Search Results (17,705)

Rift Valley fever (RVF) is a vector-borne zoonotic disease whose occurrence is strongly associated with climatic and environmental conditions, making data-driven approaches potentially valuable for epidemiological surveillance and risk assessment. Using a publicly available historical dataset comprising 180,288 monthly observations from geographically defined administrative units across Kenya between 1981 and 2010, this study investigates machine learning (ML) for the retrospective classification of reported RVF occurrence from contemporaneous climatic, environmental, topographic, and seasonal predictors under an extremely imbalanced classification setting. The dataset provides broad geographic coverage across Kenya over a 30-year historical period; however, because the outcome reflects reported events in historical surveillance records, it is not assumed to constitute a formally population-representative national sample or to capture all underlying RVF transmission. Each observation represents a geographic unit and observation month, and the response indicates whether an RVF event was reported during that corresponding period. Therefore, the present analysis should be interpreted as contemporaneous outbreak classification rather than as a fixed-horizon prospective forecast. Thirteen classifiers representing distinct learning paradigms were systematically evaluated: Logistic Regression, Linear Discriminant Analysis, K-Nearest Neighbors, Classification and Regression Tree, Naive Bayes, Support Vector Machine, Weighted Logistic Regression, XGBoost, LightGBM, CatBoost, Balanced Random Forest, EasyEnsemble, and RUSBoost. Model performance was assessed before and after SMOTENC-based rebalancing using overall and class-specific metrics, including accuracy, precision, sensitivity, specificity, F1-score, ROC–AUC, and precision–recall-based measures. Under the retrospective stratified hold-out benchmark, XGBoost, CatBoost, Balanced Random Forest, and LightGBM achieved ROC–AUC values of 0.9176, 0.9175, 0.9114, and 0.9062, respectively. Balanced Random Forest attained the highest outbreak sensitivity (0.8851), although at the cost of very low precision, illustrating that high rare-event detection can generate a substantial false-alert burden in surveillance settings. SMOTENC produced strongly model-dependent effects: it increased outbreak sensitivity for XGBoost, LightGBM, CatBoost, KNN, CART, and RUSBoost, but substantially reduced sensitivity for Balanced Random Forest and EasyEnsemble. SHAP-based interpretability analysis indicated that month, rainfall, and slope were among the most influential predictors and further showed that class rebalancing can alter the distribution of feature contributions. Overall, the findings demonstrate that modeling reported RVF occurrence under severe class imbalance requires joint evaluation of minority-class detection, false-positive behavior, discrimination, and model interpretability rather than overall accuracy alone. The present results establish a retrospective classification benchmark for climate-informed RVF risk assessment, but they should not be interpreted as an autonomous outbreak-warning system. Translation into prospective early-warning prediction will require an explicit forecasting horizon, predictors constructed exclusively from information available before the target period, temporally and geographically independent validation, and decision thresholds evaluated against an operationally acceptable false-alert burden. Full article

Arctic water–energy–food (WEF) infrastructure presents material and logistical conditions that may be unfamiliar to audiences in southern Canada. This exploratory mixed-methods study examined how 28 southern Canadian adults explored and discussed these conditions during 12 facilitated sessions of NexusQuest, a 3D serious game representing Arctic WEF systems in simplified form. The study combined qualitative analysis of discussions during gameplay, post-play interviews, and written responses with pre/post self-ratings of understanding. Participants described surprise at above-ground utilities and remote supply costs, compared piped and trucked services, and questioned the model’s realism, omissions, and potential bias. Model questioning occurred during both gameplay and facilitated reflection, while some participants identified facilitator explanations as a principal source of information. Twenty participants gave higher post-session self-ratings, eight gave unchanged ratings, and none gave lower ratings, although the pre- and post-session questions differed in wording. We interpret participants’ surprise at unfamiliar infrastructure conditions and questioning of model assumptions through “friction as pedagogy,” a sensitizing concept developed during analysis. The findings show how infrastructure arrangements and resource displays served as shared reference points for discussing Arctic WEF relationships and examining the limits of their representation within an experience combining gameplay, explanation, and reflection. Full article

Background: Immunotherapy has transformed cancer treatment and is widely used in the Chinese mainland. Though advances have been made, immune checkpoint inhibitor-related liver injury (ICILI) remains a significant clinical challenge. Existing risk models commonly lack time-specific risk stratification for ICILI. The present study aimed to develop and validate interpretable machine learning models to predict grade 2 or higher ICILI at multiple time points in patients with gastrointestinal cancer (GC). Methods: This retrospective cohort study encompassed GC patients who commenced their initial ICI medication between January 2019 and June 2023 at Zhongshan Hospital, Fudan University. Five machine learning algorithms, including Logistic Regression, Random Forest, Extreme Gradient Boosting (XGBoost), Gradient Boosting (GradientBoost), and Adaptive Boosting (AdaBoost), were utilized to develop predictive models for grade ≥ 2 ICILI at specific intervals of 3 months, 6 months, and 12 months. The evaluation of model performance was conducted using the area under the curve (AUC), accuracy, precision, recall, and F1-score. The Shapley Additive exPlanations (SHAP) method was employed to assess feature importance and interpret the final model. Results: A total of 1337, 849, and 401 patients were enrolled in the follow-up groups at 3 months, 6 months, and 12 months. The final model for grade ≥ 2 ICILI was developed with GradientBoost and achieved an AUC of 0.769 (95% CI: 0.732–0.806), with a test set accuracy of 0.834. XGBoost yielded AUCs of 0.671 (95% CI: 0.636–0.706) at 3-month indication, 0.678 (95% CI: 0.638–0.718) at 6 months, and 0.644 (95% CI: 0.589–0.699) at 12-month follow-up in the 5-fold cross-validation. The DCA curve demonstrated solid clinical benefit, whereas the calibration curve indicated good predictive reliability. SHAP analysis identified several parameters as predictive features at different intervals, which suggested that the ICILI determinants varied from acute inflammatory to host-related characteristics. Conclusions: A temporal stratification prediction model for grade ≥ 2 ICILI in GC patients was developed and validated at various intervals using ML algorithms with SHAP interpretability. This methodology facilitated early recognition of varying parameters across different treatment phases, enhancing clinical management and elevating treatment outcomes. Full article

Background/Objectives: Nonsteroidal anti-inflammatory drugs (NSAIDs) are widely used but can cause clinically significant gastrointestinal adverse events (GI-AEs). However, genetic susceptibility to NSAID-related GI toxicity remains insufficiently characterized in East Asian populations. This exploratory study aimed to identify genetic loci associated with grade ≥ 2 NSAID-related GI-AEs in Korean patients treated with commonly used NSAIDs. Methods: We conducted an exploratory genome-wide association study in a multicenter Korean cohort of NSAID-treated patients recruited from four Korean hospitals. A total of 339 patients were included, comprising 28 cases with grade ≥ 2 GI-AEs and 311 controls without documented GI-AEs. Genotyping was performed using the Korea Biobank Array, followed by imputation and quality control. Genome-wide association analysis was conducted under an additive logistic regression model adjusted for age, sex, and the first two principal components, with additional sensitivity analyses accounting for the specific NSAID type, cumulative DDD, and concomitant medications. Polygenic risk score analysis was performed using summary statistics from an arthritis-treated subset of the KoGES Ansan/Ansung cohort, and the selected variant were further evaluated using multivariable Firth penalized logistic regression. The discriminatory performance of clinical, laboratory, and genetic models was assessed using bootstrap internal validation. Results: The exploratory GWAS identified 20 suggestive variants associated with NSAID-related GI-AEs at p < 1 × 10⁻⁵, although no variant reached genome-wide significance. The KoGES-derived polygenic risk score was significantly associated with GI-AE status in the NSAID cohort, and 10 suggestive variants overlapped with SNPs included in the best-fit PRS model. rs1925245 maintained a consistent association in multivariable Firth penalized logistic regression. Conclusions: These findings suggest that NSAID-related GI-AEs in Korean patients may reflect a multifactorial susceptibility pattern involving both clinical context and polygenic risk. Although exploratory, this study provides candidate genetic signals for future replication and supports the need for larger ancestry-matched pharmacogenomic studies of NSAID-related GI toxicity. Full article

Background/Objectives: To assess associations between quantitative computed tomography (CT) features, inflammatory markers, and length of hospital stay (LOS) in acute pyelonephritis (APN). Methods: This retrospective single-center study included 82 patients with CT-confirmed APN. Two radiologists quantified renal perfusion-deficit volume and percentage and perirenal fat-stranding (PFS) area and attenuation. Associations with C-reactive protein (CRP), white blood cell (WBC) count, procalcitonin, and LOS were assessed using Spearman correlation and regression analyses. Receiver operating characteristic (ROC) curve analysis evaluated LOS ≥ 10 days. Results: Perfusion-deficit volume and percentage correlated strongly with CRP (ρ = 0.763 and 0.714; both p < 0.001) and weakly with WBC count (ρ = 0.379 and 0.374; both p < 0.001). PFS area correlated moderately with procalcitonin (ρ = 0.482, p = 0.001; n = 42). Both perfusion-deficit measures correlated moderately with LOS (ρ = 0.538 and 0.536; both p < 0.001). In adjusted linear regression, CRP remained associated with LOS (Coefficient = 0.523 days per 10 mg/L; 95% CI, 0.345–0.692; p < 0.001), whereas perfusion-deficit percentage did not (p = 0.430). In adjusted logistic regression, CRP, age, and male sex were associated with LOS ≥ 10 days. The multivariable model yielded an AUC of 0.883 (95% CI, 0.81–0.95). Results were unchanged in a sensitivity analysis excluding the seven outpatients, and bootstrap internal validation yielded an optimism-corrected AUC of 0.86. Conclusions: Quantitative renal perfusion deficits were associated with inflammatory burden and LOS in univariable analyses. CRP showed the most consistent independent association, whereas the incremental value of CT metrics requires external validation. Full article

Background and Objectives: Pediatric chest pain is usually benign, but identifying children who require admission or have a cardiac diagnosis remains challenging. We aimed to identify factors associated with these outcomes, with a focus on recurrent presentation and the combined value of C-reactive protein (CRP) and predefined red-flag findings. Materials and Methods: This retrospective cohort included children aged 0–18 years presenting with chest pain to a tertiary pediatric emergency department between January 2021 and January 2026; visits with insufficient documentation were excluded. Red-flag findings were predefined as effort-related chest pain, syncope, palpitations, dyspnea, fever, or a family history of sudden cardiac death. Independent predictors were assessed using multivariable Firth logistic regression, and cluster-robust standard errors were used in a sensitivity analysis to account for recurrent visits. Results: Of 1703 visits (1435 patients), 452 (26.5%) were recurrent presentations. Admission occurred in 64 visits (3.8%), and 91 (5.3%) had a cardiac diagnosis. Red-flag findings (OR, 5.15), abnormal ECG (OR, 4.00), CRP > 5 mg/L (OR, 3.71), male sex (OR, 2.13), and recurrent presentation (OR, 2.51) were independently associated with admission (all p ≤ 0.009). Recurrent presentation was not associated with a cardiac or psychogenic diagnosis. Admission and cardiac diagnosis rates increased from 1.4%/2.2% in children with neither elevated CRP nor red-flag findings to 22.2%/24.8% in those with both. Routine ECG, troponin, and chest radiography had low diagnostic yields (2.1–6.0%), whereas echocardiography performed selectively yielded abnormal findings in 79.7%. Conclusions: Red-flag findings, abnormal ECG, elevated CRP, and recurrence were associated with admission, whereas recurrence was not associated with diagnosis. Combining CRP with red flags may improve risk stratification, though prospective validation is needed. Full article

This study examines the Physical Execution Layer, in which AI agents extend digital decision-making into real-world tasks. It asks why Amazon Supply Chain Services (ASCS), launched in May 2026, may be structurally favored as infrastructure for this transition and evaluates it as an emerging Physical Operating System (Physical OS) candidate. Based on Amazon’s 10-K filings for 2013–2025 and an exploratory proxy series for ASCS revenue, the study applies logistic curve fitting and regression analysis with sensitivity analysis to compare the growth trajectories of Amazon Web Services (AWS) and ASCS from 2013 to 2026, including an exploratory 2026 reference year. AWS revenue is closely fitted in-sample by a logistic S-curve, whereas the ASCS proxy reflects a more physically constrained trajectory; their growth rates nevertheless exhibit substantial co-movement. The pattern is consistent with a proposed Dual-OS architecture in which AWS supplies compute, inference, and optimization while ASCS supports real-world execution within a shared technology and infrastructure investment base. CX physicalization and system-level supply-chain decarbonization may further increase the strategic importance of logistics performance. The findings are descriptive and correlational, not causal, and the ASCS series is a proxy. The Dual-OS architecture is presented as an analytical framework for future digital-physical execution networks and manufacturing reconfiguration. Full article

Introduction: Mothers of children with disabilities may experience substantial and prolonged psychophysical burden, which can negatively affect psychological well-being, health-related quality of life, and physical functioning. This study describes the protocol of a pilot and feasibility study evaluating the feasibility and acceptability of two distinct non-pharmacological interventions—craniosacral therapy (CST) and respiratory training (RT)—in mothers of children with disabilities. Methods and Analysis: This completed, retrospectively registered, three-arm, parallel-group, non-randomized controlled pilot and feasibility study enrolled 32 mothers, allocated by enrolment timing and project logistics to craniosacral therapy (CST; n = 12), respiratory training (RT; n = 10), or no-intervention control (CG; n = 10). The interventions lasted 8 weeks. Feasibility outcomes included recruitment, consent, retention, attendance/adherence, outcome completeness, fidelity indicators, withdrawals, and reported adverse events; acceptability was assessed indirectly because no dedicated acceptability instrument was administered. Clinical outcomes were exploratory. Analyses estimated pre–post change and between-group differences, with confidence intervals reported where supported by the analyses. Recruitment occurred from May to December 2025, follow-up ended on 30 April 2026, and outcome data were analyzed before submission. Ethics approval preceded enrolment (2-XII/2024); registration was retrospective (ANZCTR ACTRN12626000148370; 5 February 2026). Discussion: This pilot and feasibility study provides a structured framework for evaluating CST and RT in mothers of children with disabilities. The study is intended to inform methodological procedures, outcome selection, and the design of a future adequately powered randomized controlled trial rather than establish the clinical effectiveness or safety of either intervention. Full article

Bloodstream infections (BSIs) caused by carbapenemase-producing Enterobacterales (CPE) carry a high mortality, and comparative real-world data remain limited, particularly for metallo-β-lactamase (MBL) and dual-carbapenemase producers. We conducted a single-centre, six-year (2020–2025) retrospective study of 308 adults with CPE BSIs at a Romanian tertiary hospital, applying multivariable logistic regression, Cox proportional-hazards models with a 72 h landmark analysis, and inverse-probability-of-treatment-weighted (IPTW) regression to examine 30-day mortality and microbiological recurrence. Pre-specified sensitivity analyses addressed calendar period, carbapenemase class and incomplete outcome ascertainment. Klebsiella pneumoniae accounted for 92.5% of isolates, MBL determinants for 44.1% and dual carbapenemases for 31.8%. Thirty-day mortality was 72% among 275 patients with documented day-30 status. Vasopressor requirement, viral pneumonia and bacterial respiratory infection were independently associated with death, whereas definitive ceftazidime/avibactam with or without aztreonam (C/AVI ± AZT) was associated with lower mortality (adjusted OR 0.38; 95% CI 0.17–0.83). Mortality was 66.3% with C/AVI ± AZT, 79.3% with colistin and 67.1% with other regimens. After IPTW, C/AVI ± AZT was associated with lower 30-day mortality than colistin (OR 0.48; 95% CI 0.26–0.93); in the OXA-48/KPC subgroup, the estimate was unchanged but less precise (OR 0.49; 95% CI 0.20–1.19). The association was attenuated and no longer significant after adjustment for calendar period (OR 0.72; 95% CI 0.37–1.42), with epoch-stratified estimates of 1.14, 0.75 and 0.44 for 2020–2021, 2022–2023 and 2024–2025, respectively; crude mortality under colistin was stable across epochs (81.1%, 76.6%, and 81.2%, respectively), whereas mortality under C/AVI ± AZT fell (83.3%, 72.2%, and 60.7%, respectively). Weighted estimates were near-identical in isolates with and without an MBL determinant (0.46 and 0.48, respectively). Recurrence (10.1%) was more frequent with single OXA-48 or MBL producers. C/AVI ± AZT was associated with lower 30-day mortality than colistin, but the association was concentrated in the period when the agent was routinely available and did not survive adjustment for calendar time. These findings describe an association, not a causal survival advantage. Full article

Background/Objectives: Human immunodeficiency virus (HIV) and tuberculosis (TB) remain major public health challenges in sub-Saharan Africa, which bears the highest burden of HIV/TB co-infection. Evidence is limited on whether the severity of virological failure is associated with workplace absenteeism among employed adults with HIV/PTB co-infection. This study examined workplace absenteeism across two viral-load categories among employed adults with documented virological failure in the Govan Mbeki sub-district. Methods: A retrospective record review with secondary data analysis was conducted among 286 employed mine workers with HIV/PTB co-infection and a viral load ≥1000 copies/mL after at least six months of antiretroviral therapy (ART). Clinical data were extracted from medical, laboratory, and TB-treatment records and linked to workplace sick-leave registers; limited participant contact was used only to clarify essential missing information. The analytic dataset had an observed minimum viral load of 1200 copies/mL. Viral load was categorised as 1200–10,000 versus ≥10,001 copies/mL. Absenteeism was the number of recorded sick-leave days during the preceding three months and, for logistic regression, was classified as low (5–10 days) versus high (11–15 days). Multivariable logistic regression was adjusted for age, sex, education, and structural/socioeconomic barrier score. Results: Mean absenteeism was 10.05 days (SD = 2.44). Of 286 participants, 181 (63.3%) had a viral load ≥10,001 copies/mL and 105 (36.7%) had a viral load of 1200–10,000 copies/mL. High absenteeism occurred in 129 participants (45.1%) and lower absenteeism in 157 (54.9%). Viral load category was not associated with high absenteeism after adjustment (aOR = 1.21, 95% CI: 0.76–1.92; p = 0.432). The model was not statistically significant (χ² = 6.41, df = 5, p = 0.269), explained 3.0% of the variance (Nagelkerke R² = 0.030), and had an acceptable Hosmer–Lemeshow fit (p = 0.913). Conclusions: Within this selected employed cohort with virological failure, high-level virological failure and absenteeism were both common, but no evidence of an independent association between viral-load category and high absenteeism was detected. The cross-sectional temporal structure, healthy-worker selection, limited exposure contrast, and residual confounding constrain causal interpretation. Integrated public health and workplace interventions are suggested to improve both treatment outcomes and workforce participation in high-burden settings. Full article

Background/Objectives: To evaluate the association between the newly defined cerebroplacental–renal ratio (CPRR) and adverse perinatal outcome (APO) in pregnancy-induced hypertension (PIH), and to compare its discriminative performance with the cerebroplacental ratio (CPR) and cerebroplacental–uterine ratio (CPUR). Methods: This prospective, single-center cohort study included 105 women with PIH (preeclampsia, n = 63; gestational hypertension, n = 42). Doppler assessment was performed at ≥34 + 0 weeks and within 7 days before delivery. CPR was calculated as middle cerebral artery pulsatility index (PI)/umbilical artery PI, CPUR as CPR/mean uterine artery PI, and CPRR as CPR/right renal artery PI. APO was defined by the presence of any of the following: non-reassuring fetal status, a 5 min Apgar score <7, neonatal intensive care unit admission, or neonatal death. Receiver operating characteristic analysis and multivariable logistic regression adjusted for gestational age at ultrasound assessment and diagnosis group were performed. Results: APO occurred in 37 pregnancies (35.2%). CPUR and CPRR were significantly lower in the APO group, whereas CPR did not differ significantly. CPRR showed a higher apparent AUC for APO than CPR and CPUR (AUC 0.786, 95% CI 0.698–0.875), with significantly higher AUCs on pairwise DeLong comparisons. After adjustment, CPRR was the only Doppler index independently associated with APO (adjusted OR 1.423 per 0.1-unit decrease; 95% CI 1.162–1.743; p = 0.001). Conclusions: CPRR was independently associated with APO in PIH and showed higher discriminative performance than CPR and CPUR; however, these findings require validation in independent, multicenter studies before clinical application. Full article

Background: Effective hypertension management depends partly on medication adherence. Mental health disorders, such as anxiety and depression, may influence adherence, but evidence in older populations is limited. This secondary analysis of a cohort study identifies adherence trajectories to antihypertensive medications among older adults and assesses whether anxiety and depression are associated with trajectory membership. Patients and Methods: A cohort of 986 older adults (aged 65 and above) using antihypertensive treatments was analyzed. Adherence was measured using prescription claims over a 12-month period. Adherence patterns over time were characterized using Group-Based Trajectory Modeling (GBTM). Self-reported symptoms and diagnostic codes for anxiety and depression were used to assess for mental health disorders. Associations between depression or anxiety and adherence trajectories were investigated using logistic regression models adjusting for potential confounders. Results: We identified two stable adherence trajectories: a high-adherence group (83.6%) and a low-adherence group (16.4%). No evidence of an association was observed between the presence of anxiety (adjusted odds ratio (OR) of 1.2, 95% confidence interval (CI): 0.8–1.9) or depression (adjusted OR of 1.0, 95% CI: 0.6–1.6) and adherence trajectories. Conclusions: While most older adults in the study maintained high adherence to antihypertensive medications, a notable minority consistently demonstrated low adherence. These findings suggest that additional determinants of adherence trajectory, beyond mental health, should be investigated. Full article

Background: This study aimed to evaluate the performance of machine learning models in predicting depression risk among older adults living alone and to identify the features contributing to those predictions using explainable artificial intelligence (XAI). Methods: We analysed 2022 nationwide survey data in Korea. A total of 1007 older adults remained after excluding respondents who lived in multi-person households, were aged < 65 years, or had physician-diagnosed dementia. Depression risk was defined using the CES-D-10 (cutoff ≥ 10). After removing features with high multicollinearity, logistic LASSO selected 23 predictors. Six algorithms were fitted using the training set, with hyperparameter tuning performed by 5-fold cross-validation where applicable, and evaluated in a held-out test set following a 70/30 split. SMOTE was applied only to the training data. Performance was summarised using AUC, sensitivity, specificity and the F1 score with bootstrap 95% confidence intervals, and stability was assessed by repeated stratified cross-validation. SHAP values provided explainability. Results: LightGBM achieved an AUC of 0.802 (95% CI 0.747–0.852), followed by Random Forest (0.794) and Logistic Regression (0.779). These differences were small relative to the uncertainty of the estimates. SHAP analysis identified oral health-related quality of life, satisfaction with relationships with children, frequency of social contact, overall life satisfaction, satisfaction with health status, and age as the most influential features. IADL limitations, diabetes, hypertension, and perceived social class contributed to predictions with smaller effects. Conclusions: An explainable LightGBM model achieved an AUC of 0.802 for depression risk among older adults living alone and identified psychosocial and health-related features, particularly oral health and social connectedness, that may help inform future screening strategies. Full article

The COVID-19 pandemic redirected infection control and antimicrobial stewardship towards SARS-CoV-2, potentially increasing multidrug-resistant (MDR) bloodstream infections (BSIs). We compared MDR BSI epidemiology and outcomes in a non-COVID intensive care unit (ICU) before and after the pandemic. The MATURATE ICU study retrospectively included adults with MDR BSIs admitted to a non-COVID ICU during the pre-COVID-19 (1 January 2017–1 January 2020) and post-COVID-19 (1 January 2022–1 January 2025) eras. Clinical, microbiological and outcome data were compared. Survival analyses and multivariable logistic regression identified predictors of ICU mortality. Overall, 146 patients were included (74 pre-COVID-19 and 72 post-COVID-19). Pre-COVID-19 patients were older and had higher comorbidity burden and severity indices (all p ≤ 0.023). ICU stay after the first positive blood culture was longer in the post-COVID-19 period (p = 0.003). New organ dysfunction, acute kidney injury, and septic shock were more frequent pre-COVID-19, whereas cardiovascular events were more frequent post-COVID-19. ICU mortality was higher pre-COVID-19 (p < 0.001). Competing-risk analysis, treating ICU discharge alive as a competing event, demonstrated a significantly lower cumulative incidence of ICU death in the post-COVID-19 era (adjusted sHR 0.20, 95% CI 0.11–0.35; p < 0.001). In multivariable analysis, post-COVID-19 era independently predicted lower 28-day ICU mortality (adjusted OR = 0.11, 95% CI 0.04–0.28, p < 0.001), whereas higher CCI (adjusted OR = 1.31, 95% CI 1.04–1.69, p = 0.027) and SOFA (adjusted OR = 1.22, 95% CI 1.07–1.40, p = 0.003) independently predicted increased mortality. Despite the persistently high burden of antimicrobial resistance, improved post-COVID-19 survival suggests that, advances in ICU care and healthcare system recovery may have contributed to better outcomes. Full article

Background: Congenital heart disease (CHD) co-occurring with fetal growth restriction (FGR) is clinically important but incompletely characterized. We assessed the proportion and subtype spectrum of CHD in term-born children with FGR, and factors associated with its presence and complexity. Methods: Retrospective analysis of 375 term singleton children with documented FGR at a tertiary pediatric center. CHD required echocardiographic confirmation; isolated patent foramen ovale was excluded. Patients with multiple lesions were hierarchically classified as simple or complex. Multivariable logistic regression was restricted to characteristics available at or before birth. Results: CHD was confirmed in 96/375 children (25.6%): 71 (74.0%) simple, 25 (26.0%) complex. Atrial septal defect and patent ductus arteriosus were most frequent (each 45.8%), followed by patent foramen ovale (44.8%) and ventricular septal defect (32.3%). Severe FGR (aOR 2.012, 95% CI 1.222–3.313; p = 0.006) and genetic syndrome (aOR 5.969, 95% CI 2.565–13.889; p < 0.001) were independently associated with CHD presence. Within the CHD subgroup, genetic syndrome was also associated with complex disease (aOR 5.156, 95% CI 1.651–16.104; p = 0.005). Diagnosis occurred before six months in 95.8%. Hospitalization outcomes and comorbidity burden did not differ by complexity. Conclusions: FGR severity and genetic syndrome were both independently associated with the presence of a cardiac defect, and genetic syndrome was further associated with its complexity. These findings support targeted rather than universal echocardiographic assessment. Directionality cannot be established from postnatal data. Full article