51886 research outputs found
Sort by
Computerized tomography‐derived body composition metrics are associated with 24‐h urine lithogenic parameters
No full textBackground: The relationship between body composition and lithogenic urine parameters remains poorly defined. This study aimed to evaluate associations between computerized tomography (CT)-derived body composition metrics and 24-h urine findings.
Methods: Stone-forming patients in our Nephrolithiasis Database who underwent 24-h urine testing and CT within 120 days were retrospectively reviewed. Skeletal muscle index (SMI), visceral adipose tissue index (VATI), subcutaneous adipose tissue index (SATI) and skeletal muscle density (SMD) were calculated from segmented L3 axial images. Spearman correlations and multivariable logistic regression tested associations between body composition and 24-h urine markers.
Results: Among 443 patients, all body composition metrics demonstrated numerous correlations with 24-h urine marker values on Spearman analysis. After adjusting for confounders, higher SMI quartiles were associated with increased odds of elevated urine volume (OR 2.13-2.71), hyperoxaluria (OR 2.11), hyperuricosuria (OR 2.60) and hypernatriuria (OR 2.70). Higher VATI was associated with reduced odds of elevated urine volume (OR 0.44), SATI with elevated sodium excretion (OR 2.35-2.38) and higher SMD with decreased odds of elevated oxalate (OR 0.50) and hypocitraturia (OR 0.41).
Conclusions: CT-derived body composition metrics show distinct and clinically meaningful relationships with 24-h urine profiles. Muscle mass, adipose distribution and muscle quality each influence lithogenic risk, supporting incorporation of body composition assessment into metabolic evaluation of stone-forming patients
Preoperative GLP-1 Receptor Agonist Use and Postoperative Opioid Outcomes in Elective Surgery
No full textPreoperative GLP-1 Receptor Agonist Use and Postoperative Opioid Outcomes in Elective Surgery
Ellias Hanna, BS1, Caitlin R. Priest, MD2,3, Vidhya Gunaseelan, MBA, MS, MHA3,4, Ryan Howard, MD2,3, Joshua Adkinson, MD5, Jennifer F. Waljee, MD, MPH, MS2,3,4,5
1Indiana University School of Medicine, South Bend, IN
2Department of Surgery, University of Michigan, Ann Arbor, MI
3Center for Healthcare Outcomes and Policy, Ann Arbor, MI
4Michigan Opioid Prescribing Engagement Network, Ann Arbor, MI
5Department of Surgery, Indiana University School of Medicine, Indianapolis, IN
Objective
To examine the association between preoperative glucagon-like peptide-1 receptor agonist (GLP-1 RA) use and postoperative opioid outcomes among opioid-naïve adults undergoing elective surgery.
Background
GLP-1 receptor agonists modulate central reward pathways and have been associated with reduced substance use and opioid-related outcomes in preclinical and observational studies. However, whether preoperative GLP-1 RA use is associated with reduced postoperative opioid prescribing or persistent opioid use in surgical populations remains unclear.
Methods
We conducted a retrospective cohort study using the Merative MarketScan Commercial Database, including opioid-naïve adults aged 18–64 years undergoing one of the 20 most common elective surgical procedures between January 1, 2018, and June 30, 2023. The primary exposure was any GLP-1 RA prescription fill within 365 days prior to surgery. Secondary analyses evaluated patterns of preoperative GLP-1 RA exposure using K-means clustering based on prescription duration, continuity, and recency. Outcomes included perioperative opioid prescription fill, opioid refill, new persistent opioid use, and total opioid prescribing volume measured as morphine milligram equivalents (MMEs) within 30 and 90 days after discharge. Multivariable regression models adjusted for demographics, comorbidity burden, mental health and substance use disorders, chronic pain conditions, and surgical procedure type.
Results
Among 914,550 opioid-naïve surgical patients, 20,482 (2.2%) had preoperative GLP-1 RA exposure. In adjusted analyses, GLP-1 RA use was associated with higher odds of perioperative opioid prescription fill (adjusted odds ratio [aOR], 1.18; 95% CI, 1.13–1.22), opioid refill (aOR, 1.05; 95% CI, 1.01–1.10), and new persistent opioid use (aOR, 1.17; 95% CI, 1.05–1.30). Absolute differences in outcome rates were small, and adjusted mean MMEs at 30 and 90 days were nearly identical between GLP-1 RA users and non-users. Cluster-based analyses did not demonstrate a protective association for longer, more continuous, or more recent GLP-1 RA exposure.
Conclusions
Preoperative GLP-1 receptor agonist use was not associated with clinically meaningful reductions in postoperative opioid prescribing or opioid exposure among opioid-naïve surgical patients, suggesting limited utility of GLP-1 RAs as opioid-sparing agents in the perioperative setting
Monte Carlo Dropout for Uncertainty‐Aware Alzheimer's Disease Classification Using Transformer Models on Whole‐Genome Sequencing Data
No full textBackground:
Alzheimer's disease (AD) is a progressive neurodegenerative disorder marked by cognitive decline and memory impairment. Early and accurate detection is critical for clinical intervention. While machine and deep learning approaches have been widely used to predict AD progression, recent work emphasizes quantifying predictive uncertainty in high‐stakes medical contexts. However, many studies focus on limited genetic regions (e.g., APOE), highlighting the need for broader whole‐genome sequencing (WGS) analyses.
Method:
We obtained 1,050 WGS datasets (443 cognitively normal, 607 AD‐diagnosed) from ADNI, ADNI‐WGS‐2, and ADSP‐FUS1‐ADNI‐WGS‐2. SNPs were extracted from a region containing the APOE gene on chromosome 19, then divided into fixed‐size windows (“tokens”) for a Transformer‐based classification model. Monte Carlo (MC) Dropout was applied during training and inference to enable multiple forward passes, providing predictive variance. Models with and without MC Dropout were compared using accuracy (ACC), area under the curve (AUC), F1 score, sensitivity, specificity, expected calibration error (ECE), and reliability diagrams. Predictions were further stratified into “Uncertain” (top 25% variance) and “Certain” (bottom 75%) to examine accuracy differences.
Result:
Stratification by predictive variance revealed that the Uncertain group, with 53 samples, had an accuracy of 0.5472, while the Certain group, with 157 samples, had 0.6497. This indicates that Monte Carlo Dropout–derived variance effectively captures higher uncertainty. The Monte Carlo Dropout model showed a slight increase in accuracy from 0.6143 to 0.6238 and in area under the curve from 0.6644 to 0.6832 compared with baseline, but calibration worsened when the expected calibration error rose from 0.1024 to 0.1858. Sensitivity and specificity shifted from 0.7417 to 0.6833 and from 0.4444 to 0.5444, while the F1 score decreased from 0.6873 to 0.6749.
Conclusion:
These findings demonstrate the feasibility of using MC Dropout within a Transformer‐based WGS framework to estimate predictive uncertainty in AD classification. Variance‐based stratification effectively flagged samples with lower confidence. However, MC Dropout did not consistently improve calibration or overall accuracy, indicating a need for refining dropout settings, Transformer hyperparameters, and potentially Bayesian time‐series methods. Future work should also explore complementary approaches to balance classification accuracy with reliable uncertainty estimates
P-1435. Effectiveness of RSV Immunization among Infants in Their First RSV Season in the United States, 2023-2024 and 2024-2025 Seasons
No full textBackground:
During the 2023-2024 and 2024-2025 respiratory virus seasons, two products, a maternal vaccine (Abrysvo, by Pfizer) and an infant monoclonal antibody (nirsevimab), were recommended in the United States to protect infants against severe respiratory syncytial virus (RSV) disease during their first RSV season. We estimated product effectiveness (PE) against RSV-associated emergency department (ED) encounters and hospitalizations among infants in their first RSV season during these seasons.
Methods:
VISION includes integrated health record data with linkages to state and local immunization information systems from 127 EDs and 107 hospitals. We conducted a test-negative analysis using data from 6 VISION sites, including children with ED encounters and hospitalizations with a diagnosis of RSV-like illness (RLI) during October 2023–March 2024 and October 2024–March 2025 among infants in their first RSV season. PE was estimated for each product, comparing the odds of immunization versus no immunization among RSV-positive cases and RSV-negative controls, adjusting for age, race and ethnicity, sex, calendar day, and geographic region.
Results:
Of the 8,404 RLI ED encounters and 1,395 RLI hospitalizations among infants in their first RSV season without exposure to maternal RSV vaccine, 2,981 (35%) and 771 (55%) were RSV-positive and 1,448 (17%) and 208 (15%) had evidence of nirsevimab receipt, respectively. Nirsevimab effectiveness was 67% (95% CI: 61-72%) against RSV-associated ED encounters and 92% (95% CI: 87-95%) against RSV-associated hospitalization (Table 1). In 1,508 RLI ED encounters and 421 RLI hospitalizations among infants in their first RSV season without receipt of nirsevimab, 480 (32%) and 222 (53%) were RSV-positive and 261 (17%) and 62 (15%) had evidence of maternal RSV vaccination, respectively. Maternal RSV vaccine PE was 70% (95% CI: 57-79%) against RSV-associated ED encounters and 79% (95% CI: 58-89%) against RSV-associated hospitalization among infants in their first RSV season (Table 2).
Conclusion:
Both RSV prevention products were effective in preventing RSV-associated ED encounters and hospitalizations among infants in their first RSV season in the United States. It is important to continue monitoring RSV PE during future RSV seasons
Zoliflodacin versus ceftriaxone plus azithromycin for treatment of uncomplicated urogenital gonorrhoea: an international, randomised, controlled, open-label, phase 3, non-inferiority clinical trial
No full textBackground: Development of new treatments for gonorrhoea is a global public health priority. We aimed to evaluate the efficacy and safety of zoliflodacin versus ceftriaxone plus azithromycin in patients with uncomplicated urogenital gonorrhoea.
Methods: In this phase 3, multinational, randomised, controlled, open-label, non-inferiority clinical trial, participants aged 12 years and older with clinical suspicion of uncomplicated urogenital gonorrhoea were eligible for inclusion. The trial was done in 17 outpatient clinics in Belgium, the Netherlands, South Africa, Thailand, and the USA. Participating countries with high disease prevalence were identified for participation in the study. Sites selected for participation were led by principal investigators with research experience, who were knowledgeable in HIV or sexually transmitted infections and treatment. Feasibility questionnaires and prestudy visits assessed sexually transmitted infection case management guidelines, clinical services, and resources (ie, facility, staff, proposed composition of the study team, standard sexually transmitted infection services offered at the site, assessment of laboratory capacity, research experience and ethical review of clinical trials). Eligible participants were randomly assigned (2:1) to receive a single dose of zoliflodacin 3 g (oral) or ceftriaxone 500 mg (intramuscular) plus azithromycin 1 g (oral). Treatment assignment was known to the participants and their treating clinicians; however, microbiology laboratory staff were masked and the sponsor's central study team were masked until after database lock. The primary endpoint was the proportion of patients with microbiological cure (eradication of Neisseria gonorrhoeae, determined by urethral or endocervical culture) at test of cure (TOC; day 6 ± 2) in the microbiological intention-to-treat population. The primary efficacy analysis declared non-inferiority if the upper bound of the two-sided 95% CI for the treatment difference (comparator minus zoliflodacin) fell below the 12% non-inferiority margin. The trial is registered with ClinicalTrials.gov, NCT03959527, and EudraCT, 2019-000990-22.
Findings: Between Nov 6, 2019, and March 16, 2023, 1011 patients were screened. 81 patients did not meet screening criteria and 930 participants were randomly assigned to zoliflodacin (n=621) or comparator (n=309). The mean participant age was 29·7 years (SD 9·4). 815 (88%) of 930 participants were assigned male at birth and 115 (12%) participants were assigned female at birth. 514 (55%) of 930 participants were Black or African American, 285 (31%) were Asian, and 113 (12%) were White. Microbiological cure rates at TOC in the microbiological intention-to-treat (urogenital) population (primary efficacy endpoint) were 460 (90·9%, 95% CI 88·1-93·3) of 506 participants for zoliflodacin and 229 (96·2%, 92·9-98·3) of 238 participants for comparator. The estimated difference between groups was 5·3% (95% CI 1·4-8·6) and the upper confidence interval limit was within the prespecified non-inferiority margin of less than 12%. Zoliflodacin was generally well tolerated and adverse events were similar between treatment groups. The most frequently reported treatment-emergent adverse events included headache (61 [10%] of 619 patients), neutropenia (42 [7%]), and leukopenia (24 [4%]) in the zoliflodacin group and injection site pain (38 [12%] of 308 patients), neutropenia (24 [8%]), and diarrhoea (22 [7%]) in the comparator group. The majority of adverse events were mild or moderate in severity. No serious adverse events were reported.
Interpretation: Zoliflodacin was non-inferior to ceftriaxone plus azithromycin for the treatment of uncomplicated urogenital gonorrhoea and had a similar safety profile. These data suggest a potential role for zoliflodacin as an effective oral treatment option for uncomplicated urogenital gonorrhoea
A Biomarker‐Based Dose–Schedule Optimization Design for Immunotherapy Trials
No full textIn immunotherapy, both the dose and the schedule of drug administration can significantly influence therapeutic effects by modulating immune system activation. Incorporating immune response measures into clinical trial designs offers an opportunity to enhance decision-making by leveraging their close association with therapeutic efficacy and toxicity. Motivated by settings where biomarker data indicate improved efficacy in biomarker-positive patients, we propose a dose-schedule optimization strategy tailored to each biomarker-defined subgroup, based on elicited utility functions that capture risk-benefit tradeoffs. We introduce a joint modeling framework that simultaneously evaluates immune response, toxicity, and efficacy, enabling information sharing across outcome types and patient subgroups. Our approach utilizes parsimonious yet flexible models designed specifically to address challenges due to small sample sizes commonly encountered in early-phase trials. Simulation studies demonstrate that the proposed design achieves desirable operating characteristics and effectively informs dose-schedule optimization
Longitudinal analysis of plasma p‐Tau181 in African and African American populations from the Indianapolis‐Ibadan Dementia Cohort
No full textBackground:
The Indianapolis‐Ibadan Dementia Project (IIDP) is a longitudinal epidemiological study that evaluated subjects in Indianapolis, Indiana and Ibadan, Nigeria for prevalence and incidence of cognitive decline and dementia between 1991‐2012. Plasma collected from the 2001 wave is currently stored at the National Centralized Repository for Alzheimer's Disease and Related Dementias (NCRAD). We analyzed the IIDP plasma to assess whether the ability of p‐Tau181 to predict disease differed between subjects in these two geographically, culturally, genetically, and environmentally disparate sites.
Method:
Plasma collected from study participants was analyzed for Alzheimer's disease biomarkers using the Quanterix Simoa HD‐X pTau181 v2 Advantage kits. The mean and standard deviation of the log transformed biomarker data using all subjects with normal cognition from 2001 were used to standardize the biomarker prior to statistical analysis. For this analysis, participants with normal cognition during the 2001 wave and at least one follow‐up evaluation were included (N = 755 African Americans; N = 864 Nigerians). Incident cognitive impairment status was determined using the last wave the subject was assessed.
Result:
Using logistic regression models, p‐Tau181 was shown to predict incident cognitive impairment in the African American population (Odds Ratio (OR)=1.52, p < 0.0001) but not in the Ibadan population (OR=1.09, p = 0.42) adjusting for age, sex, education, APOE4 genotype and history of stroke. The results from the logistic regression model for incident cognitive impairment/dementia can be seen in Table 1. Mixed effects models assessing cognitive scores showed significant interactions between time and p‐Tau181, indicating that higher p‐Tau181 is associated with greater decline for total cognitive score (p = 0.0109) and memory cognitive subscale (p = 0.0031) only in the Indianapolis African American population (Table 2).
Conclusion:
While geographic location has been seen to influence APOE genotype risk, similar investigations of AD biomarkers has been limited. Our data suggest there are differences in etiology that drives cognitive impairment between the African Americans in Indianapolis and the Africans in Ibadan. Additional analyses must be done to determine what drives these differences and the most effective biomarkers or biomarker combinations for differing populations
Longitudinal Implementation of a Neurodevelopmental Screening Program for Children Born to Mothers Living With HIV in Maternal-Child Clinics in Kenya: A Mixed Methods Study
No full textBackground:
Children with perinatal HIV exposure are at increased risk for neurodevelopmental (ND) delays, yet little is known about ND screening implementation for this population.
Methods:
This longitudinal study evaluated ND screening implementation at a health clinic in Kenya, from 9/2021 to 8/2023. Children aged 18-36 months with perinatal HIV exposure were screened using a 12-item general ND tool. Implementation outcomes—acceptability, feasibility, fidelity, and sustainability—were assessed through time-motion observations, clinic records, and semi-structured interviews with caregivers and staff.
Results:
Of 507 eligible children, 405 (80%) were screened. Screening rates were consistent over 24 months, with average time reduced to under 5 min. Facilitators included staff collaboration and caregiver support; barriers included time constraints and child temperament.
Conclusions:
ND screening was acceptable, feasible, and sustainable. Policymakers should embed ND screening within national child health programs, invest in workforce training and task-sharing models, strengthen referral and follow-up systems, and ensure affordable access to services
Safety and pharmacokinetics of teplizumab in children less than 8 years of age with stage 2 type 1 diabetes
No full textAims/hypothesis: Teplizumab is approved in the USA and seven other countries to delay stage 3 type 1 diabetes onset in individuals ≥8 years of age with stage 2 type 1 diabetes. As part of a US Food and Drug Administration post-marketing requirement, this study evaluated the safety, tolerability and pharmacokinetics of teplizumab in children aged <8 years with stage 2 type 1 diabetes.
Methods: The PETITE-T1D trial is a 2 year single-arm, open-label, multicentre study of 23 children <8 years of age with stage 2 type 1 diabetes. Participants received a 14 day teplizumab course. A prespecified interim analysis was performed after 15 participants completed 1 year of follow-up and included all 23 participants. Primary endpoints included treatment-emergent adverse events (TEAEs), TEAEs causing treatment discontinuation, and serious adverse events (SAEs). Other endpoints assessed immunogenicity, pharmacokinetics, pharmacodynamics and time from study treatment to stage 3 type 1 diabetes.
Results: Mean participant age was 4.8 years (range 1.7-6.8). Median follow-up duration was 51.9 weeks (range 3.9-77.1). All participants experienced one or more TEAE, with most being mild to moderate. No grade 4 or 5 TEAEs were reported. Three participants (13%) had TEAEs leading to teplizumab discontinuation: anaemia, elevated liver enzymes and maculo-papular rash. Two participants (9%) each had two SAEs. Serum teplizumab concentrations peaked at day 14. Two participants progressed to stage 3 type 1 diabetes. The estimated probability of lack of progression to stage 3 was 89.6% (95% CI 64.3%, 97.3%) at the time of interim analysis.
Conclusions/interpretation: Teplizumab was safe and well tolerated in children <8 years of age with stage 2 type 1 diabetes. Adverse events were consistent with those seen in previous studies, with no new safety risks identified. Two participants progressed to stage 3 type 1 diabetes during the observation period; surveillance is ongoing
Advice from adolescents with cancer, their caregivers, and clinicians on utilizing and improving patient portals in adolescent oncology care
No full textBackground: Adolescents with cancer have unique communication needs from other patients with cancer and adolescents without cancer. Electronic health record (EHR) patient portals offer opportunities for communication and health management, but pose challenges in balancing transparency, privacy, and health management. The 21st Century Cures Act amplified these challenges by increasing the availability of health records in patient portals. To inform portal use and policies, we analyzed advice from adolescents with cancer, their parents, and oncology clinicians regarding portal access and use.
Methods: We purposefully sampled adolescents aged 12-17 with cancer and their parents from large academic pediatric oncology centers and used convenience and snowball sampling to recruit clinicians from a national professional group. We conducted semi-structured interviews, descriptively coded transcripts to generate preliminary codes, categorized themes, and applied them across transcripts.
Results: We interviewed 38 adolescent-parent dyads and 53 clinicians. Interviews revealed agreement and tensions among stakeholders. Participants recommended a standard introduction of the portal, portal modifications to improve users' experiences, and encouraging portal use for health management and communication while maintaining patient access to clinicians. Adolescents and parents suggested clinicians modify notes to have a more hopeful tone. Clinicians favored delayed release of results, while parents preferred immediate release. Adolescents and clinicians encouraged adolescent control over parental proxy access, while parents favored default parental access.
Conclusion: This qualitative study provides actionable recommendations for improving portal use, policy, and design for adolescents with cancer. Implementation of these findings and future research can optimize adolescent and caregiver engagement with the portal and portal usability