Category: Statistics

JMIR Res Protoc. 2026 Aug 21;15:e94740. doi: 10.2196/94740.

ABSTRACT

BACKGROUND: The widespread use of pesticides in agriculture raises concerns about exposure among populations living near treated fields. Environmental contamination may lead to prolonged pesticide exposure, with potential health effects, yet significant knowledge gaps remain regarding exposure levels and pathways in agricultural communities.

OBJECTIVE: The Exposure to Pesticides Used in Regions of Agriculture (EPURA) study aims to characterize pesticide exposure patterns and identify their key sources and determinants through an integrated approach combining biological and environmental measurements.

METHODS: We recruited 400 participants from agricultural communities living within 1500 meters of vegetable production fields in Quebec, Canada, throughout the 2 years of data collection. Over 4 summer months and 3 winter months, participants provided monthly urine samples and floor wipes and completed questionnaires on diet, water consumption, household characteristics, pesticide use, and lifestyles. Summer and winter urine samples were pooled for each participant, and summer and winter wipe samples were pooled for each household. At the end of the summer, we collected additional samples, including indoor dust, yard soil, and tap water. Environmental and biological samples will be analyzed for 51 pesticides and their metabolites. Pesticide applications near residences will be estimated using agricultural land use data integrated into a geographic information system. Pesticide intake from food and water will be assessed by combining consumption data with pesticide residues measured in water and in food monitoring programs. Residential contamination will be estimated by measuring pesticides in indoor dust and yard soil. Urinary pesticide concentrations will be used to estimate exposure doses. Urinary concentrations will be compared across population subgroups and seasons and with available Canadian biomonitoring data. Associations between exposure doses and dietary and water intake, indoor and outdoor contamination, proximity to treated fields, and household-level determinants will be examined using mixed models. Variable selection will be performed using high-dimensional methods.

RESULTS: As of June 1, 2026, participant recruitment and data collection are complete. A total of 173 households (372 participants) were included in the study, with 1049 wipe samples, 170 sealed petri dishes, 157 water samples, 170 yard soil samples, and 2257 urine samples collected. Completion of all laboratory analyses is expected in 2027, after which the statistical analyses will be conducted in accordance with the study objectives.

CONCLUSIONS: This study will generate crucial data on pesticide exposure in agricultural communities. Its findings will help identify major determinants of household contamination and key risk factors contributing to residents’ pesticide exposure. Ultimately, the EPURA study will support the development of targeted strategies to reduce pesticide-related health risks and inform public health policies.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/94740.

PMID:42628033 | DOI:10.2196/94740

JMIR Form Res. 2026 Aug 21;10:e94172. doi: 10.2196/94172.

ABSTRACT

BACKGROUND: Cesarean section (C-section) is the most common surgical procedure in the United States, yet its use varies widely across regions and institutions. Although clinical risk factors are central to delivery decisions, geographic context, health system capacity, and local practice patterns may also influence C-section use. Understanding both the determinants and predictability of C-section delivery is important for improving obstetric quality and equity.

OBJECTIVE: This study aims to document geographic variation in C-section use across the United States, identify maternal and county-level factors associated with C-section delivery, and evaluate the predictive performance of machine learning models across clinically defined risk groups.

METHODS: This population-based study used 38,133,279 US births from the 2013-2022 National Vital Statistics System Natality Detailed Files. County identifiers were linked to national county-level measures of insurance coverage, health care capacity, and socioeconomic conditions. Logistic regression models with county fixed effects were used for feature interpretation, and supervised machine learning models were used for prediction. Analyses were conducted separately for the full sample, a low-risk sample (n=17,760,772), and a high-risk sample (n=20,372,438). Predictive performance was evaluated using accuracy, precision, recall, F₁-score, and area under the receiver operating characteristic curve (AUC), with 200-bootstrap 95% CIs. Temporal validation was also conducted using training on earlier years and testing on later years.

RESULTS: County-level C-section rates declined modestly from 32.35% in 2013 to 31.49% in 2022, but substantial geographic variation persisted, with consistently higher rates in the US South. In the full sample, model discrimination was good, with AUC values ranging from 0.8310 for logistic regression to 0.8401 for extreme gradient boosting (XGBoost). Predictive performance was substantially weaker in the low-risk sample (AUC 0.7246-0.7410) than in the high-risk sample (AUC 0.8404-0.8568). In the high-risk sample, XGBoost achieved the highest AUC (0.8568) and F₁-score (0.7579), while random forest achieved the highest recall (0.7142). Temporal validation yielded similar results in the full sample (AUC 0.8335-0.8387), temporal low-risk sample (AUC 0.7364-0.7432), and temporal high-risk sample (AUC 0.8360-0.8479), indicating stable performance over time.

CONCLUSIONS: C-section use in the United States is shaped by both maternal clinical risk and geographic context. Machine learning models perform well overall and especially well in high-risk pregnancies, but prediction is substantially more difficult in low-risk pregnancies, where discretionary and contextual influences may play a larger role. These findings support the use of risk-adjusted, context-aware prediction tools for audit, benchmarking, and clinical decision support while underscoring the need for cautious implementation, subgroup monitoring, and further external validation.

PMID:42628029 | DOI:10.2196/94172

JMIR Res Protoc. 2026 Aug 21;15:e90946. doi: 10.2196/90946.

ABSTRACT

BACKGROUND: Rehabilitation is beneficial for individuals with cancer across all phases of care, but access remains limited for many survivors, including individuals with melanoma. As new treatment strategies for melanoma are introduced, there is an opportunity to assess perioperative rehabilitation strategies for patients undergoing surgery during and after neoadjuvant treatment.

OBJECTIVE: This paper describes the protocol for a 2-armed pilot randomized controlled trial with the overall objective of determining the feasibility and preliminary effects of a perioperative rehabilitation strategy for individuals with melanoma at a single cancer institution in Ontario, Canada.

METHODS: In total, 30 participants, including English-speaking adults with a current melanoma diagnosis scheduled to undergo surgery after neoadjuvant immunotherapy, will be randomized 1:1 to either the intervention or control group. The intervention will include 4 perioperative intervention sessions (2 sessions before surgery and 2 sessions after surgery) led by a trained physiotherapist or kinesiologist. Participants in the intervention group will receive education on the benefits of exercise; how to exercise safely; and rehabilitation techniques to maximize range of motion, strength, and function. Participants will also receive a tailored exercise program and will establish goals and action plans at each session. The control group will receive usual care (no rehabilitation). The primary outcome is feasibility (measured via recruitment, retention, and adherence rates). Secondary outcomes will be collected before and after the intervention and include overall impairment score (measured via the Edmonton Symptom Assessment Scale), physical activity level (Godin Leisure Time Exercise Questionnaire), functional mobility (6-minute walk test and 30-second sit-to-stand test), range of motion (goniometry), grip strength (handheld dynamometry), quality of life (Functional Assessment of Cancer Therapy-Melanoma), perception of health status (EQ-5D-3L), and medication use. Feasibility will be analyzed using descriptive statistics (mean [SDs] or frequency [%]), as appropriate. Secondary outcomes will be reported using means, SD, and CIs of outcomes. Cohen d effect sizes will be calculated and reported for each secondary outcome to inform a future full-scale trial. STATA/MP (version 14) will be used for all statistical analysis, with significance set at P<.05.

RESULTS: Trial enrollment began in June 2026, and 3 participants have been recruited to date. Recruitment is expected to be completed by December 2026. Results will be submitted for publication by mid or late 2027.

CONCLUSIONS: This pilot trial aims to address the evidence gap on the feasibility and preliminary effects of rehabilitation for individuals with melanoma, laying the groundwork for a larger randomized controlled trial and future implementation of rehabilitation for this population. By implementing rehabilitation around the surgical phase, we aspire to enhance patient outcomes by maximizing function and reducing side effects, facilitating a faster return to meaningful daily activities.

TRIAL REGISTRATION: ClinicalTrials.gov NCT07320222; https://clinicaltrials.gov/study/NCT07320222.

INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): PRR1-10.2196/90946.

PMID:42628027 | DOI:10.2196/90946

J Clin Rheumatol. 2026 Aug 20. doi: 10.1097/RHU.0000000000002400. Online ahead of print.

ABSTRACT

INTRODUCTION: In recent years, PANLAR has addressed gender equity.

OBJECTIVES: To analyze the characteristics of speakers participating in PANLAR congresses and to compare the distribution of speakers by gender.

METHODS: Speakers who participated in the annual PANLAR congresses from 2021 to 2024 were identified by reviewing the corresponding scientific programs. The following data were recorded: age, sex, nationality, country of residence, speaker category, type of participation (moderator, speaker at general conferences, or speaker at industry-sponsored symposia), years in the specialty, and workplace. Statistical analyses included descriptive statistics, the χ2 or Fisher exact test, and the Student t test.

RESULTS: A total of 430 speakers participated in the four annual PANLAR congresses held from 2021 to 2024. The median age was 48 years (IQR: 40 to 60), and 231 speakers (53.7%) were men. Of these, 337 were adult rheumatologists (79.3%). The main countries of origin of the speakers were Colombia (21.6%), Brazil (20.9%), and Argentina (13.7%). Among the speakers, 114 participated as moderators, 337 as speakers at general conferences, and 67 as speakers at pharmaceutical industry-sponsored symposia. Among rheumatologists who treated adult patients, the proportion of men was higher than that of women in both moderator roles and general conferences, although these differences were not statistically significant (men 58.4% vs. women 41.2%, and men 53.6% vs. women 46.4%, respectively; p=NS). Men participated significantly more often than women in pharmaceutical industry-sponsored symposia (men 71.4% vs. women 28.6%; p=0.005). The mean numbers of moderator roles, general lectures, and industry-sponsored symposia were also significantly higher among men than among women.

CONCLUSION: Men accounted for a higher proportion of moderators and speakers at PANLAR annual meetings during the study period, with a more pronounced gender disparity in pharmaceutical industry-sponsored symposia.

PMID:42628024 | DOI:10.1097/RHU.0000000000002400

JMIR Hum Factors. 2026 Aug 21;13:e83371. doi: 10.2196/83371.

ABSTRACT

BACKGROUND: With Türkiye’s aging population and widespread internet usage, the excessive seeking of health information through online platforms, formally identified as cyberchondria, has emerged as a concern affecting older individuals’ health perception and psychological well-being.

OBJECTIVE: This study aimed to evaluate the association of cyberchondria with health perceptions and psychological well-being among older participants in a third-age university program.

METHODS: This cross-sectional study included 352 participants aged ≥60 (mean 67.4, SD 4.7) years from Tazelenme University, Antalya, Türkiye, and was conducted between November and December 2024. Data were collected using the Cyberchondria Severity Scale Short Form (CSS-12), Individual Health Perception Scale, and Psychological Well-Being Scale for Older People (PWBS-OP). Statistical analyses included correlation and multiple regression analyses.

RESULTS: The mean age of the 352 participants was 67.4 (SD 4.7, range 60-87) years, 38.6% (n=136) were first-year students, 69.6% (n=245) were men, 58.8% (n=207) were married, 48.9% (n=172) were university graduates or above, and 65.1% (n=229) had chronic diseases. Significant negative correlations were found between CSS-12 distress, compulsion, and total scores with both Individual Health Perception Scale and PWBS-OP scores (P<.05). According to the multiple linear regression analysis, the presence of chronic disease was the only significant positive predictor of higher health perception levels (B=2.16, 95% CI 1.721-3.593; P=.003), whereas factors such as age, gender, education level, and CSS-12 total scores did not demonstrate a statistically significant impact (P>.05). Psychological well-being (PWBS-OP) was significantly and positively predicted by age (B=0.22, 95% CI 0.037-0.399; P=.02), marital status (B=1.67, 95% CI 1.064-2.716; P=.002), economic status (B=2.10, 95% CI 1.729-3.464; P=.003), and the thought of having an undiagnosed disease (B=3.38, 95% CI 1.038-5.830; P=.007), whereas it was significantly and negatively predicted by medical examinations in the past year (B=-1.29, 95% CI -2.093 to 0.485; P=.002), undergoing examinations without a physician’s recommendation (B=-2.04, 95% CI -3.794 to -0.290; P=.02), searching for health-related topics on the internet (B=-1.26, 95% CI -2.152 to -0.360; P=.006), and CSS-12 total scores (B=-0.13, 95% CI -0.261 to -0.060; P=.04).

CONCLUSIONS: High cyberchondria levels significantly impair older adults’ psychological well-being and health perception. While demographic factors positively influence health perception, excessive internet-based health seeking deteriorates psychological well-being. Digital health literacy programs, professional online health counseling, and psychological support should be recommended to target cyberchondria risks in older populations.

PMID:42628018 | DOI:10.2196/83371

JMIR Med Inform. 2026 Aug 21;14:e83099. doi: 10.2196/83099.

ABSTRACT

BACKGROUND: Acute chest pain (ACP) is one of the most common chief complaints in the emergency department (ED), accounting for approximately 8% of all ED visits. However, among patients presenting with chest pain suggestive of cardiac origin, fewer than 10% are ultimately diagnosed with acute coronary syndrome (ACS).

OBJECTIVE: AI models were developed to support clinical decision-making for triage level-2 ED patients presenting with ACP. These models aim to accurately detect ACS and reliably identify low-risk patients based on a single high-sensitivity cardiac troponin T test result. By integrating this AI-assisted strategy into clinical workflows, we aim to reduce ED length of stay, alleviate crowding, and improve health care efficiency while maintaining high safety standards.

METHODS: We conducted a retrospective study using single-center data from a tertiary teaching hospital between January 2016 and December 2022. Models based on artificial neural networks (ANNs) were trained to classify patients with ACP at triage level 2 into 3 clinical classes: critical patients with ACS (subgroup GA), critical patients without ACS (subgroup GB1), and low-risk patients (subgroup GB2). The models were trained and internally validated via a 5-fold cross-validation protocol using data from 2016 to 2020. Model performance was then evaluated on the hold-out testing data (2021-2022) using area under the receiver operating characteristic curve, area under the precision-recall curve, and subgroup-specific metrics. Feature selection methods and the Shapley Additive Explanations value analysis were applied to identify features that drive the model’s predictive power. The study was approved by the institutional review board of the National Cheng Kung University Hospital, Tainan, Taiwan (A-ER-111-199).

RESULTS: After excluding ED visits with missing triage data, incomplete medical histories, or unsuitable dispositions, 17,935 visits were included, with 1209 GA, 3873 GB1, and 12,853 GB2. Rendering prediction based on 24 feature variables (2 demographics, 6 vital signs, 4 blood test results, and 12 medical history), all ANN models demonstrated strong testing AUROC (95% CI) performance of 0.942 (0.920-0.965) for GA classification, 0.824 (0.808-0.841) for GB1, and 0.893 (0.884-0.902) for GB2. Regarding multiclassification, ANN-S3 achieved balanced performance, with ACS sensitivity of 0.941 (95% CI 0.908-0.973) and low-risk positive predictive value and sensitivity of 0.911 (95% CI 0.9-0.921) and 0.837 (95% CI 0.824-0.85), respectively. A sensitivity-prioritized variant, ANN-S3-L, increased ACS sensitivity to 0.966 (95% CI 0.94-0.991) and negative predictive value to 0.998 (95% CI 0.996-0.999), but at the cost of lower specificity and reduced low-risk sensitivity, indicating a safety-efficiency trade-off.

CONCLUSIONS: These findings suggest that ANN-based classifiers can effectively support clinical risk stratification and disposition decision-making in ACP care for patients presenting ≥3 hours after symptom onset. However, because even the sensitivity-prioritized ANN-S3-L variant falls short of the stringent sensitivity threshold (>0.99) typically required for a standalone ED rule-out tool, this system should be interpreted as a clinical decision-support aid rather than an independent rule-out strategy.

PMID:42628009 | DOI:10.2196/83099

JMIR AI. 2026 Aug 21;5:e68317. doi: 10.2196/68317.

ABSTRACT

BACKGROUND: Palliative care is increasingly recognized as essential for an aging population and rising life-limiting illnesses. Machine learning (ML) has been widely applied in this field, primarily for prognostication. However, recent literature suggests broader applications that may enhance patient-centered care and optimize system-level processes.

OBJECTIVE: This study aimed to map and summarize the evolving landscape of ML applications in palliative care through a scoping review, identifying how studies extend beyond mortality prediction into new domains, while assessing explainability, equity, and implementation readiness.

METHODS: We conducted a scoping review following the Arksey and O’Malley framework and PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines. Six databases (MEDLINE, PsycINFO, Embase, CINAHL, Scopus, and Web of Science) were searched from inception to April 15, 2021, with an update through February 9, 2026. Included studies were peer-reviewed primary studies applying ML to palliative care contexts. Each study was coded for explainable AI (XAI) methods, equity considerations, and implementation readiness. Two reviewers independently screened and extracted data. Synthesis combined descriptive statistics and inductive thematic analysis. Consistent with scoping review methodology, no formal risk-of-bias assessment was performed.

RESULTS: We included 121 studies (2015-2026) spanning 24 countries, with 69.4% (84/121) published from 2021 onward. The United States contributed the largest share (66/121, 54.5%), followed by Japan, Taiwan, and China (22/121, 18.2%). Cancer was the most commonly studied population (52/121, 43%). Supervised classification was the most common approach (84/121, 69.4%), followed by natural language processing and text mining (16/121, 13.2%). Six application domains were identified: mortality and survival prediction (51/121, 42.1%), health care use (25/121, 20.7%), symptom assessment and phenotyping (20/121, 16.5%), communication and natural language processing (16/121, 13.2%), clinical decision support and care quality (6/121, 5%), and other (3/121, 2.5%). Approximately half of the studies (61/121, 50.4%) used at least one XAI technique, most commonly feature importance rankings and SHAP (Shapley Additive Explanations) values. Among the 121 studies, equity in model performance was fully addressed in only 8 (6.6%) studies, partially in 8 (6.6%) studies, and not addressed in 105 (86.8%) studies. Two-thirds of studies (80/121, 66.1%) remained at the proof-of-concept stage, while 16.5% (20/121) achieved external validation and 17.4% (21/121) reached prospective deployment or clinical integration.

CONCLUSIONS: ML applications in palliative care are expanding beyond prognostication toward patient-centered uses, including symptom management, clinical decision support, and resource planning. The persistent gap in equity reporting (105/121, 86.8% did not report equity considerations) signals that the field risks developing tools that may not perform equitably across diverse populations. While half of studies now use XAI techniques, fewer than 1 in 5 studies (21/121, 17.4%) have reached clinical integration. Bridging this translational gap requires systematic attention to implementation science, equity auditing, and explainability reporting.

PMID:42628008 | DOI:10.2196/68317

JCO Precis Oncol. 2026 Aug;10(8):e2501012. doi: 10.1200/PO-25-01012. Epub 2026 Aug 21.

ABSTRACT

PURPOSE: A six-gene panel involved in androgen production, uptake, and conversion (APUC-6: HSD3B1, HSD3B2, CYP3A43, CYP11A1, CYP11B1, CYP17A1) may define distinct clinical outcomes in metastatic prostate cancer. This study evaluated the prognostic and predictive value of APUC-6 expression in metastatic castration-sensitive prostate cancer (mCSPC).

METHODS: Transcriptomic data from 160 patients in the phase III Eastern Cooperative Oncology Group-ACRIN E3805 CHAARTED trial were analyzed. Patients were stratified into four subgroups based on APUC-6 and androgen receptor (AR) gene expression. Clinical outcomes including time to clinical progression (ttCP), time to castration resistance (ttCR), and overall survival (OS) were assessed within the androgen deprivation therapy (ADT) and docetaxel plus ADT (D-ADT) arms. Interaction testing evaluated the predictive value of APUC-6/AR status. Subgroup analyses were also performed by Decipher genomic risk.

RESULTS: Patients with APUC-6-high/AR-low expression showed significantly improved ttCP (hazard ratio [HR], 0.43 [95% CI, 0.22 to 0.82]; P = .0092), ttCR (HR, 0.56 [95% CI, 0.32 to 1]; P = .049), and OS (HR, 0.31 [95% CI, 0.16 to 0.61]; P = .0003), but did not appear to benefit from the addition of docetaxel. Conversely, APUC-6-low/AR-low patients demonstrated significantly improved outcomes with D-ADT (ttCP: HR, 0.37 [95% CI, 0.21 to 0.65]; P = .00036; ttCR: HR, 0.38 [95% CI, 0.23 to 0.64]; P = .00017; OS: HR, 0.39 [95% CI, 0.21 to 0.72]; P = .0019). Statistical interaction analysis supported that patients with APUC-6-low/AR-low expression were more likely to benefit from docetaxel (HR, 0.36 [95% CI, 0.19 to 0.66]; P = .001). For patients with a high Decipher genomic classifier score, those in the APUC-6-low subgroup had improved survival with D-ADT, whereas those in the APUC-6-high subgroup did not gain any additional benefit from docetaxel.

CONCLUSION: APUC-6 expression and AR gene expression define clinically relevant subgroups in mCSPC and may enhance precision medicine treatment approaches in combination with other transcriptomic biomarkers.

PMID:42628007 | DOI:10.1200/PO-25-01012

JMIR Res Protoc. 2026 Aug 21;15:e87715. doi: 10.2196/87715.

ABSTRACT

BACKGROUND: Sacroiliac joint (SIJ) dysfunction is associated with significant pain, disability, and altered postural control in adults. Manual therapies, including Mulligan mobilization, Maitland mobilization, high-velocity thrust (HVT), and muscle energy technique (MET), are routinely applied for the management of SIJ dysfunction. However, comparative evidence regarding their effects on pain, disability, and center-of-gravity (COG) displacement remains insufficient.

OBJECTIVE: This study aims to compare the efficacy of Mulligan mobilization, Maitland mobilization, HVT, and MET in improving pain, functional disability, and COG displacement in individuals with SIJ dysfunction.

METHODS: A randomized controlled, parallel-group trial will be conducted on participants diagnosed with unilateral SIJ dysfunction. Participants will be assigned in equal numbers to 4 groups, each receiving Mulligan mobilization, Maitland mobilization, HVT, or MET interventions administered 5 days per week for 3 weeks, in addition to a standardized exercise regimen. The primary outcomes will be pain (assessed using a visual analog scale [VAS]) and functional disability (assessed using the Modified Oswestry Disability Index [MODI]), and the secondary outcome will be COG displacement (measured by sway velocity, lift-up index on the left and right, COG alignment, movement time, and impact index).

RESULTS: We hypothesize that all interventions will lead to statistically significant reductions in pain and disability. Results will be analyzed after the completion of the study.

CONCLUSIONS: This protocol is designed to provide robust comparative data on manual therapy approaches for SIJ dysfunction. Findings may guide physiotherapy practice by informing evidence-based recommendations for optimal management strategies.

PMID:42627998 | DOI:10.2196/87715

Lab Med. 2026 Aug 4;57(5):lmag051. doi: 10.1093/labmed/lmag051.

ABSTRACT

INTRODUCTION: Errors during blood transfusion continue to contribute to preventable adverse events, and nurses often have gaps in their transfusion-related knowledge and practices. This study evaluated the effect of simulation-based education on nurses’ knowledge and practices regarding blood transfusion.

METHODS: A quasiexperimental, single-group, pretest/post-test study was conducted among 227 registered nurses at a tertiary-care hospital. A structured questionnaire was used to assess knowledge (23 items, maximum score = 30) and self-reported practices (15 items, maximum score = 75) before and after 4-hour simulation-based education. The intervention consisted of didactic review, hands-on skills stations, and high-fidelity simulation scenarios. Data were analyzed using paired t tests, independent t tests, and 1-way analysis of variance.

RESULTS: The mean (SD) postintervention total score (54.8 [7.1]) demonstrated a statistically significant increase of 20.0 points from the preintervention score (34.8 [10.5]) (t226 = -61.5; P < .001], with a large effect size (Cohen d = 2.24). Statistically significant improvements were observed in both the knowledge and practice domains (P < .001). Analysis revealed that nurses with less than 5 years of clinical experience and those in the 20 to 29 years of age group had statistically significantly lower baseline scores but showed the greatest magnitude of improvement (P < .001). Prior formal training showed no statistically significant association with baseline knowledge or improvement.

DISCUSSION: Simulation-based education substantially improved nurses’ transfusion-related knowledge and practices.

PMID:42627984 | DOI:10.1093/labmed/lmag051