Category: Statistics

Environ Sci Technol. 2026 Apr 22. doi: 10.1021/acs.est.6c03813. Online ahead of print.

ABSTRACT

Temperature regulates mollusk physiology and can alter metal bioaccumulation through filtration, uptake, growth dilution, and elimination. Yet many toxicokinetic (TK) applications treat temperature as a simple correction to a subset of rates and rarely account for trait- and context-dependence, limiting transferability across studies, seasons, and warming scenarios. Here we synthesize experimental evidence for marine univalve and bivalve mollusks and develop a temperature-aware framework that couples one-compartment mass-balance TK with temperature-dependent filtration and growth, while statistically linking absorption efficiency and elimination to temperature, species traits (e.g., body size), and metal chemical properties. Thermal responses in filtration and growth were captured with unimodal performance functions; machine learning was used for predictor screening. Evaluated against an independent data set, the framework reproduced internal concentrations across multiple orders of magnitude with good agreement (R² ≈ 0.73). Across the compiled evidence, filtration and growth showed strong species-specific thermal sensitivity, while metal chemistry primarily structured uptake. In the limited multitemperature TK calibration data set, a positive association between temperature and elimination was observed, but this relationship should be regarded as provisional pending additional multitemperature uptake-depuration data sets. By explicitly representing temperature-sensitive filtration and turnover pathways, the approach enables scenario testing for warming and heat extremes and provides a practical basis for improving the interpretation of bioaccumulation factors, seasonal biomonitoring, and temperature-aware risk assessment under climate change.

PMID:42019011 | DOI:10.1021/acs.est.6c03813

JMIR Mhealth Uhealth. 2026 Apr 22;14:e86836. doi: 10.2196/86836.

ABSTRACT

BACKGROUND: Cancer poses a significant threat to children’s health, and mobile health (mHealth) is emerging as a key tool for remote disease management, health education, and follow-up. However, evidence of its effectiveness remains limited.

OBJECTIVE: This study aimed to summarize the effects of mHealth interventions for pediatric cancer compared with usual care, providing evidence-based support for optimizing intervention models and improving patient outcomes.

METHODS: A systematic search of 14 databases identified randomized controlled trials (RCTs) on mHealth apps for pediatric patients with cancer from inception to August 1, 2025. Two reviewers independently screened studies, extracted data, assessed bias risk, and graded evidence quality. The meta-analysis was conducted using RevMan 5.4 and Stata 15.

RESULTS: A total of 24 RCTs involving 2645 patients were included. This review found that mHealth interventions significantly reduced infection rates (odds ratio [OR] 0.25, 95% CI 0.10-0.60; P=.002) and the overall incidence of peripherally inserted central catheter (PICC) complications (OR 0.16, 95% CI 0.10-0.24; P<.001), while improving quality of life (standardized mean difference [SMD] 1.34, 95% CI 0.13-2.55; P=.03), self-management ability (SMD 6.39, 95% CI 1.26-11.53; P=.01), and treatment adherence (OR 2.83, 95% CI 1.41-5.66; P=.003). However, mHealth interventions had no significant effect on PICC catheter displacement (OR 0.44, 95% CI 0.15-1.29; P=.13) or health knowledge (SMD 4.44, 95% CI -2.40 to 11.29; P=.20). Further high-quality studies are needed to verify their impact in these areas. The intervention components covered 9 behavior change techniques: goals and planning, feedback and monitoring, social support, shaping knowledge, repetition and substitution, reward and threat, comparison of outcomes, natural consequences, and regulation.

CONCLUSIONS: This systematic review and meta-analysis synthesized evidence from RCTs. The findings support the use of mHealth to reduce infections and PICC-related complications among pediatric patients with cancer while improving quality of life, self-management capabilities, and treatment adherence. These results underscore the importance of incorporating mHealth strategies into pediatric cancer care and guide the development and enhancement of future mHealth interventions.

PMID:42018994 | DOI:10.2196/86836

JMIR Cancer. 2026 Apr 22;12:e81579. doi: 10.2196/81579.

ABSTRACT

BACKGROUND: Fear of cancer recurrence (FCR) is prevalent among cancer survivors, affecting between 39% and 97% of patients. FCR is associated with impaired concentration, sleep disturbances, decreased quality of life, and increased psychological distress and health care use. To date, the literature lacks a review that summarizes the breadth of psychological interventions available for reducing fear of recurrence.

OBJECTIVE: This review aims to identify and summarize the evidence on psychological interventions for addressing FCR across all cancers.

METHODS: The Joanna Briggs Institute method for scoping reviews guided the processes, and we reported the review following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) checklist. We searched 5 databases (CINAHL, PsycInfo, the Cochrane Central Register of Controlled Trials, Embase, and MEDLINE) and 2 gray literature sources (ProQuest Dissertations & Theses Global and the World Health Organization International Clinical Trials Registry). Eligible studies included adults (≥18 years) diagnosed with cancer and evaluated psychological interventions aimed at reducing FCR. Data extraction captured study characteristics, intervention details, outcome effectiveness, and follow-up durations. We synthesized the findings using descriptive summaries and narrative analysis.

RESULTS: Overall, 5131 articles were screened, and 122 were included in this review; 48 (39.3%) involved patients with breast cancer, 47 (38.5%) focused on patients with multiple cancer types; over half of the studies (n=64, 52.5%) were randomized controlled trials. Only 28 (23%) studies explicitly reported the definition of FCR. Eighteen different measurement tools were used. Blended interventions (different combinations of cognitive behavioral therapy, mindfulness, acceptance and commitment therapy, and other strategies) formed the largest intervention category (n=38, 31.1%), followed by cognitive behavioral therapy interventions (n=26, 21.3%) and mindfulness-based interventions (n=24, 19.7%). Of the included studies, 104 (85.2%) demonstrated significant reductions in FCR. Most interventions were delivered face-to-face by disciplinary specialists (n=75, 61.5%), while some were delivered remotely (n=34, 27.9%), with the majority of these delivered via the website (n=18, 52.9%). Follow-up duration ranged from postintervention to 3 years.

CONCLUSIONS: FCR has been the focus of an increasing number of studies since 2009, with the majority being randomized controlled trials. Most interventions are delivered face-to-face and rely on trained specialists. Most have had statistically significant results. However, the included studies demonstrated heterogeneity in terms of delivery, duration, and dose, requiring cautious interpretation of intervention effects. Future research should develop consistent guidelines to standardize the definition of FCR, the measurement tools used, and the timing of follow-up assessments. Long-term follow-up data are needed to evaluate the sustained effects.

PMID:42018993 | DOI:10.2196/81579

JMIR AI. 2026 Apr 22;5:e78764. doi: 10.2196/78764.

ABSTRACT

BACKGROUND: In recent years, there has been increasing interest in developing machine and deep learning models capable of annotating clinical documents with semantically relevant labels. However, the complex nature of these models often leads to significant challenges regarding interpretability and transparency.

OBJECTIVE: This study aims to improve the interpretability of transformer models and evaluate the explainability of a deep learning-based annotation of coded clinical documents derived from death certificates. Specifically, the focus is on interpreting and explaining model behavior and predictions by leveraging calibrated confidence, saliency maps, and measures of instance difficulty applied to textualized representations coded using the International Statistical Classification of Diseases and Related Health Problems (ICD). In particular, the instance difficulty approach has previously proven effective in interpreting image-based models.

METHODS: We used disease language bidirectional encoder representations from transformers, a domain-specific bidirectional encoder representations from transformers model pretrained on ICD classification-related data, to analyze reverse-coded representations of death certificates from the US National Center for Health Statistics, covering the years 2014 to 2017 and comprising 12,919,268 records. The model inputs consist of textualized representations of ICD-coded fields derived from death certificates, obtained by mapping codes to the corresponding ICD concept titles. For this study, we extracted a subset of 400,000 certificates for training, 100,000 for testing, and 10,000 for validation. We assessed the model’s calibration and applied a temperature scaling post-hoc calibration method to improve the reliability of its confidence scores. Additionally, we introduced mechanisms to rank instances by difficulty using Variance of Gradients scores, which also facilitate the detection of out-of-distribution cases. Saliency maps were also used to enhance interpretability by highlighting which tokens in the input text most influenced the model’s predictions.

RESULTS: Experimental results on a pre-fine-tuned model for predicting the underlying cause of death from reverse-coded death certificate representations, which already achieves high accuracy (0.990), show good out-of-the-box calibration with respect to expected calibration error (1.40), though less so for maximum calibration error (30.91). Temperature scaling further reduces expected calibration error (1.13) while significantly increasing maximum calibration error (42.17). We report detailed Variance of Gradients analyses at the ICD category and chapter levels, including distributions of target and input categories, and provide word-level attributions using Integrated Gradients for both correctly classified and failure cases.

CONCLUSIONS: This study demonstrates that enhancing interpretability and explainability in deep learning models can improve their practical utility in clinical document annotation. By addressing reliability and transparency, the proposed approaches support more informed and trustworthy application of machine learning in mission-critical medical settings. The results also highlight the ongoing need to address data limitations and ensure robust performance, especially for rare or complex cases.

PMID:42018992 | DOI:10.2196/78764

JMIR Res Protoc. 2026 Apr 22;15:e83225. doi: 10.2196/83225.

ABSTRACT

BACKGROUND: Research is needed to better understand mental health (MH) problems among higher education (HE) students and how to address them. The Observatory on Student Mental Health in Higher Education (OSMHHE) brings together 350 members across Quebec (Canada) and internationally. Its mission is to develop, promote, and disseminate knowledge to foster a culture that supports student MH in HE.

OBJECTIVE: This study aims to describe the OSMHHE’s research protocol, which consists of three objectives: (1) establishing a portrait of students’ MH and its determinants, (2) identifying and evaluating a variety of MH practices, and (3) assessing the implementation and impacts of the OSMHHE as a knowledge mobilization infrastructure.

METHODS: Objective 1 will be achieved through a provincial survey using a cross-sectional, repeated-measures design with 2 data collections (November 2024 and 2026) targeting the entire Quebec HE student population. Dimensions, indicators, and scales were selected based on conceptual frameworks, a systematic literature review, and Delphi methods. Analyses will include descriptive statistics by education levels; inferential analyses comparing subpopulations; multiple regressions, logistic models, and linear mixed models to identify MH determinants; and repeated-measures ANOVA to examine temporal changes. Objective 2 will evaluate the implementation, sustainability, scale-up, and impacts of MH practices using mixed methods. Analyses may include descriptive and comparative statistics, correlations, structural equations modeling (path analysis), and qualitative content or thematic analyses. Objective 3 will draw on the framework of Ziam et al to assess knowledge mobilization strategies. A developmental evaluation approach and convergent mixed methods design within a case study will be used to assess the OSMHHE’s implementation and impacts. Qualitative data will include semistructured individual and group interviews with OSMHHE members, addressing topics such as roles, decision-making processes, facilitators and barriers, and outcomes. Additional qualitative sources will include diverse documents (eg, meeting agendas, reports). Quantitative data will come from questionnaires completed by members examining levels of engagement and satisfaction, challenges and barriers, and impacts of the OSMHHE’s activities and knowledge mobilization practices. Qualitative data will be analyzed using content analyses. Quantitative data will be examined using descriptive, comparative, and correlational analyses.

RESULTS: This project is funded from February 2023 to February 2028. The first provincial survey took place in November 2024, collecting data from 32,212 students in 77 HE institutions. Analyses are underway, and a first report was released in November 2025. Approximately 20 student MH practices are currently being evaluated.

CONCLUSIONS: The OSMHHE provincial survey will provide portraits of students’ MH in HE in Quebec and its determinants to better guide MH practices and institutional decision-making. Evaluating MH practices will advance knowledge of their effectiveness. Assessing the implementation of the OSMHHE will help deepen our understanding of knowledge mobilization infrastructures designed to support student MH in HE.

PMID:42018986 | DOI:10.2196/83225

JMIR Hum Factors. 2026 Apr 22;13:e87846. doi: 10.2196/87846.

ABSTRACT

BACKGROUND: Near miss events can reveal system problems before patients are harmed, but current reviews are inconsistent and often rely on simple counts that are distorted by patient volume and reporting culture. Consequently, leaders cannot tell whether a rise in reports means that safety is getting worse or that staff are reporting more, and current systems are not strong enough to clearly separate real safety risks from random variation.

OBJECTIVE: This study developed a 3-level near miss framework (NM³), a conceptual framework that converts descriptive near miss data into decision-grade intelligence through a structured, evidence-based process, including baseline measurement and advanced interpretation and governance.

METHODS: NM³ was developed to provide decision-grade analytics for acute inpatient hospital settings. The framework was designed as a maturity model, progressing from baseline measurement to advanced interpretation. It integrates standardized definitions, rate calculations, statistical process control, severity weighting, and learning metrics.

RESULTS: Level 1 establishes an organizational baseline through near miss rates per 1000 patient-days and near miss-to-harm ratios monitored with control charts. Level 2 introduces domain-specific denominators and unit-level charts to detect local variation. Level 3 applies severity weighting to generate a Near Miss Index; incorporates learning yields at 90 and 180 days; and triangulates near miss trends with harm events, exposure, reporting volume, and culture measures. A synthetic example demonstrates how the framework converts raw reports into stable rates, weighted indices, and learning metrics.

CONCLUSIONS: NM³ provides a structured pathway for organizations to strengthen near miss analytics. By progressing through maturity levels, leaders can improve the interpretation of safety signals, prioritize high-consequence risks, and integrate near miss reporting into governance.

PMID:42018985 | DOI:10.2196/87846

JMIR Hum Factors. 2026 Apr 22;13:e78939. doi: 10.2196/78939.

ABSTRACT

BACKGROUND: Mobile health (mHealth) apps are increasingly used to support healthy lifestyle behaviors through features such as health tracking and personalized reminders. Personalized messaging, tailored to users’ profiles, has been shown to improve engagement and retention in health-related contexts. Prior research has linked personality traits, based on the Big Five model, to preferences for specific app mechanisms, leading to the development of a preference matrix for personalizing mHealth apps. This matrix comprises 15 mechanisms derived from behavior change techniques and gamification elements, intended to guide developers in optimizing engagement according to user profiles.

OBJECTIVE: This study aimed to validate this preference matrix by examining whether the associations between mechanisms and Big Five personality traits reported in the literature align with user preferences observed in an experimental setting.

METHODS: A cross-sectional study was conducted using an online survey that collected demographic data, mHealth app usage, and personality traits. Participants were presented with mockups illustrating 15 mechanisms and were asked to select their preferred options. Logistic regression and ordinal logistic regression analyses were performed to examine associations between personality traits, mechanism selection, and motivation scores. All analyses were adjusted using the Bonferroni correction to account for multiple comparisons.

RESULTS: A total of 214 participants completed the survey (mean age 29.42, SD 10.41 y; n=118, 55.1% women; n=89, 41.6% men; n=5, 2% identifying as other; and n=2, 1% nonrespondents). Higher conscientiousness significantly increased the likelihood of selecting the collection mechanism (eg, collecting badges or points; odds ratio [OR] 1.87, 95% CI 1.27-2.75). For competition (eg, competing with other users), conscientiousness (OR 3.22, 95% CI 1.73-6.00) and agreeableness (OR 1.93, 95% CI 1.08-3.45) were significant predictors. Preferences for rewards (eg, virtual incentives such as points or virtual currency) were associated with conscientiousness (OR 2.36, 95% CI 1.53-3.63) and neuroticism (OR 1.97, 95% CI 1.36-2.86). Additionally, 4 mechanisms-self-monitoring, progression, challenge, and quest-were selected by more than half of the participants, independent of personality traits.

CONCLUSIONS: The findings partially validate the proposed preference matrix. Conscientiousness consistently emerged as a key predictor of preference across multiple mechanisms, highlighting its central role in engagement with gamified mHealth features. While some mechanisms appear to have universal appeal, others show personality-specific preferences, underscoring the value of combining baseline mechanisms with targeted personalization strategies in mHealth app design.

PMID:42018983 | DOI:10.2196/78939

JMIR Res Protoc. 2026 Apr 22;15:e91976. doi: 10.2196/91976.

ABSTRACT

BACKGROUND: The COVID-19 pandemic has a lasting impact on health care utilization, as both the acute infection and post COVID condition (PCC) can lead to increased demand for medical services due to ongoing symptoms.

OBJECTIVE: The aim of this study is to systematically examine health care utilization among individuals after acute SARS-CoV-2 infection in Bavaria, Germany, with a particular focus on PCC. The study combines claims data analysis with qualitative interviews to improve the understanding of objective care pathways and patients’ subjective experiences within the health care system.

METHODS: The research project ‘SOLongCOVID’ employs a mixed-methods design consisting of two subprojects: (1) a retrospective cohort study using claims data from the Bavarian Association of Statutory Health Insurance Physicians (KVB) to analyze care pathways through state sequence analysis, (2) a qualitative study based on semistructured interviews and focus groups with patients with PCC concerning their subjective care experiences. A synthesis process involving a focus group discussion will combine the information from the two subprojects, providing a comprehensive understanding of the care processes of patients with PCC.

RESULTS: The study was funded by the German Federal Joint Committee Innovation Fund in October 2024. Statutory health insurance claims data cover the period from 2019 to 2022, and qualitative interview data collection is planned from May 2025 to August 2026. As of manuscript submission, study preparation and ethics approvals have been completed, and 14 participants have been recruited for the qualitative interviews. Study findings are anticipated to be published from July 2026 to August 2027.

CONCLUSIONS: The results are expected to enhance the understanding of existing barriers and challenges and to support evidence-based recommendations for improving care pathways for patients with specific care needs.

PMID:42018978 | DOI:10.2196/91976

JMIR Res Protoc. 2026 Apr 22;15:e84909. doi: 10.2196/84909.

ABSTRACT

BACKGROUND: Cesarean section (CS) requires perioperative antibiotic prophylaxis (PAP) for the prevention of surgical site infections. However, systemic antibiotics during the peripartum period may induce compositional perturbations of the maternal gut microbiome, a system already characterized by reduced resilience. Data on maternal gut microbiome dynamics after CS with PAP are scarce, largely due to logistical and feasibility barriers that limit the participation of pregnant women and new mothers in conventional clinical studies.

OBJECTIVE: This protocol primarily aims to evaluate the feasibility of a fully decentralized, remote study design for longitudinal gut microbiome research in the peripartum period. Secondary exploratory objectives include the comparative analyses of microbiome composition between CS with PAP and vaginal delivery (VD) without antibiotic exposure to inform future adequately powered studies.

METHODS: The MAMA (Microbiome Changes Due to Antibiotic Prophylaxis in Mothers at Birth) study is a prospective, 2-arm observational cohort study conducted entirely off-site. Women in the third trimester of pregnancy were recruited at 2 German level-1 perinatal centers and affiliated outpatient facilities. Participants underwent either CS with PAP (single dose cefuroxime 1.5 g intravenously) or VD without antibiotics. Stool samples were self-collected at home and returned by mail at 3 predefined time points: late pregnancy (T0), 2 to 3 days post partum (T1), and 90±10 days post partum (T2). Primary outcomes are feasibility indicators, including recruitment rate, sample and questionnaire return rates at each time point, adherence to sampling windows, and participant retention across follow-up. Secondary outcomes are exploratory microbiome measures based on 16S rRNA gene sequencing (V3-V4), including alpha diversity indices, beta diversity metrics, and relative taxonomic abundances. Microbiome analyses are explicitly compositional and hypothesis-generating. Group comparisons and longitudinal within-individual changes will be assessed using nonparametric diversity metrics and multivariate distance-based methods. No confirmatory hypothesis testing is planned.

RESULTS: Recruitment occurred between May 2022 and October 2023, with 37 women enrolled (25 CSs and 12 VDs). Follow-up was completed with receipt of the final stool sample in March 2024. DNA extraction and sequencing were completed in a single batch in October 2024. Bioinformatic processing and statistical analyses were initiated in June 2025 and are ongoing as of December 2025. Results from the exploratory microbiome analyses are expected to be published in 2026.

CONCLUSIONS: This protocol demonstrates the feasibility of conducting fully decentralized, longitudinal microbiome research in a peripartum population without requiring on-site visits. By integrating study procedures into maternal realities, the remote design reduces participation barriers and addresses a clinically relevant research gap that has remained largely unexamined despite routine use of PAP. While microbiome-related outcomes are exploratory, the methodological framework established here provides a scalable model for future maternal and postpartum research, supporting ethically grounded, participant-centered study designs and evidence-informed care strategies.

PMID:42018976 | DOI:10.2196/84909

Neurology. 2026 May 26;106(10):e214989. doi: 10.1212/WNL.0000000000214989. Epub 2026 Apr 22.

ABSTRACT

BACKGROUND AND OBJECTIVES: Implementation of telemedicine expanded options for outpatient neurology care. It remains uncertain which new neurology patients can be appropriately evaluated virtually. We compared subsequent health care utilization after virtual vs in-person new patient neurology visits across 3 academic medical centers.

METHODS: We conducted a retrospective multicenter cohort study of adults with a new outpatient neurology visit from September 2020 through December 2021 using the Vizient Clinical Data Base and Clinical Practice Solutions Center databases. Virtual and in-person patients were matched 1:1 using propensity scores incorporating demographics, clinical characteristics, time period, and previous health care utilization. Outcomes were analyzed overall and stratified by neurologic chief complaint category and institution. We compared rates of subsequent neurologic clinic follow-up, emergency department (ED) visits, and hospitalizations after virtual and in-person encounters. Testing and all-cause ED visits/hospitalizations were also assessed.

RESULTS: We identified 10,428 virtual and 36,767 in-person neurology new outpatient visits. After propensity score matching, 8,202 virtual visits were matched to 8,202 in-person visits. Neurology follow-up within 90 days did not differ between virtual and in-person visits (24.6% vs 23.7%, p = 0.18). Thirty-day neurology clinic follow-up was slightly lower after virtual visits, whereas follow-up at 6 months and 1 year was similar between groups. Neurologic ED visits and hospitalizations within 90 days were similar (0.9% vs 0.8%, p = 0.23 and 1.8% vs 1.7%, p = 0.47, respectively). All-cause ED visits and hospitalizations within 90 days were also comparable (1.8% vs 1.7%, p = 0.59 and 2.2% vs 1.8%, p = 0.13, respectively). Analyses by chief complaint found that 90-day follow-up was higher after in-person visits for dementia, whereas 30- and 90-day follow-up was higher after virtual visits for Parkinson disease and multiple sclerosis, and 90-day follow-up was higher after virtual visits for headache. Testing was more frequent after in-person visits for certain chief complaints.

DISCUSSION: In this propensity score-matched multicenter cohort, new neurology patients seen virtually had similar downstream utilization as those seen in-person, including comparable 90-day follow-up and similar neurologic and all-cause ED visits and hospitalizations. Although follow-up varied modestly by chief complaint and testing was more frequent after some in-person visits, no major differences emerged overall.

PMID:42018961 | DOI:10.1212/WNL.0000000000214989