Tag: statistics

Curr Med Imaging. 2026;22:e15734056495507. doi: 10.2174/0115734056495507260515193943.

ABSTRACT

INTRODUCTION: Soft tissue metastases are rare lesions seen in the course of systemic malignancies and pose diagnostic challenges. Radiological findings are heterogeneous, and standardized criteria are needed for reliable differentiation from benign lesions. While the diagnostic importance of perilesional changes is emphasized in the current literature, comprehensive analyses considering the clustered data structure are lacking. In this study, we aimed to determine effective radiological parameters for distinguishing soft tissue metastases from benign soft tissue lesions and to reveal the independent predictive value of perilesional findings.

METHODS: Soft tissue lesions detected by computed tomography (CT) and magnetic resonance imaging (MRI) between January 2015 and December 2023 were retrospectively evaluated in this single-center study. The study included 57 benign lesions (55 patients) and 139 metastatic lesions (65 patients). Lesion size, contour characteristics, morphological shape, anatomical localization, perilesional edema, and perilesional vascularity were evaluated. Due to the clustered data structure, the Generalized Estimating Equations (GEE) methodology was used. Model performance was evaluated using ROC curve analysis, precision-recall curve, and Brier score. Statistical analyses were performed using Jamovi v2.6.44, JASP v0.19.3, and R v4.5.1 software.

RESULTS: Metastatic lesions were significantly smaller than benign lesions (median 17.0 mm vs. 33.3 mm; p<0.001). In the GEE analysis, the presence of perilesional edema increased the likelihood of metastasis by 35 times (OR=35.25; 95% CI: 7.58-164.00; p<0.001), and perilesional vascularity increased the likelihood of metastasis by 45 times (OR=44.54; 95% CI: 1.86-1066.00; p = 0.016). Abdominal-pelvic localization showed a 133- fold (OR=133.00; 95% CI: 10.90-1622.00; p<0.001) higher likelihood of metastasis compared to the extremities, while thoracic-anterior chest wall localization showed a 35-fold (OR=35.22; 95% CI: 2.41-514.00; p = 0.007). Each unit increase in standardized size reduced the likelihood of metastasis by 90% (OR=0.10; 95% CI: 0.02-0.42; p = 0.001). The model demonstrated excellent discrimination (AUC-ROC=0.947) and calibration (Brier score=0.075) performance.

DISCUSSION: Our results show that perilesional edema and perilesional vascularity are key diagnostic signs of metastatic lesions. The combined assessment of perilesional findings and anatomical localization can significantly enhance diagnostic accuracy in daily practice. Furthermore, the relationship between lesion size and the metastatic process emphasizes the need for more careful evaluation of smaller lesions.

CONCLUSION: Perilesional edema, perilesional vascularity, and trunk region localization were associated with soft tissue metastasis. These findings may be helpful in the radiological differentiation of metastasis and benign soft tissue lesions.

PMID:42273870 | DOI:10.2174/0115734056495507260515193943

Dan Med J. 2026 May 20;73(6):A11250903. doi: 10.61409/A11250903.

ABSTRACT

INTRODUCTION: Rectal prolapse significantly impairs the quality of life, particularly in elderly women. Minimally invasive ventral mesh rectopexy (VMR) has emerged as a preferred surgery, offering anatomical correction and improved post-operative outcomes. This study evaluated operative details, complications and recurrence rates following laparoscopic VMR (LVMR) and robotic VMR (RVMR) at a single tertiary care centre over a five-year period.

METHODS: A retrospective analysis of patients undergoing LVMR or RVMR between 2019 and 2024 was conducted. The two groups were compared with respect to demographics, intraoperative details, post-operative complications, recurrence and long-term functional outcomes.

RESULTS: A total of 88 patients had undergone VMR, of whom 69 underwent RVMR; 19 underwent LVMR. All patients were female with a median age of 74 years. External prolapse was observed in 97.7% of the patients, whereas only 2.3% had internal prolapse. No statistically significant difference was observed between LVMR and RVMR with regard to length of stay or mean operative time. The conversion rate was 0%. Post-operative complications were classified according to Clavien-Dindo (CD); six patients had a CD IIIb, requiring return to surgical theatres. A single patient from the RVMR group had a recurrence after one year.

CONCLUSIONS: Minimally invasive VMR is a long-term, effective and safe surgical method for rectal prolapse, with a low recurrence rate, a relatively low complication rate and a significant improvement in functional outcomes.

FUNDING: None.

TRIAL REGISTRATION: Not relevant.

PMID:42273863 | DOI:10.61409/A11250903

Cancer Med. 2026 Jun;15(6):e71963. doi: 10.1002/cam4.71963.

ABSTRACT

AIM: We examine the impact of interferon (IFN)-based and IFN-free treatment on the prognosis of hepatocellular carcinoma (HCC) after hepatitis C virus (HCV) sustained virological response (SVR).

METHODS: Clinical information was collected on 311 cases of HCC after HCV-SVR from 16 facilities affiliated with the Kyushu Liver Surgery Study Group. Clinical factors and the tumor microenvironment of HCC after SVR treatment with IFN-based and IFN-free treatments were analyzed.

RESULTS: No statistically significant differences were observed in the recurrence rate and overall survival (OS) between the two groups. Propensity score-matched analysis similarly showed no statistically significant differences in recurrence and OS. In the IFN-based treatment group, OS time was significantly shorter for the programmed death-ligand 1(PD-L1)-positive HCC than the PD-L1-negative HCC group (p = 0.0183). No significant difference was observed in the recurrence rate between PD-L1-positive and PD-L1-negative HCC groups. In the IFN-based treatment group, the recurrence rate in the cluster of differentiation (CD) 8-positive group was significantly lower than in the CD8-negative group (p = 0.0292). There was no difference in OS time between the CD8-positive and CD8-negative groups. In the IFN-free treatment group, PD-L1 and CD8 were not associated with recurrence rate or OS.

CONCLUSIONS: No statistically significant differences were observed in recurrence or OS rate after HCC resection in the IFN-free treatment group compared with the IFN-based treatment group. In the IFN-based treatment group, PD-L1 and CD8 expression on cancer cells might be prognostic factors.

PMID:42273857 | DOI:10.1002/cam4.71963

Inquiry. 2026 Jan-Dec;63:469580261444668. doi: 10.1177/00469580261444668. Epub 2026 Jun 11.

ABSTRACT

Relative deprivation has been associated with adverse health outcomes, yet the psychological pathways linking perceived socioeconomic disadvantage to health remain insufficiently understood, particularly in the Chinese context. This study therefore examines whether subjective well-being mediates the relationship between relative deprivation and health among Chinese adults. This cross-sectional study analyzed 5604 adults aged ≥ 18 using data from the 2023 Chinese General Social Survey (CGSS) conducted in China. After computing descriptive statistics and correlations, stepwise regression models tested whether subjective well-being mediated the relationship between relative deprivation and health. Indirect effects were estimated using bias-corrected bootstrap procedures (5000 replications). Urban-rural stratification examined geographic heterogeneity in mediation pathways, while robustness checks included ordered logit models and a sensitivity analysis that dichotomized subjective well-being into low and high categories. STATA 18.0 was used throughout. The results indicated that relative deprivation demonstrated a significant negative association with Chinese adults’ health. Furthermore, subjective well-being was independently associated with better health outcomes. Mediation analysis revealed that subjective well-being significantly mediated 20.5% of the total effect between relative deprivation and health. Regional stratification showed that the indirect pathway through subjective well-being did not differ significantly between urban and rural residents, whereas the direct effect of relative deprivation on health was significantly stronger among rural residents. These findings highlight the potential role of subjective well-being as a psychological mechanism linking perceived relative deprivation to health. While this pathway appears comparable across urban and rural contexts, the association between relative deprivation and health is stronger among rural residents. Addressing health disparities therefore requires attention to both psychosocial processes and the broader structural conditions that shape socioeconomic inequality.

PMID:42273849 | DOI:10.1177/00469580261444668

J Chem Inf Model. 2026 Jun 11. doi: 10.1021/acs.jcim.6c00675. Online ahead of print.

ABSTRACT

Two-dimensional (2D) inorganic crystals are a class of materials that are gaining significant attention for use in electronic and optoelectronic devices. Among many other exciting applications, 2D materials offer a range of beneficial properties to such devices owing to charge carrier confinement, high carrier mobility, tunable band gaps, strong light-matter interactions, and atomically thin geometries that enable excellent electrostatic control and mechanical flexibility. In parallel, data-driven approaches to predictions of inorganic material properties have gained considerable attention as computationally lightweight surrogate models for properties of interest. This is particularly important when screening candidate materials for particular sets of structure-property relationships. Many of these approaches have targeted three-dimensional bulk crystalline materials. In this work, we develop a set of data-driven models for predicting the properties of 2D layered, van der Waals, and ultrathin film materials, namely, thermodynamic stability, metallicity, and electronic band gap. We train the models on materials sourced from open-source computational databases of 2D materials (Alexandria_2D, C2DB, MC2D, and 2DMatpedia) and use chemically relevant elemental, physical, and compositional features as input. The large feature space is reduced to a subset of critical features by a statistical and gradient-boosted feature selection strategy. The models are fully interpretable with feature relevance scores and SHapley Additive exPlanations analysis assessing the global and local influence of feature values. The classifiers for thermodynamic stability and metallicity achieve F₁-scores of 0.832 and 0.870 and accuracies of 89.7% and 89.7%, respectively. The regressor model for the band gap achieves an R² of 0.883, a mean-absolute error (MAE) of 0.317 eV, and a root-mean-squared error (RMSE) of 0.485 eV on the in-distribution test set. We assess the band gap predictor regressor against a 2D material band gap data set (N ∈ 177) manually extracted from the academic literature to quantify the model’s ability to predict outside of the training distribution, achieving an R² of 0.334, an MAE of 0.675 eV, and an RMSE of 0.961 eV. These results demonstrate the efficacy of feature selection in producing fully explainable machine learning surrogate models for high-throughput property prediction for 2D materials.

PMID:42273839 | DOI:10.1021/acs.jcim.6c00675

Afr J Reprod Health. 2026 Jun 11;(11):20-27. doi: 10.29063/ajrh2026/v30i11.2.

ABSTRACT

Contraceptive failure remains a global public health challenge among HIV positive women. Therefore, this study aimed to propose a contraceptive failure prediction framework based on the findings from HIV-positive women of reproductive age in South Africa and the perception of their healthcare providers. A mixed-methods approach was utilized, with participants purposively selected based on their knowledge and experiences. Qualitative data gathered from two focus group discussions (FGDs) involving seven (n = 7) healthcare workers (HCWs) and in-depth interviews involving ten (n = 10) HIV-positive women as well as quantitative data comprising a cohort of one hundred and seventy-three (n = 173) HIV-positive women who participated in the South African National HIV Prevalence, Incidence, Behavior Survey (SABSSM) were used to identify contraceptive failure determinants. Participant demographic information was analyzed using descriptive statistics, while inferential analysis was done using multiple regression statistics. The qualitative data were thematically analyzed. Healthcare workers’ (HCWs’) support, knowledge of contraceptive use, access to preferred contraceptives, and income collectively predicted contraceptive failure. These findings facilitated the development of a Contraceptive Failure Prediction Framework (CFPF), interpreted here as a determinants framework.

PMID:42273838 | DOI:10.29063/ajrh2026/v30i11.2

J Gerontol A Biol Sci Med Sci. 2026 Jun 10:glag143. doi: 10.1093/gerona/glag143. Online ahead of print.

ABSTRACT

BACKGROUND: Physical activity (PA), sedentary behavior (SB), and sleep are modifiable lifestyle factors linked to cognitive health. Prior research has largely relied on self-report to measure these constructs or ignored the compositional nature of these data. We applied compositional data analysis (CoDA) to quantify the association between accelerometry-derived measures of time in major activity domains and cognitive outcomes among older adults.

METHODS: Participants were 927 older adults (M age = 82.71 years, SD = 3.99, range = 76-95) from the Atherosclerosis Risk in Communities (ARIC) study who wore wrist-worn accelerometers and completed cognitive assessments. Major activity domains were time in bed and time out of bed, which were divided into four categories: sedentary behavior (SB), low-light (low-LIPA), high-light (high-LIPA), and moderate-to-vigorous physical activity (MVPA). Cognitive outcomes included global cognition, memory, executive function, and language performance.

RESULTS: Across all cognitive domains, greater time spent in MVPA relative to other activity domains, including out-of-bed time, sedentary behavior, and low-LIPA, was associated with better cognitive performance. In contrast, greater time spent in low-LIPA relative to other activity domains was consistently associated with poorer performance across all cognitive domains, and these associations were attenuated, but remained statistically significant, after adjustment for gait speed. Sedentary behavior was negatively associated with executive function only.

CONCLUSIONS: The trade-off between time spent in MVPA and in low-LIPA is the objectively measured physical activity that was most strongly associated with cognitive abilities. Time spent in low-LIPA may be an important biomarker of physical health and cognitive impairment.

PMID:42268661 | DOI:10.1093/gerona/glag143

JAMA Dermatol. 2026 Jun 10. doi: 10.1001/jamadermatol.2026.1527. Online ahead of print.

ABSTRACT

IMPORTANCE: In Germany, a population-based skin cancer screening (SCS) program was implemented in 2008. The benefit of the intervention is unclear.

OBJECTIVE: To determine whether the German SCS program was associated with reduced melanoma mortality.

DESIGN, SETTING, AND PARTICIPANTS: This population-based, ecological comparative effectiveness study compared trends in melanoma mortality from 2009 to 2022 in Germany with those in neighboring countries without population-based SCS using data on melanoma mortality from the official cause-of-death statistics. The total populations of 15 federal states of Germany and 9 neighboring countries were examined. The German federal state of Schleswig-Holstein was excluded from the analysis due to a potential lasting effect of a preceding pilot project, conducted in 2003 and 2004. Data were analyzed from November 7, 2025, to February 11, 2026.

EXPOSURE: The German SCS program entitles men and women aged 35 years and older to a visual skin examination every 2 years, regardless of their individual skin cancer risk. The 2-year participation rate was estimated at approximately 32%.

MAIN OUTCOMES AND MEASURES: Pooled estimates of annual percentage changes (APCs) in age-standardized melanoma mortality rates in German and non-German control regions were calculated using a random-effects model.

RESULTS: Data include a mean of 79.1 million inhabitants in Germany and 164.8 million inhabitants in 9 neighboring countries. Between 2009 and 2022, age-standardized melanoma mortality rates decreased in each included region. In Germany, APCs ranged from -3.8% (95% CI, -5.5% to -2.2%) to -0.1% (95% CI, -1.7% to 1.5%). In the control regions, mortality rates decreased between -3.8% (95% CI, -4.9% to -2.7%) and -1.0% (95% CI, -1.9% to -0.2%) per year. Pooled APC estimates are -1.8% (95% CI, -2.3% to -1.4%) for Germany and -2.2% (95% CI, -2.8% to -1.6%) for the non-German control regions; the difference was not statistically significant (P = .42).

CONCLUSIONS AND RELEVANCE: The findings of this ecological study are in line with previous studies that failed to show a melanoma mortality benefit associated with the German SCS program. To enable a well-founded decision on the future of the program, the causes of its poor performance should be investigated.

PMID:42268621 | DOI:10.1001/jamadermatol.2026.1527

JAMA Netw Open. 2026 Jun 1;9(6):e2617684. doi: 10.1001/jamanetworkopen.2026.17684.

ABSTRACT

IMPORTANCE: Effectiveness of seasonal influenza vaccine must be assessed annually as influenza viruses evolve, requiring updated vaccine components. Evidence is limited for direct vaccine benefit in lowering risk of death following influenza virus infection.

OBJECTIVE: To evaluate the association between current-season influenza vaccination and laboratory-confirmed influenza virus infection and, among persons with laboratory-confirmed influenza, the association with influenza-associated death.

DESIGN, SETTING, AND PARTICIPANTS: A case-control analysis of California residents aged 6 months or older with influenza diagnostic testing ordered between October 1, 2024, and May 31, 2025, with data on diagnostic test result and 2024 to 2025 influenza vaccination. Influenza-associated deaths within 30 days of testing among persons with laboratory-confirmed influenza were identified from vital statistics records.

EXPOSURES: Influenza vaccination from October 1, 2024, to May 31, 2025.

MAIN OUTCOMES AND MEASURES: Associations were evaluated between 2024 to 2025 influenza vaccination and (1) laboratory-confirmed influenza, in which case individuals tested positive and control individuals tested negative for laboratory confirmed influenza; and (2) influenza-associated death, in which deaths were compared with surviving cases among persons with laboratory-confirmed influenza. Associations were assessed with mixed-effects logistic regression.

RESULTS: Among 1 106 628 persons tested for laboratory-confirmed influenza (610 093 female participants [55.1%]), 234 715 were case individuals with laboratory-confirmed influenza (median [IQR] age, 28 [10-52] years) and 871 913 were control individuals (median [IQR] age, 42 [19-67] years); 45 441 influenza cases (19.4%) and 255 605 control individuals (29.3%) had received 2024 to 2025 influenza vaccination. Vaccination was significantly associated with decreased likelihood of laboratory-confirmed influenza (estimated vaccine effectiveness, 40%; 95% CI, 39%-41%). Among persons with laboratory confirmed influenza, there were 801 influenza-associated deaths. Influenza-associated death among persons aged 65 years or older was associated with lower likelihood of influenza vaccination (adjusted odds ratio, 0.71; 95% CI, 0.60-0.84) compared with persons with nonfatal infection.

CONCLUSIONS AND RELEVANCE: In this case-control study of estimated vaccine effectiveness, influenza vaccination was associated with decreased likelihood of laboratory-confirmed influenza and influenza-associated deaths among those with laboratory-confirmed influenza. Electronic laboratory reporting linked to immunization registry and vital statistics provide tools to assess vaccine effectiveness and risk of death, a rare influenza-associated outcome.

PMID:42268610 | DOI:10.1001/jamanetworkopen.2026.17684

FASEB J. 2026 Jun 30;40(12):e71993. doi: 10.1096/fj.202600816RR.

ABSTRACT

Biological noise is ubiquitous in living systems; yet, it is often neglected in cell-based experiments, potentially biasing data interpretation. We provide a quantitative characterization of single-cell equivalent diameter distributions in six human cell types using cell counting. By analyzing thousands of cells over passage number, we show that cell size is phenotype dependent and varies significantly with passage, with most cell types exhibiting a progressive reduction in median diameter. This variability is structured: When diameters are converted to masses, the distributions obey scaling laws, revealing conserved statistical properties of human cells in culture. Because key physiological processes such as metabolism scale with cell mass, passage-dependent shifts in diameter distributions can propagate into functional readouts. We show that changes in cell size are sufficient to bias estimates of construct-level metabolic rate, potentially confounding the interpretation of size-normalized assays and treatment effects. Our results highlight that biological noise in vitro is a source of statistical structure that enables scaling analyses and a dynamic property that, if ignored, can lead to systematic misinterpretation of experimental outcomes. Accounting for size distributions and their evolution over passages may therefore improve experimental design, data interpretation, and, ultimately, the translatability of in vitro models.

PMID:42268592 | DOI:10.1096/fj.202600816RR