SEARCH · Search PubMed

Results for “Model interpretability”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

460 recordsLinked to original sources

PubMed2025

Enhanced fracture detection on radiographs with AI assistance for clinicians: a systematic review and meta-analysis.

BACKGROUND: Emergency radiographic interpretation for fractures is prone to missed or misdiagnoses. Artificial intelligence (AI) is expected to become a powerful tool to assist clinicians in fracture detection. PURPOSE: A systematic review and meta-analysis was performed to assess whether AI improves clinicians' ability to detect fractures on radiographs. MATERIALS AND METHODS: A literature search was conducted in PubMed, Web of Science, and Cochrane Library for studies published between January 1, 2010, and October 10, 2025. A meta-analysis of diagnostic accuracy studies was performed using a Summary Receiver Operating Characteristic (SROC) curve. The quality of included studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) tool. Subgroup analysis and meta-regression were conducted to explore potential sources of heterogeneity. RESULTS: A total of 26 studies were included . The pooled sensitivity of clinicians increased from 77% (95% CI: 72-81) to 87% (95% CI: 83-90) with AI assistance, while the pooled specificity improved from 88% (95% CI: 85-90) to 92% (95% CI: 89-94). The corresponding AUC values were 0.90 (95% CI: 0.87-0.92) before and 0.95 (95% CI: 0.93-0.97) after AI assistance. Eight studies were rated as high risk of bias. Subgroup analysis and meta-regression identified potential sources of heterogeneity, including fracture location, AI model type, high risk of bias, and reference standards. CONCLUSION: AI assistance significantly improves clinicians' diagnostic performance in detecting fractures on radiographs for extremity and trunk fractures.

Humans↗

PubMed2026

A mechanism-guided framework for prioritizing membrane-interaction anti-Vibrio peptides from peptidomics data.

A mechanism-guided framework for prioritizing membrane-interaction antimicrobial peptide candidates from proteomics-derived peptide mixtures is presented. The framework integrates conservative machine-learning-based antimicrobial peptide (AMP) screening with a literature-derived membrane-interaction plausibility (MAP) assessment and a data-driven membrane-interaction ranking function (AIPx), followed by structural visualization for interpretability. MAP encodes physicochemical characteristics commonly associated with peptide-membrane interaction and provides a graded plausibility assessment. Building upon this physicochemically interpretable framework, AIPx ranks peptides using feature weights calibrated from experimentally characterized anti-Vibrio peptides, where minimum inhibitory concentration (MIC) values are used as a coarse-grained ranking reference rather than a direct prediction target. In a peptidomics-based peptide fractionation study targeting Vibrio spp., AIPx exhibited a consistent relationship with experimentally observed antibacterial activity. Distributional analysis revealed that peptide fractions exhibiting high anti-Vibrio activity are characterized by enrichment of high-ranking peptides rather than by AMP abundance alone. By structuring AMP identification and prioritization as sequential stages, the MAP + AIPx framework enables interpretable and experimentally actionable candidate selection by reducing biologically implausible candidates. The framework facilitates species-oriented prioritization of AMP candidates, addressing a key challenge in antimicrobial peptide discovery where activity may depend on target-specific membrane characteristics. Moreover, the approach is extensible through species-specific calibration and supports interpretable, mechanism-informed prioritization in antimicrobial peptide discovery.

Proteomics↗

PubMed2026

Systematic evaluation of one-dimensional-to-two-dimensional near-infrared spectroscopy transformations with deep learning for quantifying coconut sap adulteration.

Near-infrared (NIR) spectroscopy have limitations when combined with deep learning (DL) algorithms because they rely on low-dimensional datasets. Therefore, we investigated the potential of transforming one-dimensional (1D) NIR spectra into two-dimensional (2D) spectrograms using synchronous and asynchronous techniques and the continuous wavelet transform (CWT) and their effectiveness by integrating with DL for detecting adulteration in coconut sap. NIR spectra (12,500-4000 cm-1) were collected from binary mixtures (0%-100%;w/w). The performance of all DL (convolutional neural networks-CNN, AlexNet and ResNet) models was compared with that of partial least squares (PLS). The models were ranked in the mentioned order based on their performances: 2D-CWT > 2D-asynchronous > 2D-synchronous > 1D/2D-PLS. The important features of the best model can be explained and visualized using gradient-weighted-class-activation-mapping. The findings highlight that the 1D-to-2D NIR data transformation combined with DL is a highly robust approach because it addresses the feature representation gap in NIR data and effectively captures the spatial-spectral correlations.

Spectroscopy, Near-Infrared↗

PubMed2026

AI echo INSIGHT study: A prospective blinded randomized trial of artificial intelligence echocardiogram interpretation.

BACKGROUND: Transthoracic echocardiography (TTE) is the most commonly performed cardiac imaging modality with over 30 million studies annually. Demand for timely expert interpretation continues to outpace capacity, creating diagnostic delays and inter-observer variability that impact patient care. Recent research has suggested computer vision artificial intelligence (AI) models can generate accurate preliminary comprehensive TTE reports, however, prospective evaluation is needed to determine whether AI-assisted TTE interpretation can improve clinician efficiency while preserving diagnostic accuracy. METHODS: AI ECHO INSIGHT is a prospective randomized blinded clinical trial conducted at Kaiser Permanente Northern California that will evaluate 1200 historical TTE studies (1000 consecutive unselected studies plus 200 with moderate or greater valvular disease) interpreted using three workflows: (1) AI-generated preliminary report finalized by a blinded cardiologist (AI-assisted); (2) cardiologist-generated preliminary report finalized by a blinded cardiologist (cardiologist-assisted); and (3) sonographer-generated preliminary report finalized by a blinded cardiologist (sonographer-assisted). The primary outcome is the rate of substantial change between preliminary and final reports, comparing the AI-assisted workflow to the pooled cardiologist-assisted and sonographer-assisted workflows. Secondary outcomes include cardiologist interpretation time for report finalization, superiority testing for diagnostic accuracy, and reporting consistency. CONCLUSION: AI ECHO INSIGHT is a prospective randomized blinded clinical trial evaluating the clinical impact of AI-assisted TTE interpretation on diagnostic accuracy, cardiologist efficiency, and reporting consistency in real-world echocardiography workflows. TRIAL REGISTRATION: ClinicalTrials.gov registration number NCT07229300.

Humans↗

PubMed2026

Machine learning-based prediction of unplanned readmission and construction of an online calculator for elderly patients with mild ischemic stroke.

OBJECTIVE: To screen for independent risk factors for unplanned readmission in elderly patients with mild ischemic stroke, and to construct and validate an online risk prediction calculator based on an interpretable machine learning model, thereby providing a promising practical tool for accurate clinical assessment of 30‑day all‑cause unplanned readmission risk in this population. METHODS: A prospective cohort study was conducted, including 1050 patients aged ≥ 60 years with mild ischemic stroke admitted between August 2023 and September 2024. Participants were randomly divided into a training set (840 cases) and a test set (210 cases) at a ratio of 8:2. Risk factors were screened by univariate analysis and multivariable Logistic regression. Four machine learning models, namely LightGBM, XGBoost, Random Forest, and K‑Nearest Neighbors (KNN), were developed and their performance was evaluated using AUC, accuracy, sensitivity, and specificity as metrics. The SHAP framework was used for interpretability analysis, and an online calculator was subsequently developed based on the optimal model. RESULTS: Univariate analysis showed significant differences (P < 0.05) in 13 factors including age, smoking, AIP, TyG index, HALP score, etc. Multivariable Logistic regression identified age (OR = 9.752), smoking (OR = 5.171), AIP (OR = 6.691), TyG index (OR = 4.393), HALP score (OR = 2.831), and ≥ 2 comorbidities (OR = 3.664) as independent risk factors. All four machine learning models demonstrated good predictive performance. Based on a comprehensive evaluation of multiple metrics and computational efficiency, the LightGBM model exhibited the best predictive performance (AUC = 0.884, accuracy = 0.829, sensitivity = 0.812, specificity = 0.875). SHAP analysis showed that age, AIP, TyG index, smoking, and HALP score were key predictors. An online calculator developed based on this model enables individualized risk predictions. CONCLUSION: Key risk factors associated with 30‑day unplanned readmission in elderly patients with mild ischemic stroke were identified. The LightGBM model demonstrated high predictive accuracy, and together with the interpretability analysis and online calculator, offers a practical tool to support clinical risk assessment. However, this tool requires future external validation.

Humans↗

PubMed2026

Predicting ACL injury risk in athletes: A systematic review of machine learning-based models.

BACKGROUND: Early ACL injury risk identification in athletes is essential. This systematic review examines machine learning (ML) models for predicting ACL injuries, evaluating their methodological quality, performance, and reliability. METHOD: A comprehensive electronic search was conducted across PubMed, Scopus, Web of Science, and IEEE Xplore databases, supplemented by Google Scholar for grey literature, covering articles published between January 1, 2015, and August 30, 2025. Eligible studies were appraised using the Prediction Model Study Risk of Bias Assessment Tool (PROBAST) for methodological quality and risk of bias, and the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) guidelines for quality of evidence. RESULTS: Ten studies were included. PROBAST showed eight studies had moderate risk of bias and two low risk. TRIPOD found only two studies met quality criteria. ML models included logistic regression (n = 5), support vector machines (n = 4), k-nearest neighbor (n = 3), decision trees (n = 3), random forests (n = 5), neural networks (n = 2), linear discriminant analysis (n = 1), and pre-trained CNNs (n = 1). AUC ranged from 0.63 to 0.98. Accuracy (reported in six studies) ranged from 26% to 95%; however, these values should be interpreted with caution due to the absence of confidence intervals, lack of class imbalance handling, and limited external validation across studies. Tree-based ensemble methods such as random forest achieved competitive accuracy (74-86%), while SVM, a non-ensemble classifier, reported accuracy ranging from 71% to 95%; however, the highest values were obtained in studies with notably small sample sizes (n = 12 to n = 39), raising concerns about overfitting and generalizability. CONCLUSION: Current ML algorithms show promise for identifying athletes at high ACL injury risk and detecting relevant risk factors. Although study quality was generally satisfactory, future research should prioritize external validation and model interpretability to support clinical translation.

Humans↗

PubMed2026

AI-enabled viral genomics: from virus discovery to host prediction and emerging variant forecasting.

The rapid expansion of metagenomic sequencing has generated vast repositories of viral sequence data that far outpace our capacity to interpret them using conventional approaches. Highly divergent sequences, sparse functional annotation, and taxonomically uneven sampling present fundamental challenges for reference-dependent methods, which lose sensitivity precisely for novel and understudied viruses with high public health relevance. Artificial intelligence (AI) provides a new avenue to address these challenges by enabling predictive inference from viral genomes and proteins while reducing dependence on sequence similarity. In this Review, we discuss representative advances in AI for virus discovery, taxonomic classification and functional annotation, prediction of host range and zoonotic potential, and efforts toward forecasting emerging variants. These advances are transforming viral genomics from a largely descriptive discipline into one with increasing predictive capability. We also critically assess the major challenges that constrain current approaches, including the availability of high-quality and representative datasets, rigorous model evaluation, biological interpretability and responsible governance for increasingly capable AI models.

Artificial Intelligence↗

PubMed2026

The landscape of pruning for large language models: A systematic review and unified taxonomy.

Confronting the inherent tension between the exceptional capabilities and the immense computational costs of Large Language Models (LLMs), pruning has become a crucial technique for achieving efficient deployment. However, a systematic analytical framework dedicated specifically to LLM pruning remains absent. In this paper, we aim to bridge this gap. We first elucidate the theoretical foundations that underpin the effectiveness of pruning, namely overparameterization and redundancy, and then propose a multidimensional taxonomy that organizes existing approaches along the axes of granularity, timing, and criteria. Building upon this unified perspective, we further analyze performance recovery mechanisms and the broader evaluation ecosystem, while also exploring forward-looking challenges such as interpretability, automation, and hardware-algorithm co-design. Through this comprehensive synthesis, we seek to provide an integrated and coherent analytical lens for advancing both research and practice in LLM pruning.

Large Language Models↗

PubMed2026

Making waves: toward systems-level interpretation of hormonal and endogenous biomarkers in wastewater-based epidemiology.

Wastewater-based epidemiology (WBE) has proven invaluable for population health monitoring, most notably during the COVID-19 pandemic. Yet current WBE largely relies on exogenous markers such as drugs, pathogens, and their metabolites, limiting surveillance to what communities are exposed to. We argue for expanding WBE towards endogenous biomarkers, particularly hormones, which provide insights into physiological stress, metabolic function, and endocrine activity. Hormone-based WBE offers new opportunities to capture population-level biological responses to societal and environmental stressors, disasters, and chronic disease burdens at the community scale. This perspective outlines a systems-level framework for integrating hormonal signals in wastewater with clinical data, behavioral indicators, environmental factors, and digital markers to support more robust and context-aware public health surveillance. We highlight key technical considerations, interpretive challenges, and opportunities for translational pilot studies. By moving beyond exposure tracking toward more integrated interpretation of biological responses, hormone-informed WBE may contribute to more resilient, inclusive, and actionable public health infrastructure.

Humans↗

PubMed2026

Innovation-related perception as a key driver of alternative protein acceptance: evidence from an early-stage model for cultivated meat and algae-/microalgae-based alternative protein products in Italy.

Alternative proteins are increasingly considered part of the transition toward more sustainable food systems, yet their diffusion depends critically on consumer acceptance. This study investigates the early-stage acceptance of two alternative protein categories in Italy-cultivated meat and algae-/microalgae-based alternative protein products. Focusing on the first three phases of acceptance, the analysis examines how innovation-related perception (IRP) shapes consumer perceived value (CPV), consumer perceived risk (CPR), and subsequent affective (AFF), cognitive (COG), and conative (CON) responses. Data were collected through an online survey administered to 238 Italian respondents and analysed using partial least squares structural equation modelling (PLS-SEM). The results show that IRP is the main upstream driver of early-stage acceptance in both product domains: more favourable perceptions strongly increase perceived value and reduce perceived risk. In turn, CPV exerts a much stronger influence than CPR on both affective and cognitive attitudes. A tentative cross-model comparison suggests only a descriptive variation in the final transition toward conative acceptance: affective and cognitive responses were both significant in the two models, with a relatively larger affective coefficient for cultivated meat and more balanced coefficients for algae-/microalgae-based products. Overall, the findings support a process-based interpretation of alternative protein acceptance and highlight the central role of innovation-related perception in shaping early consumer responses. These results provide relevant implications for communication strategies, product positioning, and policy actions aimed at improving the acceptability of alternative proteins in food cultures characterised by strong culinary traditions.

Italy↗

PubMed2026

ORBIT: Oncogenic Representation Learning via Bi-Prototype Contrastive Learning in Hyperbolic Space for cancer driver gene identification.

Accurate identification of cancer driver genes is crucial for precision oncology but remains challenging due to the complexity of integrating heterogeneous data and modeling dynamic biological systems. To address these limitations, we propose ORBIT (Oncogenic Representation Learning via Bi-Prototype Contrastive Learning in Hyperbolic Space). Our framework synergistically fuses multi-omics profiles with functional network data using a context-adaptive graph reweighting mechanism to capture cancer-specific dynamics. The model employs a bi-prototype contrastive learning strategy within hyperbolic space, which aligns gene representations around distinct driver and non-driver semantic anchors while preserving the intrinsic hierarchy of biological networks. Comprehensive evaluations demonstrate that ORBIT achieves highly competitive stability in pan-cancer analysis while consistently outperforming state-of-the-art methods in cancer-specific predictions. Furthermore, functional enrichment analysis confirms that the model effectively segregates core cancer pathways, and drug sensitivity profiling validates the clinical relevance of the identified drivers. By integrating hyperbolic geometry with context-adaptive learning, ORBIT offers a robust and interpretable paradigm for precision medicine. The source codes and datasets are publicly accessible at https://github.com/spcho-dev/ORBIT.

Humans↗

PubMed2026

Beyond species trees: pervasive gene flow limits phylogenomic resolution in the diversification of Juniperus from the Qinghai-Tibet Plateau.

Understanding how lineages diversify despite persistent ancestral polymorphism and recurrent gene flow remains a central challenge in evolutionary biology. Juniperus distributed across the Qinghai-Tibet Plateau provide an ideal system for addressing this question because repeated geological uplift and climatic oscillations have likely promoted cycles of lineage divergence, range shifts, and secondary contact. Here, we combined approximately 1.08 million genome-wide SNPs from 164 individuals representing thirteen Juniperus lineages with phylogenomic datasets comprising 3,381 nuclear single-copy genes and nearly complete plastomes. We detected extensive phylogenomic discordance and cytonuclear incongruence across genomic datasets. Topology weighting, coalescent simulations, quartet-based tests, and analyses of gene flow and reticulation collectively support the interpretation that these patterns were shaped by the combined effects of prolonged incomplete lineage sorting and gene flow during lineage diversification. Ecological niche analyses further provide a spatial and climatic context in which environmentally similar lineages may have had greater opportunities for secondary contact during historical range shifts. Collectively, our results reveal that the evolutionary history of Qinghai-Tibet Plateau Juniperus is characterized by reticulate diversification rather than strictly bifurcating evolution, and demonstrate how genome-wide discordance can provide biological insights into the evolutionary processes underlying lineage diversification.

Gene Flow↗

PubMed2026

Artificial intelligence for anticancer drug discovery from natural products of macroalgae and sponges: A systematic review.

Marine natural products (MNPs) from macroalgae and marine sponges have inspired clinically important anticancer agents, including the cytarabine pharmacophore and the eribulin scaffold, while cyanobacterial dolastatin chemistry supplies the auristatin payloads of several marine-inspired antibody-drug conjugates (ADCs) such as brentuximab vedotin. Artificial intelligence (AI) methods, encompassing both classical machine learning (ML) with hand-engineered features and modern deep learning (DL) with many-layered neural networks, are increasingly supporting key decisions in natural-product anticancer drug discovery, including bioactivity prediction, target identification, absorption, distribution, metabolism, excretion and toxicity (ADMET) filtering, generative analogue design, and the selection of preclinical candidates. DL architectures relevant to this field include graph neural networks, transformer-based molecular generators, diffusion models for protein-ligand docking, and convolutional networks for mass spectrometry, while classical ML contributes interpretable fingerprint-based bioactivity models and molecular networking for dereplication. This review follows a systematic literature review methodology to organize the landscape of AI methods now applied to MNP anticancer discovery, distinguishing ML and DL approaches where relevant, situating them within the chemical context of macroalgal and sponge-derived oncology leads, and critically examining published case studies, including validation level (computational, in vitro, in vivo, clinical). The principal bottleneck for medical translation has shifted partly from algorithmic capability toward data infrastructure and experimental validation. Sparse, heterogeneous, and taxonomically biased bioactivity records limit what current models can learn and reduce the reliability of AI-prioritized candidates entering the preclinical pipeline. A roadmap is proposed that prioritizes open MNP-specific benchmarks, symbiont-aware modeling, and active learning loops with synthesizability and ADMET constraints. These AI workflows may accelerate the prioritization of marine-derived anticancer leads and support earlier, more evidence-based translational decisions in oncology drug development.

Biological Products↗

PubMed2026

Analysis of deep learning techniques in computer-aided diagnosis for meniscus injuries: a systematic literature review.

Meniscus informatics is a growing subject of study in the healthcare industry. One of the major hindrances to the healthcare system's transformation is obtaining knowledge and meaningful information from complicated, high-dimensional and diverse sources. Modern biomedical research, for instance, has seen an increase in the use of complex, dissimilar, poorly documented, and generally unstructured electronic health records, imaging, sensor data and text, even after many current techniques have been used to extract more robust and useful elements from the data for analysis. New efficient standards for building end-to-end learning models from complex data are therefore needed. Therefore, the current study aims to examine the most recent research on the use of deep learning techniques for diagnosing meniscus tears and recommend creating comprehensive and meaningful interpretable structures that might benefit the healthcare industry. We also draw attention to shortcomings and the need for better technique development, and we provide new perspectives about this exciting new development in the field.

Humans↗

PubMed2026

Quo vadis, BGA? A collaborative EDNAP exercise on the challenges and progress in forensic biogeographical ancestry inference.

There is a broad consensus that forensic tests for the prediction of externally visible characteristics (EVC) and analysis of biogeographic ancestry (BGA) of an individual are technically reliable. However, interpretation of the results and population-specific genotype distribution patterns remains challenging. EVC and BGA analyses provide valuable information for population genetics studies and as investigative leads for criminal cases, as well as for historical and contemporary identification tests. However, inaccurate or incorrect predictions, for example, from subjective bias in the interpretations made, have the potential to misdirect police investigations. The legal situation regarding EVC and BGA testing varies by country: ranging from countries where it is explicitly prohibited, to those without specific regulations on biogeographic ancestry prediction, and others that have already enacted laws governing its use. The reluctance to utilize these analyses is not only due to legal restrictions and data protection concerns, but also to initial limited sets of sufficiently comprehensive forensic DNA assays. Forensic BGA marker panels typically contain up to ∼300 SNPs. This relatively small number of genetic markers, along with limited reference population data, complicates the interpretation of results from donors of unknown origin. This paper presents the results of a collaborative EDNAP study, which, for the first time, evaluated the approach to reporting EVC and BGA data between international laboratories. For the study, DNA from nine individuals with self-reported ancestry was collected and analysed using various forensic panels differing in the number and composition of ancestry-informative markers genotyped, comprising: the Precision ID mtDNA Whole Genome Panel, the VISAGE Basic Tool and the VISAGE Enhanced Tool for Appearance and Ancestry Prediction, and the Ion AmpliSeq™ PhenoTrivium Panel. To ensure full data protection, all SNP genotypes and uniparental marker haplotypes obtained were not shared with third parties. Instead, the genetic data were analysed using a range of commonly used population analysis software packages. These analysis outcomes were then distributed to twelve European forensic laboratories (both academic and law enforcement institutions), who were asked to prepare reports based on their interpretation of the phenotypes and ancestry they inferred from the analysis data. A questionnaire sent alongside the genetic information, aimed to evaluate which difficulties were encountered by the participants in processing the BGA analysis data they were given.

Humans↗

PubMed2026

Association of time-averaged systemic immune-inflammation indices with in-hospital mortality after intracerebral hemorrhage: a retrospective study.

BACKGROUND: Systemic inflammation plays a central role in secondary brain injury following intracerebral hemorrhage (ICH). Although inflammatory indices such as the neutrophil-to-lymphocyte ratio (NLR), systemic immune-inflammation index (SII), and systemic inflammation response index (SIRI) are linked to poor outcomes, their associations with mortality are commonly assumed to be linear, potentially overlooking nonlinear patterns where mortality risk rises steeply at higher levels. METHODS: We conducted a retrospective study using the MIMIC-IV database, including 440 patients with non-traumatic ICH who were alive and remained in the ICU for at least 72 h after admission. Mean NLR, SII, and SIRI were calculated from measurements obtained during this period. Multivariable logistic regression and restricted cubic spline (RCS) analyses were applied to assess their independent and nonlinear associations with in-hospital mortality. Model discrimination and calibration were internally validated using 1,000 bootstrap resamples. RESULTS: The in-hospital mortality rate was 26.1%. After multivariable adjustment, NLR and SIRI remained independently associated with mortality. Patients in the highest SIRI quartile had the highest risk of death (aOR = 5.12; 95% CI: 2.57-12.24; p < 0.001). RCS analysis revealed a significant nonlinear association between SIRI and mortality (p-nonlinearity < 0.05), showing a steep risk increase at higher SIRI levels. Adding SIRI to the base model provided a modest improvement in discrimination (AUC 0.762 to 0.785, p = 0.045) and significantly improved risk reclassification (cNRI = 0.4778, p < 0.001; IDI = 0.0240, p = 0.0151). CONCLUSIONS: Among patients with ICH who met the 72-hour eligibility criterion, higher 72-hour average SIRI was independently associated with in-hospital mortality. As a time-averaged measure, SIRI should be interpreted as a dynamic marker integrating the initial inflammatory state and the early clinical course rather than as a purely baseline prognostic factor. Although adding SIRI to the base model modestly improved discrimination and risk reclassification, it should be considered a candidate prognostic marker requiring external validation before clinical application.

Humans↗

PubMed2026

Approaches to observational study designs and analytical options to evaluate the safety of multi-dose vaccines: a systematic review.

INTRODUCTION: Observational studies require careful considerations when evaluating the safety of multidose vaccines. We reviewed design and analytical approaches in observational studies evaluating the safety of multidose vaccines in the post-licensure phase. METHODS: EMBASE, MEDLINE, Web of Science, and Scopus (2018-2022) were searched for hypothesis-testing studies evaluating the safety of multidose vaccines. Key features from frequently used designs were extracted. RESULTS: Among 123 eligible studies, cohort (46%) and self-controlled case series (SCCS)/self-controlled risk interval (SCRI) (40%) followed by case-control (12%) were the most common designs, and 15% of studies used multiple designs. Among cohort studies evaluating multiple doses, vaccination date (36%) and cohort entry with time-updated exposure status (32%) were frequent approaches used to define time zero. Twenty-eight percent of cohort studies did not report time zero; all but one evaluated COVID-19 vaccine effect on post-delivery and fertility-related outcomes. For SCCS/SCRI, 64% of studies accounted for event-dependent exposures, mainly by including pre-exposure periods (53%) and modified SCCS model (48%), while 20% employed multiple correction strategies. Among studies using multiple designs, 68% reached consistent conclusions. CONCLUSIONS: SCCS/SCRI and cohort designs dominate multidose vaccine safety studies. Clear reporting on time zero in pregnancy and fertility-related cohort studies, and on addressing event-dependent exposures in SCCS/SCRI studies is needed, along with guidance on interpreting results from multiple designs.

Humans↗

PubMed2026

The relationship between composite dietary antioxidant index and odds of depression: A systematic review and dose-response meta-analysis of observational studies.

BACKGROUND: Recently, the association between the composite dietary antioxidant index (CDAI) score and depression has received special attention from researchers. However, this possible association has not been comprehensively examined. Thus, the present study aims to comprehensively assess the relationship between CDAI score and depression odds. METHODS: A thorough literature search was performed using Scopus, PubMed, and Web of Science up to November 2025. Observational studies assessing the relationship between CDAI score and depression odds were included. The odds ratios (ORs) and their 95% confidence intervals (CIs) reported in each study were extracted, and a pooled effect size was calculated using a random-effects model with inverse-variance weighting. RESULTS: Four observational studies were included in the current systematic review and meta-analysis. The pooled effect size indicated that a higher CDAI score was associated with lower odds of depression (OR:0.62; 95%CI: 0.51-0.76, I2 = 48.5%; P-value = 0.121). In an exploratory linear dose-response analysis based on two eligible studies, no significant association was observed between a one-unit increase in CDAI score and the odds of depression (OR = 0.96; 95% CI: 0.91-1.02; I2 = 92.2%; P-value<0.001). Given the limited number of studies and substantial heterogeneity, these findings should be interpreted cautiously. CONCLUSIONS: Our findings suggest that higher CDAI scores, reflecting greater adherence to antioxidant-rich dietary patterns, are associated with lower odds of depression.

Humans↗