Tag: statistics

J Environ Radioact. 2025 Jul 31;289:107769. doi: 10.1016/j.jenvrad.2025.107769. Online ahead of print.

ABSTRACT

The existing data gaps in the sorption and desorption parameters of naturally occurring radionuclides (e.g., radium (Ra)) challenge the use of radioecological risk assessment models. We present two alternatives for deriving Ra distribution coefficients (K_d (Ra)) in soils when the physicochemical information on the solid and liquid phases involved is too scarce to apply parametric prediction models: the deduction of sorption parameters from those of chemical analogues (such as Ba and Sr) and the proposal of best estimate K_d (Ra) values deduced from probabilistic distribution functions of data grouped according to relevant factors affecting K_d (Ra) variability. Regarding the use of chemical analogues, partial least squares regression analysis and univariate linear correlations revealed that Ba and Ra sorption in soils was governed by the same soil properties (K_d (Ca + Mg) and Mn_am), related to exchangeable sites on the soil surface. The derivation of K_d (Ra) values from K_d (Ba) and also K_d (Sr) is feasible by applying suitable correction factors. Furthermore, several K_d (Ra) best estimates were derived from the distribution functions of K_d (Ra) datasets obtained from own and literature data. Statistical differences were noticed for the sorption and desorption datasets (the latter significantly affected by data from native Ra), leading to the proposal of distinct K_d (Ra) values (870 and 2760 L kg^-1, respectively) for uptake and remobilisation scenarios. Regarding the desorption data, diverse K_d (Ra) best estimates were suggested for acidic (1540 L kg^-1) and alkaline (6440 L kg^-1) soils. For the sorption data, statistically different K_d (Ra) best estimates were suggested according to pH (100 and 1240 L kg^-1 for pH < 4.5 and pH ≥ 7, respectively) and water-soluble Ca + Mg content, allowing for the selection of the most appropriate best estimate values for use in risk assessment models depending on the available information.

PMID:40749306 | DOI:10.1016/j.jenvrad.2025.107769

Med Image Anal. 2025 Jul 25;105:103741. doi: 10.1016/j.media.2025.103741. Online ahead of print.

ABSTRACT

The performance of a well-trained segmentation model is often trapped by domain shift caused by acquisition variance. Existing efforts are devoted to expanding the diversity of single-source samples, as well as learning domain-invariant representations. Essentially, they are still modeling the statistical dependence between sample-label pairs to achieve a superficial portrayal of reality. On the contrary, we propose a Spectrum Intervention based Invariant Causal Representation Learning (SI²CRL) framework, to unify the data generation and representation learning from causal view. Specifically, for the data generation, the unknown object elements can be reified in frequency domain as phase variables, then we propose an amplitude-based intervention module to generate low-frequency perturbations via random-weighted multilayer convolutional network. For the causal representations, a two-stage causal synergy modeling process is proposed to derive unobservable causal factors. In the first stage, the style-sensitive non-causal factors lying in the shallow layer of encoder are filtered out by contrastive-based causal decoupling mechanism. In the second stage, the hierarchical features in decoder are first factorized with cross-covariance regularization to ensure channel-wise independence; Subsequently, we introduce an adversarial-based causal purification module, which encourages the decoder to iteratively update causally sufficient information and make domain-robust predictions. We evaluate our SI²CRL against the state-of-the-art methods on cross-site prostate MRI segmentation, cross-modality (CT-MRI) abdominal multi-organ segmentation, and cross-sequence (MRI) cardiac segmentation. Our approach achieves consistent performance gains compared to these peer methods.

PMID:40749275 | DOI:10.1016/j.media.2025.103741

Accid Anal Prev. 2025 Jul 30;220:108172. doi: 10.1016/j.aap.2025.108172. Online ahead of print.

ABSTRACT

OBJECTIVES: Higher age is often seen as a key factor in the decline of driving skills. Moreover, there is some evidence that overestimation is related to both higher age and poorer on-road performance in older drivers. However, it is unknown how the extent of overestimation or underestimation affects on-road driving performance beyond age.

METHODS: 112 older drivers with and without mild cognitive impairment participated in this prospective on-road study. All participants underwent a standardized on-road driving assessment, neuropsychological testing, collection of driving-related data and different self-assessments. Statistical analyses included a hierarchical regression analysis to predict on-road driving performance by adding age and sex in the first step and the validity of self-assessment (VSA) in the second step. Correlation analyses focused on the association between VSA and cognitive and driving-related behavioral factors.

RESULTS: Results revealed that the combination of age and sex significantly predicted on-road driving skills (R²_adjusted = 0.320). The inclusion of VSA led to a significant increase of explained variance in the criterion (R²_adjusted = 0.639). Moreover, the degree of overestimation correlated with higher age, lower cognitive performance and more risky driving behavior.

DISCUSSION: Our results highlight the importance of the VSA for on-road driving performance beyond the effects of age and sex. Moreover, the VSA appears to decrease towards overestimation with increasing age and decreasing cognition and should thus be a focus of safety research.

PMID:40749272 | DOI:10.1016/j.aap.2025.108172

ESMO Open. 2025 Jul 31;10(8):105516. doi: 10.1016/j.esmoop.2025.105516. Online ahead of print.

ABSTRACT

BACKGROUND: Several studies showed that in mismatch repair-deficient (dMMR)/microsatellite instability (MSI) nonmetastatic colon cancer (CC), neoadjuvant immune checkpoint inhibitors (ICIs) were associated with exceptional rates of pathological major response (pMR) and complete response (pCR). Patients included in these trials, however, were highly selected, and real-life data are now needed to better evaluate the efficacy and tolerability of neoadjuvant ICIs in routine clinical practice.

PATIENTS AND METHODS: This retrospective observational study aimed to describe the clinical practices regarding ICIs in the neoadjuvant setting for patients with dMMR/MSI nonmetastatic CC, and to evaluate their efficacy and safety in real-world conditions. Patients receiving ICIs as part of a therapeutic trial were excluded.

RESULTS: Between 2019 and 2024, 32 patients were included across six French centers [median age 70 years (range 51-76 years), Lynch syndrome 31%]. Twenty-four patients had right-sided CC (85%), including three (9%) with two primary tumor locations. Ten patients (31%) received ipilimumab + nivolumab (NICHE regimen), while 22 (69%) were treated with pembrolizumab monotherapy. Grade ≥3 toxicities were observed in six patients (19%), including one toxic death and three toxicities (9%) leading to treatment discontinuation (one rheumatoid polyarthritis-like syndrome and two tumor fistulizations). Three patients developed bowel obstruction while receiving ICIs, two of whom underwent surgery showing pCR and pMR. Thirty patients were resected and 33 tumors were analyzed histologically with pMR in 21 cases (64%) including pCR in 14 cases (42%). The presence of an independent-cells contingent was statistically associated with poor pathological response.

CONCLUSIONS: This retrospective real-world study confirms the excellent results of neoadjuvant ICIs in dMMR/MSI nonmetastatic CC patients. pMR and pCR rates were, however, lower than those published in previous studies (64% and 42%, respectively), with higher rates of grade ≥3 toxicity, including one potential toxic death and three treatment discontinuations.

PMID:40749247 | DOI:10.1016/j.esmoop.2025.105516

J Neurosurg Pediatr. 2025 Aug 1:1-8. doi: 10.3171/2025.4.PEDS25100. Online ahead of print.

ABSTRACT

OBJECTIVE: This study aimed to investigate the effect of intrathecal baclofen (ITB) on BMI over time in a large pediatric cohort of patients with cerebral palsy (CP).

METHODS: The medical records of pediatric patients diagnosed with CP who underwent ITB pump placement at Texas Children’s Hospital between 2007 and 2024 were retrospectively reviewed. Pre- and postoperative BMI, demographic information, and clinical characteristics were collected. Multiple BMI velocities were calculated. A linear mixed-effects model was used to account for interpatient variability.

RESULTS: Among 237 patients, the average BMI was 17.90 (SD 4.00) kg/m2 preoperatively and 19.13 (SD 4.58) kg/m2 postoperatively, showing a significant difference (p = 0.001, Kruskal-Wallis test) but a small effect size (η2 = 0.02, 95% CI 0.001-0.05). The average BMI velocity was 0.55 (SD 3.5) kg/m2/yr presurgery and 0.46 (SD 3.5) kg/m2/yr postsurgery, showing no significant difference (p = 0.52, t-test). The mixed-effects model found no statistically significant effect of ITB surgery on BMI rate of change by catheter level. Specifically, cervical (p = 0.97), high thoracic (p = 0.41), midthoracic (p = 0.63), and low thoracic (p = 0.84) catheter levels were nonsignificant in effect on BMI.

CONCLUSIONS: Although there was an absolute increase in BMI postoperatively, the small effect size and results of the linear mixed-effects model-accounting for clinical confounders, within-patient variability, and catheter level-demonstrated that ITB surgery does not significantly affect BMI. The authors conclude that improved tone control may not substantially impact BMI, necessitating further nutritional intervention to ensure optimal BMI.

PMID:40749238 | DOI:10.3171/2025.4.PEDS25100

J Neurosurg Spine. 2025 Aug 1:1-9. doi: 10.3171/2025.5.SPINE25304. Online ahead of print.

ABSTRACT

OBJECTIVE: Despite growing interest in using biometric measures as objective tools to monitor recovery after spine surgery, there is sparse guidance on appropriate thresholds to evaluate perioperative changes in these domains. The authors defined thresholds of activity and heart rate (HR) measures that mark improved early recovery and evaluated their association with longer-term surgical outcomes.

METHODS: A prospective observational cohort study recruited patients undergoing elective surgery for lumbar degenerative spine disease at a single institution from February 2021 to June 2023. Patient-reported outcome measures (PROMs) were collected preoperatively and at 1 month and 6 months after surgery. The authors evaluated 50 metrics of physical fitness available on Fitbit fitness trackers that assessed activity amount, activity intensity, and physiological status based on the responsiveness of each after surgery and correlation with clinical metrics. Anchor-based methods were used to calculate Fitbit metric cutoffs that delineated clinical improvement, with Patient-Reported Outcome Measurement Information System physical function (PF) and pain interference scores used as anchors at 1 month postoperatively. The receiver operating characteristic (ROC) analysis was performed to identify optimal thresholds. Six-month clinical outcomes were compared between patient groups based on the newly defined Fitbit metric cutoffs. Multivariable regression, with 6-month PROMs as the dependent variable, were used to control for clinical confounders.

RESULTS: Sixty-two patients (mean age 62 [SD 12] years, 33 [53%] female) were included. PROMs at 6 months were significantly improved for all variables. Among 50 candidate Fitbit metrics, 4 were selected: active time (cutoff: 7.6% decrease, area under the ROC curve [AUC] 0.71), daily steps (cutoff: 1.55 steps/min decrease, AUC 0.70), time spent moving at ≥ 40 steps/min (cutoff: 0.7% decrease, AUC 0.67), and HR stability (cutoff: 15% increase, AUC 0.72). When separating patients into groups based on newly defined Fitbit cutoffs, these Fitbit metrics had at least one significant association with 6-month PROMs. These measures improved predictions of 6-month outcomes based on R2 and, except for daily steps, remained statistically significant on multivariable regression.

CONCLUSIONS: By monitoring the dynamic nature of a patient’s PF through wearable trackers, clinicians can gain valuable insights into functional recovery that can complement traditional PROMs. The cutoffs derived from this study can potentially serve as important benchmarks for future studies examining physiological markers of surgical recovery.

PMID:40749235 | DOI:10.3171/2025.5.SPINE25304

JMIR Form Res. 2025 Aug 1;9:e75120. doi: 10.2196/75120.

ABSTRACT

BACKGROUND: The rise of short-video platforms, such as TikTok (Douyin in China) and Bilibili, has significantly influenced how health information is disseminated to the public. However, the quality, reliability, and effectiveness of health-related content on these platforms, particularly regarding uterine fibroids, remain underexplored. Uterine fibroids are a common medical condition that affects a substantial proportion of women worldwide. While these platforms have become vital sources of health education, misinformation and incomplete content may undermine their efficacy.

OBJECTIVE: This study aims to address these gaps by evaluating the quality and dissemination effectiveness of uterine fibroid-related health information on TikTok and Bilibili.

METHODS: A total of 200 uterine fibroid-related videos (100 from TikTok and 100 from Bilibili) were selected through a keyword search. The videos were evaluated by 2 trained gynecological experts using the Global Quality Score (GQS) and a modified DISCERN (mDISCERN) tool. In addition, the Patient Education Materials Assessment Tool for Audio and Visual Materials was used to assess the understandability and actionability of the videos. Statistical analyses, including the Mann-Whitney U test, Spearman rank correlation, and stepwise regression analysis, were used to assess differences between platforms and identify predictors of video quality.

RESULTS: The results indicated that TikTok outperformed Bilibili in terms of user engagement metrics, such as likes, comments, shares, and followers (all P<.001). However, Bilibili videos were generally longer than those on TikTok (P<.001). The videos on both platforms demonstrated suboptimal overall quality and reliability, reflected by median GQS score of 3 (IQR 3-4) for TikTok and the median GQS score of Bilibili is 3 (IQR 2-4). The median modified DISCERN scores were also low: 2 (IQR 2-2) for TikTok and 2 (IQR 2-2) for Bilibili, with no significant differences between the 2 platforms (P=.62 for GQS; P=.18 for mDISCERN). The videos on both platforms yielded comparable median scores for Patient Education Materials Assessment Tool-Understandability (PEMAT-U) and Patient Education Materials Assessment Tool-Actionability (PEMAT-A). The median score of PEMAT-U was 77% (IQR 69%-83%) for TikTok and 77% (IQR 69%-85%) for Bilibili. The PEMAT-A yielded a median score of 67% (IQR 33%-67%) for TikTok and 67% (IQR 0-67%) for Bilibili. Videos uploaded by medical professionals on TikTok had significantly higher quality scores compared to those uploaded by nonprofessionals. A moderate positive correlation was observed between the GQS and mDISCERN scores (r=0.41, P<.01), indicating an interrelationship between quality and reliability. Stepwise regression analysis identified “completeness score,” “source,” and “PEMAT scores” as significant predictors of video quality.

CONCLUSIONS: This study highlights the generally low quality of uterine fibroid-related health information on short-video platforms, although TikTok showed better performance in terms of engagement and quality. The involvement of medical professionals was found to enhance video quality. These findings underscore the need for improved oversight of health content on social media platforms and greater involvement of health care professionals to ensure the dissemination of accurate and reliable health information.

PMID:40749219 | DOI:10.2196/75120

JMIR Serious Games. 2025 Aug 1;13:e76357. doi: 10.2196/76357.

ABSTRACT

BACKGROUND: Sarcopenia in older adults is associated with reduced muscle mass and function, leading to frailty, increased fall risk, and decreased quality of life (QOL). Mixed reality (MR)-based interventions have emerged as promising tools to enhance physical therapy engagement and effectiveness through immersive, interactive environments.

OBJECTIVE: This study aimed to investigate the effects of a Mixed Reality-Based Physical Therapy (Mr.PT) platform on quadricep muscle thickness, balance confidence, activities of daily living, and QOL in older adults with sarcopenia.

METHODS: A preliminary randomized controlled trial was conducted involving 30 older women (mean age 75.3, SD 9.9 y) diagnosed with sarcopenia based on the Asian Working Group for Sarcopenia criteria. Participants were randomly assigned to either the Mr.PT group or a conventional physical activity (CPA) group. Both groups participated in 30-minute exercise sessions, 3 times per week, over 4 weeks. The Mr.PT program used head-mounted MR devices with gamified, interactive training, while the CPA group received standard therapist-led exercises using resistance bands. Outcome measures included ultrasound imaging of quadricep muscle thickness, the Korean version of the Activities of Daily Living scale (Katz Index of Independence in Activities of Daily Living), the Activities-Specific Balance Confidence (ABC) scale, and the 12-Item Short-Form Health Survey. Statistical analysis was performed using repeated-measures ANOVA and Tukey post hoc tests.

RESULTS: The Mr.PT group showed significantly greater improvement in quadricep muscle thickness than the CPA group (P=.001). Within-group improvements in balance confidence (ABC scale) and daily functioning (Katz Index of Independence in Activities of Daily Living) were observed in both groups (P<.05), though between-group differences were not statistically significant. However, the Mr.PT group demonstrated significantly greater gains in QOL as measured by 12-Item Short-Form Health Survey (P=.02). All participants completed the intervention without dropouts or adverse events.

CONCLUSIONS: MR-based exercise using the Mr.PT platform appears effective in increasing muscle mass and enhancing QOL among older adults with sarcopenia. Its interactive and adaptive features may improve engagement and motivation, suggesting potential advantages over traditional programs. Further research with larger cohorts and longer follow-up is recommended to confirm these preliminary findings and explore long-term outcomes.

PMID:40749218 | DOI:10.2196/76357

J Contin Educ Nurs. 2025 Aug;56(8):322-328. doi: 10.3928/00220124-20250612-02. Epub 2025 Aug 1.

ABSTRACT

BACKGROUND: Early career nurses with 1 to 2 years of experience are offered minimal structured support after onboarding. During this period, nurses often contemplate their next career steps without guidance, sometimes leaving an organization.

METHOD: A quantitative quasi-experimental study was conducted to determine whether career coaching focused on professional development and mentorship would (a) increase retention and (b) improve engagement of pediatric nurses with 1 to 2 years of experience.

RESULTS: Participant retention remained 100% throughout the study. Nonparticipant retention dropped to 96% at 12 weeks postintervention and 94% at 20 weeks postintervention. Engagement scores increased after the career coaching sessions, with no statistically significant difference in mean values for engagement pre- and postintervention (t = -1.06, df = 8.00, p = .320).

CONCLUSION: Engagement improved postintervention. Retention was greater for participants than for nonparticipants. Increasing nurse retention positively impacts professional, financial, and patient care outcomes for nurses and the health care community.

PMID:40749214 | DOI:10.3928/00220124-20250612-02

J Med Internet Res. 2025 Aug 1;27:e60546. doi: 10.2196/60546.

ABSTRACT

BACKGROUND: Internet medical services (IMS) expanded rapidly in China during the COVID-19 pandemic. Unfortunately, disparities in internet medical services utilization (IMSU) have marginalized disadvantaged groups of Chinese patients from digital health benefits. The extent and nature of these disparities remain poorly understood, with no research comprehensively addressing how unfavorable predictors, including demographic, socioeconomic, and health-related factors, shape IMSU status, patterns, and preferences in China after the COVID-19 pandemic.

OBJECTIVE: This study aims to gain a deeper understanding of the disparities and unfavorable predictors that limit IMSU among Chinese adult patients following the COVID-19 pandemic, providing key reference points for advancing equitable IMSU.

METHODS: This cross-sectional study used a triangular approach, combining data from a provincial field survey conducted in July 2023 and a national online survey conducted in March 2024. Participants were Chinese adult outpatients aged 18 years or older. Descriptive and comparative analyses were used to examine disparities in IMSU status, patterns, and preferences across different demographic, socioeconomic, and health status-related factors. Binary logistic regression models were applied to assess the associations between unfavorable predictors (constructed from selected demographic, socioeconomic, and health status-related factors) and IMSU status, patterns, and preferences.

RESULTS: Of the 2011 eligible participants in the Jiangsu provincial field survey, 787 (39.13%) reported using IMS at least once in the previous 12 months. Among the 1611 eligible participants in the national online survey, all reported accessing IMS during the same period; however, only 481 (29.86%) reported high-frequency use (defined as usage in the third quartile or above, ie, ≥6 times). Overall, participants with unfavorable predictors were less likely to engage in IMSU, deep IMS were used less frequently than shallow IMS, and participants with 6 or more cumulative unfavorable predictors had the lowest total of IMSU preference score (mean 48.98 and 57.37 in the provincial field and national online surveys, respectively). Based on combined data from the provincial field and national online surveys, significantly negative associations were observed between unfavorable predictors and IMSU status, patterns, and preferences. In particular, participants aged 60 years or above (odds ratio [OR] 0.40, 95% CI 0.25-0.63, P<.001) and those without everyday internet access (OR 0.54, 95% CI 0.41-0.71, P<.001) were the least likely to use IMS at a medium-to-high frequency. Participants without private health insurance (OR 0.59, 95% CI 0.44-0.79, P<.001) were the least likely to utilize deep IMS. Moreover, participants aged 60 years or above (OR 0.45, 95% CI 0.33-0.63, P<.001) and those with a high school education or less (OR 0.67, 95% CI 0.55-0.82, P<.001) were the least likely to prefer IMS to a moderate-to-strong degree.

CONCLUSIONS: Widespread disparities in IMSU status, patterns, and preferences persisted among Chinese adult patients after the COVID-19 pandemic. More pro-disadvantaged patient policies may be warranted to narrow these disparities in IMSU, such as reimbursement for IMSU, to promote digital health equity in China.

PMID:40749207 | DOI:10.2196/60546