social media listening (SML) is a tool that can enhance the understanding of those with lived experience by offering firsthand accounts that are irreproducible from controlled studies.

OBJECTIVE: This study aims to capture the candid experiences of individuals with PIs and caregivers through social media.

METHODS: A noninterventional qualitative descriptive analysis was conducted using SML. Social media posts made on X (formerly Twitter), Reddit, and YouTube between January and December 2022 were compiled using SML tools X Pro (formerly TweetDeck) and Awario, and using Boolean search terms. Posts were manually screened for relevance, and duplicates were removed. Relevant posts were hand-coded by two independent reviewers. Inductive content analysis was used to analyze the posts.

RESULTS: The search yielded 666 relevant posts from 498 unique social media users. Most posts were made in the United States (170/666, 25.5%), the United Kingdom (150/666, 22.5%), and Canada (62/666, 9.3%). Social media users provided detailed descriptions of the PIs, including the setting in which the PI occurred, the cause of the PI, and how the PI was managed. The majority of PIs (197/666, 29.6%) were reported to have occurred in the hospital setting due to a perceived lack of care from care providers, and local wound care was often cited (99/666, 14.9%) as a PI management strategy. Three key themes were developed regarding living with or caring for someone with a PI: (1) challenges experienced when living with or caring for a PI, (2) needs related to PI prevention and management, and (3) emotions experienced when living with or caring for a PI. Social media users frequently discussed challenges associated with living with a PI, including negative personal impacts and poor perceived treatment quality. Users also described a critical need for health care, education, and social support. Finally, users often expressed anger and/or sadness related to living with or caring for a PI.

CONCLUSIONS: SML captured candid insights into the experiences, challenges, and needs of individuals living with PIs and their caregivers globally that may not be gleaned from controlled studies. Individuals with lived experience and their caregivers often struggled with negative personal impacts regarding their physical health and daily functioning related to PIs, further highlighting the urgent need to address barriers to appropriate PI care. Clinicians and policymakers should consider practices and policies that optimize the delivery of person-centered PI care in order to overcome challenges and needs identified in this study.

PMID:42302309 | DOI:10.2196/76682

JMIR Med Educ. 2026 Jun 16;12:e81250. doi: 10.2196/81250.

ABSTRACT

BACKGROUND: In 2021, the American Academy of Pediatrics released a policy statement spotlighting the health-promoting and stress-buffering effects of early relational health (ERH) and calling for universal ERH promotion in pediatric primary care. However, little educational content for the observation and promotion of ERH is available, highlighting the need for scalable ERH training modules.

OBJECTIVE: This study aims to investigate the acceptability, feasibility, and impact of the “Lens of Emotional Connection,” a self-paced, asynchronous, ~45-minute ERH training module codeveloped by Reach Out and Read and the Center for Early Relational Health at Columbia University. The module introduces practitioners to emotional connection, an observable component of ERH, through written and video didactic content and experiential rating of emotional connection in videos of parent-child dyads interacting face-to-face.

METHODS: The evaluation was conducted by the Carolinas Collaborative. Pediatric providers across 8 clinical sites were invited to participate and responded to embedded pre-post surveys. Focus groups conducted with participants further examined the educational experience.

RESULTS: Of 653 invited clinicians, 207 (31.7%) participated in the module. Individual survey responses were available for 44-75 participants, depending on the question. Of responders, 64 out of 69 (93%) reported the module was a good way to learn about emotional connection, and 63 out of 69 (91%) felt the module provided valuable knowledge. Overall, 60 out of 69 respondents (87%) reported satisfaction with the module length, and 36 out of 44 respondents (82%) reported they would recommend this training to other clinicians. Focus groups echoed these findings. Comparison of pre-post data showed the greatest changes were in familiarity with emotional connection (n=75, pre mean 54.20, SD 18.59; post mean 73.99, SD 14.73; Cohen d=1.14; P<.001) and confidence in observing the quality of the parent- or caregiver-infant relationship during well-child visits (n=75, pre mean 55.36, SD 18.49; post mean 74.20, SD 14.07; Cohen d=1.28; P<.001). Suggested areas for improvement included more thorough explanations of specific components of emotional connection identified in parent-child interaction videos, a desire for synchronous live training, and additional content addressing what to do if low emotional connection is identified.

CONCLUSIONS: In this evaluation of a training module designed to introduce pediatric practitioners to ERH and emotional connection, acceptability among participants was found to be high, with most responders reporting it as valuable and reporting they would recommend it. Statistically significant impact was noted in both perceptions of the importance of information about emotional connection and perceived knowledge acquisition. Feasibility of widespread implementation with voluntary participation, as here, was relatively low, with only a minority completing the module. Critically, Reach Out and Read’s commitment to iterative creation, validation, and eventual delivery of ERH training creates a scalable avenue for wide-scale implementation, given the organization’s presence in >6500 clinics across the 50 states.

PMID:42302308 | DOI:10.2196/81250

J Med Internet Res. 2026 Jun 16;28:e92852. doi: 10.2196/92852.

ABSTRACT

BACKGROUND: Public health medical education is increasingly important in the low-resource, high-altitude Xizang Autonomous Region (Tibet). Traditional authoritative textbooks do not meet modern needs for accessibility and interactivity, whereas general large language models (LLMs) may hallucinate in specialized medical domains. Developing specialized LLMs for low-resource regions is also expensive and difficult.

OBJECTIVE: This study aimed to explore a novel approach to high-altitude public health medical education in the low-resource Xizang Autonomous Region that integrates modern LLMs and authoritative textbooks, using a comprehensive benchmark evaluation across multiple dimensions and retrieval-augmented generation (RAG) technology.

METHODS: We conducted a 2-stage cross-sectional comparative evaluation study to benchmark publicly available LLMs and evaluate the added value of textbook-augmented retrieval under standardized generation settings and blinded expert assessment. First, 4 publicly available LLMs (GPT-5.2 [OpenAI], Gemini 3.0 Pro [Google], DeepSeek R1 [DeepSeek], and Tencent HY 2.0 [Tencent]) were benchmarked using an 80-question benchmark on high-altitude public health medicine developed by authoritative medical specialists. Each question was asked 3 times, yielding 960 outputs; first responses (n=320) were scored under blinded conditions by 2 independent 8-member physician panels. A clinically weighted evaluation of multidimensional first-response scores (including comprehensiveness, accuracy, clarity, and relevance) and a composite consistency metric (including semantic similarity and algorithmic similarity) was administered. Second, 4 specific and prevalent authoritative textbooks on high-altitude public health medicine-Ward, Milledge and West’s High Altitude Medicine and Physiology, High Altitude Medicine: A Case-Based Approach, High Altitude Medicine, and High Altitude Medical Protection-were deployed as the external knowledge base for the evaluation-optimized model. Statistical analyses included Spearman ρ, Cronbach α, intraclass correlation coefficients, Friedman tests with Dunn multiple comparisons, and paired Wilcoxon signed-rank tests. The significance threshold was set at α=.05.

RESULTS: DeepSeek R1 was selected as the optimal base model for achieving the highest weighted score (5.61/10.00), followed by GPT-5.2 (5.51/10.00), Gemini 3.0 Pro (5.39/10.00), and Tencent HY 2.0 (4.71/10.00). The deployed retrieval-augmented model integrating the authoritative textbooks and the optimal LLM DeepSeek R1, HPHME-Xplus-RAG, achieved remarkable improvement in multidimensional scores compared to baseline DeepSeek R1 (median 8.00 [IQR 7.88-8.00] vs median 7.63 [IQR 7.38-7.88]; P<.001, r_rb=0.68, indicating a large effect).

CONCLUSIONS: Integrating authoritative textbooks with an evaluation-optimized general LLM through an RAG framework showed strong performance for medical education in the low-resource Xizang Autonomous Region. Unlike prior studies that mainly evaluated general LLMs or used clinical guidelines to build RAG systems for diagnosis and treatment, this study used authoritative textbooks for the broader, guideline-scarce field of public health medical education. This work provides a replicable workflow-domain-authoritative knowledge+RAG+model optimization and evaluation-for low-resource settings, with practical implications for medical instructors and students, hospitals, and public health services seeking cost-effective, convenient, and trustworthy educational support.

PMID:42302307 | DOI:10.2196/92852

JMIR Aging. 2026 Jun 16;9:e90705. doi: 10.2196/90705.

ABSTRACT

BACKGROUND: Postoperative frailty is highly prevalent among older adults undergoing hip surgery and is closely linked to poor clinical outcomes. Despite growing interest in understanding its progression, the temporal patterns of frailty remain underexplored. Moreover, there is a lack of validated models that can predict frailty trajectories and stratify patients by risk in the early postoperative period.

OBJECTIVE: This study aimed to identify distinct frailty trajectories within 6 months following hip surgery in older adults and to explore their associated predictors. An interpretable machine-learning model was developed and internally validated for individualized risk prediction and was implemented as a clinically accessible web-based calculator.

METHODS: This prospective longitudinal observational study was conducted among older adults undergoing hip surgery at a tertiary hospital in China. Frailty assessments were performed preoperatively and at 1, 3, and 6 months postoperatively. A total of 209 participants who completed the 6-month follow-up were included in the analysis. Frailty was assessed using the Frailty Index, and group-based trajectory modeling was applied to identify distinct frailty progression patterns. Predictive variables were selected using the least absolute shrinkage and selection operator regression. An interpretable Extreme Gradient Boosting (XGBoost) model was developed using a 60:40 training-test data split. Model performance was evaluated in terms of discrimination, calibration, and clinical utility. Interpretability was assessed using SHAP (Shapley Additive Explanations) at both the global and individual levels.

RESULTS: Three distinct frailty trajectories were identified: low-fluctuation frailty (55/209, 26%), high-improvement frailty (81/209, 39%), and high-deterioration frailty (73/209, 35%). Twelve predictors grounded in the Health Ecology Model were selected, spanning individual characteristics, interpersonal networks, and the living environment. The XGBoost model demonstrated excellent discrimination, with a microaverage area under the receiver operating characteristic curve of 0.98 (95% CI 0.96-0.99) in the training set and 0.93 (95% CI 0.90-0.96) in the test set. Calibration was acceptable, with a weighted Brier score of 0.0852. Decision curve analysis showed favorable clinical utility across a range of threshold probabilities. A web-based risk calculator was developed to facilitate personalized frailty trajectory prediction.

CONCLUSIONS: The XGBoost model demonstrated strong predictive performance and interpretability, enabling the early identification of older patients at risk for adverse frailty trajectories following hip surgery. This tool may support targeted interventions and improve perioperative care in geriatric populations.

PMID:42302293 | DOI:10.2196/90705

JMIR Med Educ. 2026 Jun 16;12:e90064. doi: 10.2196/90064.

ABSTRACT

BACKGROUND: Feedback is essential for medical students’ learning during clinical clerkships; yet, supervising physicians often struggle to provide meaningful written feedback due to time constraints. Large language models offer a promising approach to supplement human feedback, but how artificial intelligence (AI)-generated and human feedback differ in authentic clinical settings remains unclear, as most comparisons have been conducted in classroom or simulation contexts.

OBJECTIVE: The aim of the study is to examine how AI-generated feedback and supervisor-provided feedback differ when applied to medical students’ clinical clerkship logs, by identifying the distinct characteristics and complementary strengths of each feedback type.

METHODS: This cross-sectional convergent mixed methods study included 161 weekly clinical clerkship logs from 47 fifth- and sixth-year medical students across 12 clinical departments at Nagoya University, Japan (January-May 2024). Of 164 eligible logs, 3 were excluded because supervisors entered contact messages rather than substantive feedback. AI feedback was generated using GPT-4o. In total, 10 faculty physicians and 10 medical students evaluated both feedback types in blinded, randomized order using a validated 5-category rubric (criteria-based, clear direction, accuracy, prioritization, and supportive tone), followed by open-ended comments and source identification. Quantitative analyses (paired 2-tailed t tests, cumulative link mixed-effects models; α=.05 with Bonferroni correction) were complemented by qualitative thematic analysis and integrated using joint display analysis.

RESULTS: AI feedback was significantly longer than supervisor feedback (mean 382.02, SD 81.82 vs mean 98.87, SD 73.66 characters; Cohen d=2.84, 95% CI 2.50-3.19; P<.001). Cumulative link mixed-effects models showed that AI scored higher on criteria-based (odds ratio [OR] 11.81, 95% CI 7.64-18.27; P<.001) and clear direction (OR 6.61, 95% CI 4.35-10.06; P<.001), with no significant differences on accuracy (OR 1.35, 95% CI 0.91-2.00; P>.99), prioritization (OR 1.70, 95% CI 1.16-2.50; P=.10), or supportive tone (OR 1.34, 95% CI 0.87-2.06; P>.99). AI feedback showed greater consistency (variance ratio 3.9:1; Levene F1,320=73.20; P<.001). All 20 evaluators correctly identified feedback sources. Qualitative analysis revealed that AI provided structured, text-anchored feedback addressing rubric criteria, while supervisors offered experience-based feedback grounded in clinical context and professional expertise.

CONCLUSIONS: This study extends the comparison of AI-generated and supervisor feedback to an authentic clinical clerkship environment, moving beyond classroom and simulation settings examined in prior work. Through integrated mixed methods analysis, a key distinction emerged between text-anchored AI feedback, which systematically addresses written log content in alignment with rubric criteria, and experience-based supervisor feedback, which draws on clinical observation and professional judgment. AI consistently delivered structured feedback addressing gaps that arise when time-pressured supervisors provide brief comments, while supervisors contributed clinically grounded insights that AI cannot replicate. These complementary strengths suggest that AI feedback should supplement rather than replace supervisor feedback, and that hybrid models leveraging each type’s advantages warrant investigation in clinical education.

PMID:42302283 | DOI:10.2196/90064

JMIR Cancer. 2026 Jun 16;12:e89823. doi: 10.2196/89823.

ABSTRACT

BACKGROUND: Lung cancer remains the leading cause of cancer deaths in the United States; however, uptake of lung cancer screening (LCS) with low-dose computed tomography (LDCT) among eligible individuals remains low. Evidence suggests that limited knowledge, stigma, and false health beliefs contribute to the underuse of LDCT screening.

OBJECTIVE: This pilot study aimed to examine an online educational intervention designed to improve knowledge, attitudes, health beliefs, behavioral intentions, perceived importance, and confidence related to LCS among high-risk individuals.

METHODS: A single-group preintervention and postintervention design was used. High-risk individuals who smoke, defined according to the US Preventive Services Task Force criteria, completed baseline questionnaires followed by 5 self-directed online educational modules delivered through Research Electronic Data Capture (REDCap). Postintervention questionnaires assessed changes in lung cancer and screening knowledge, lung cancer stigma, health beliefs based on the health belief model and precaution adoption process model, and intentions, perceived importance, and confidence regarding LDCT screening. LCS uptake was assessed via follow-up email 3 months after the intervention. Data were analyzed using descriptive statistics and paired-samples two-tailed t tests.

RESULTS: A total of 25 participants completed the intervention. Significant improvements were observed across all major study outcomes. Knowledge scores increased markedly (score=3.76-8.60; P<.001), while lung cancer stigma decreased (score=25.52-19.16; P<.001). Health belief model constructs showed significant improvements, including perceived susceptibility, perceived benefits, cues to action, and self-efficacy, alongside reductions in perceived barriers and perceived severity (all P<.001). Self-reported intentions, perceived importance, and confidence related to obtaining LDCT screening increased significantly. Of the 22 (88%) participants who completed the 3-month follow-up, 13 (59.1%) reported obtaining LDCT screening. Participant satisfaction with the intervention was high, with a mean score of 18.32 (SD 2.33) out of 20.

CONCLUSIONS: Findings from this pilot study support the feasibility, acceptability, and preliminary efficacy of an online educational intervention created to promote LCS among high-risk individuals. The intervention improved knowledge; reduced stigma; positively influenced health beliefs; and increased screening intentions, perceived importance, confidence, and uptake. Results provide a foundation for a larger-scale study and suggest that online educational platforms may be an effective strategy to reach geographically diverse high-risk populations and promote LDCT screening.

PMID:42302275 | DOI:10.2196/89823