Tag: statistics

BJC Rep. 2025 Sep 16;3(1):61. doi: 10.1038/s44276-025-00178-7.

ABSTRACT

BACKGROUND: Immune checkpoint inhibitors (ICIs) have revolutionized the treatment of advanced renal cell carcinoma (RCC), but their use is associated with immune-related adverse events, including hepatic adverse events (irHAEs).

METHODS: We retrospectively analysed 105 RCC patients treated with ICIs as first-line therapy between 2018 and 2023 at the University Hospital of Essen. Patients were categorized by the development of irHAE, defined per CTCAE grading v5.0. Multivariable logistic regression was used to identify risk factors, while Kaplan-Meier survival analyses evaluated PFS and OS.

RESULTS: Among the cohort, 16.19% (n = 17) developed irHAE, while 8.57% (n = 9) experienced higher-grade events. Combination therapy with tyrosine kinase inhibitors (TKIs) was associated with a higher likelihood of irHAE (OR: 7.69, p = 0.037) compared to ICI-only regimens, with cabozantinib showing a significantly shorter time to onset (35 vs. 84 days; p < 0.001). Patients with a BMI ≥ 25 had a significantly increased risk (p = 0.011). Differences in PFS (18.63 vs. 19.87 months; p = 0.099) and OS (27.80 vs. 23.87 months; p = 0.36) were not statistically significant.

CONCLUSIONS: The combination of ICI with TKI posed higher risks for irHAE in RCC patients. While survival outcomes were unaffected, the results underscore the need for tailored monitoring and management. Prospective studies are warranted to refine therapeutic approaches.

PMID:40957947 | DOI:10.1038/s44276-025-00178-7

Nat Rev Genet. 2025 Oct;26(10):661-662. doi: 10.1038/s41576-025-00885-4.

NO ABSTRACT

PMID:40957943 | DOI:10.1038/s41576-025-00885-4

Nat Commun. 2025 Sep 16;16(1):8296. doi: 10.1038/s41467-025-64015-8.

ABSTRACT

Floods displace an average of 12 million people every year, and are responsible for 54% of all disaster-induced displacements. Displacement risk scales with the vulnerability of exposed populations, but this vulnerability is poorly understood at a global scale. Here we show that measures of human development and rural areas explain more of the variance of displacement vulnerability than income levels measured by gross domestic product. We combine global flood and displacement data to estimate vulnerability, as the ratio of displacement to exposure, for over 300 historical flood events. We find that this vulnerability varies by several orders of magnitude both between and within countries. A random forest regression shows that infant mortality rate and population density are among the most important predictors of displacement vulnerability at national level and within countries, respectively, highlighting the vulnerability of low-income and marginalized populations and of rural communities. Our results indicate that, rather than relying on overall economic development alone, targeted investments are needed to improve living conditions and coping capacities for the most vulnerable groups, particularly outside of large cities, and to prepare for increasing flood hazards due to climate change.

PMID:40957930 | DOI:10.1038/s41467-025-64015-8

J Med Internet Res. 2025 Sep 16;27:e77014. doi: 10.2196/77014.

ABSTRACT

BACKGROUND: Telemedicine is an effective and promising strategy, especially for the initial stages of a home-based therapeutic exercise program.

OBJECTIVE: The objectives of this study were 2-fold: first, to assess whether Tai Chi practice combined with wearable device-based monitoring improves cognitive function in this population, and second, to explore the underlying mechanisms for any improvements observed, including changes in physical activity levels and sleep patterns.

METHODS: The study was a randomized controlled trial in which participants were randomized (1:1:1) to receive usual care, fitness walking, or Tai Chi exercise. All indicators were assessed at baseline and 12-week follow-up. The usual care includes traditional diabetes education. Participants in the fitness walking group performed walking exercises on a treadmill under the supervision of a researcher 3 times a week for 12 weeks. Participants in the Tai Chi group practiced 24-style Simplified Tai Chi through live video streaming under the guidance of professors and professionals. In this 12-week program, participants underwent continuous glucose monitoring (CGM) using Guardian Sensors 3, CGM sensors attached to the upper arm. All participants carried bracelets to record their heart rate, sleep parameters, and steps. The primary outcome was the Montreal Cognitive Assessment (MoCA) at 12 weeks. Secondary outcomes included other cognitive subdomain tests and blood metabolic indices. The MoCA is a tool designed for rapid screening for mild cognitive impairment (MCI) and early dementia, with the core advantage of being more sensitive to early cognitive problems. The MoCA has a total score of 30. Lower scores may indicate the presence of cognitive dysfunction.

RESULTS: After 12 weeks of intervention, the Tai Chi exercise group showed a significant improvement in MoCA scores from baseline (mean difference 23.83, 95% CI 17.79-25.66 vs 21.42, 95% CI 17.11-24.74; P=.03). The fitness walking exercise group showed an improvement in MoCA scores (22.94, 95% CI 18.05-23.98 vs 21.58, 95% CI 17.35-24.12; P.08), but this did not reach statistical significance. Furthermore, there was a statistical difference in the improvement of MoCA scores between the Tai Chi and fitness walking groups (2.65, 95% CI 0.34-4.41 vs 1.44, 95% CI 0.89-2.87; P<.05). The usual care group showed the least change in score at both points (0.23, 95% CI -0.02 to 1.39; P=.83). Compared with the MQ in the fitness walking group (91.93, 95% CI 77.83-97.47) vs 88.62, 95% CI 77.14-95.84; P=.45), Trail Making Test Part B (TMT-B) (220.81, 95% CI 210.03-233.49 vs 223.66, 95% CI 215.04-230.27; P=.33), the Tai Chi group was more effective in improving the MQ (99.23, 95% CI 80.55-107.69 vs 89.23, 95% CI 78.16-96.08; P=.001), TMT-B (207.33, 95% CI 200.26-220.82 vs 225.58, 95% CI 214.12-234.94; P=.001) scores, and there were significant differences between the two groups.

CONCLUSIONS: In summary, this study demonstrated that web-based exercise therapy for patients may enhance the effectiveness of exercise therapy in improving cognitive function among older individuals with type 2 diabetes mellitus. Tai Chi has significant advantages in improving cognitive function and sleep quality, while fitness walking, although also beneficial, is relatively weak in these areas.

PMID:40957073 | DOI:10.2196/77014

Clin Exp Dent Res. 2025 Oct;11(5):e70222. doi: 10.1002/cre2.70222.

ABSTRACT

OBJECTIVES: This in vitro experimental study aimed to evaluate the effects of laboratory contamination of implant abutment screws and connection surfaces on reverse torque values (RTVs), as an indicator of screw preload loss.

MATERIAL AND METHODS: Forty-five Dentis One Q implants and 45 CCM UCLA abutments were randomly assigned into three groups (n = 15 per group). Group 1 (control) involved uncontaminated abutments and screws with no restorations. Group 2 (screw contamination) used new abutments attached with screws contaminated by laboratory materials (porcelain powder, metal debris, and polishing paste). Group 3 (connection contamination) included screw-retained restorations fabricated under contaminated conditions and attached using new screws. All samples were subjected to standardized torque (250 N·cm), thermocycling (1500 cycles between 5°C and 55°C), and subsequent RTV measurement. One-way ANOVA and Tukey’s post hoc tests were used for statistical analysis (α = 0.05).

RESULTS: Mean RTVs (SD) were 218 (0.15) N·cm (control), 181 (0.14) N·cm (screw contamination), and 207 (0.11) N·cm (connection contamination). RTVs in the screw contamination group were significantly lower than both the control (p < 0.001) and connection contamination groups (p < 0.001). The difference between the control and connection contamination groups was not statistically significant (p = 0.08).

CONCLUSIONS: Laboratory contamination of implant components can significantly reduce reverse torque values, particularly in cases of screw contamination, indicating an increased risk of screw loosening. Contamination control during prosthetic procedures is essential to maintaining implant stability.

PMID:40957072 | DOI:10.1002/cre2.70222

Int J Prosthodont. 2025 Sep 16;0(0):1-28. doi: 10.11607/ijp.9306. Online ahead of print.

ABSTRACT

PURPOSE: This study aims to assess the accuracy of one-piece implant placement using stackable surgical guides by means of a novel stent defined as ‘number guide’ screwed onto the fixed base template in order to improve the intra-surgical registration of the scan bodies.

MATERIAL AND METHODS: A digital workflow was used for one-piece implant planning and placement. Participants were selected based on specific criteria, and stackable guides were used for fully guided implant placement with immediate provisional loading. The accuracy of implant placement was assessed by comparing pre- and post-intervention data registered with the aid of the number guide.

RESULTS: Ten participants were treated with forty-seven one-piece implants placed in the maxilla and mandible. All implants achieved adequate primary stability, allowing immediate loading. Accuracy was measured by linear and angular deviations, showing greater precision in the mandible. Implant characteristics yielded statistically significant differences in terms of accuracy.

CONCLUSIONS: Fully guided digital workflow ensured precise implant placement and immediate provisional loading. The additional use of the number guide allowed effective accuracy assessments of the system in terms of linear and angular deviations, highlighting its potential to assess post-operative implant placement.

PMID:40957069 | DOI:10.11607/ijp.9306

JMIR Med Educ. 2025 Sep 16;11:e75916. doi: 10.2196/75916.

ABSTRACT

BACKGROUND: Effective diabetes management requires behavioral change support from primary care providers. However, general practitioners (GPs) often lack training in patient-centered communication methods such as motivational interviewing (MI), especially in time-constrained settings. While brief MI offers a practical alternative, evidence on its impact among GPs and patient outcomes remains limited.

OBJECTIVE: This study aimed to evaluate the effectiveness of a structured MI educational program for GPs and GP trainees on their MI knowledge and confidence, and its impact on clinical outcomes among patients with type 2 diabetes in primary care settings.

METHODS: A mixed methods study was conducted using a before-and-after two-group design with quantitative assessments of GPs’ knowledge and patients’ biomarkers, supplemented by qualitative interviews. The intervention group (n=35) received a 4-hour interactive MI workshop, optional web-based modules, and brief MI guides. The control group received standard care. A total of 149 and 167 patients with diabetes were included in the study and control groups, respectively.

RESULTS: A paired-sample t test was conducted to evaluate the impact of the MI course on the learners’ knowledge. There was a statistically significant difference in the knowledge test scores from Time 1 (mean 11.46, SD 3.48) to Time 2 (mean 15.04, SD 2.35), t28= -7.74; P<.001 (2-tailed). The mean increase in knowledge score was 3.57 (SD 2.44), with a 95% CI of 2.62 to 4.52, indicating a large and statistically significant effect. The eta-squared statistic indicated a large effect size (eta-squared=0.85). Patients in the intervention group had greater improvements in HbA1c (mean difference= -0.50, 95% CI -0.91 to -0.09; P=.02) and diastolic blood pressure (mean difference= -5.96 mmHg, 95% CI -8.66 to -3.25; P<.001) compared to controls. Qualitative feedback highlighted the usefulness of brief MI, along with challenges in mastering advanced techniques and time constraints.

CONCLUSIONS: The MI educational program improved GP trainees’ MI knowledge and patient outcomes. Brief MI appears feasible in primary care but requires ongoing support for skill development and implementation.

PMID:40957062 | DOI:10.2196/75916

JMIR Public Health Surveill. 2025 Sep 16;11:e75291. doi: 10.2196/75291.

ABSTRACT

BACKGROUND: Incident HIV infection is a critical indicator of an ongoing epidemic, particularly in high-burden regions such as Liangshan Yi Autonomous Prefecture in China, where HIV prevalence exceeds 1% in 4 key counties (Butuo, Zhaojue, Meigu, and Yuexi). Identifying spatial clusters and drivers of recent infections is essential for implementing targeted interventions. Despite advancements in geospatial analyses of HIV prevalence, studies identifying drivers of incident HIV clustering remain limited, especially in low-resource settings.

OBJECTIVE: This study aims to identify spatial clusters of recent HIV infections and investigate potential driving factors in 4 key counties of the Liangshan Yi Autonomous Prefecture to inform targeted intervention strategies.

METHODS: From November 2017 to June 2018, we identified 246 (4.42%) recent HIV infection cases from 5555 newly diagnosed cases through expanded testing of the whole population in 4 key counties of Liangshan Yi Autonomous Prefecture. Recent infection cases were confirmed using limiting antigen avidity enzyme immunoassays or documented seroconversion within 6 months. The spatial distribution of incident HIV infection cases was analyzed using kernel density. Potential drivers, including population density, HIV prevalence, elevation, nighttime light index, urban proximity, and antiretroviral therapy (ART) coverage, were analyzed. The spatial lag regression model was used to identify factors associated with clustering of recent infection cases. The Geodetector q-statistic was used to quantify nonlinear interactive effects among these factors.

RESULTS: Significant spatial autocorrelation was observed in the distribution of recent HIV cases (Moran I=0.11; P<.01). Six spatial clusters were identified, and all were located near urban centers or major roads. Furthermore, 5 factors were identified by the spatial lag regression model as being significantly correlated with the clustering of recent HIV infection cases, including population density (β=0.59; P<.001), HIV prevalence (β=0.02; P<.001), distance to local urban area (β=-3.10; P=.01), SD of elevation (β=-0.15; P=.02), and ART coverage rate (β=183.80; P<.01). Geodetector analysis revealed strong interactive effects among these 5 factors, with population density and HIV prevalence exhibiting the largest interactive effect (q=0.69).

CONCLUSIONS: This study reveals that besides HIV prevalence, urbanization-related factors (population density and proximity to urban area) and transportation accessibility drive incident HIV clustering in Liangshan Yi Autonomous Prefecture. Paradoxically, higher ART coverage was associated with increased transmission, suggesting the need for integrated prevention strategies beyond ART expansion. Furthermore, the township-level geospatial approach provides a valuable model for pinpointing transmission hot spots and tailoring interventions in high-burden regions globally.

PMID:40957018 | DOI:10.2196/75291

JMIR Form Res. 2025 Sep 16;9:e72331. doi: 10.2196/72331.

ABSTRACT

BACKGROUND: Digital health measurement offers opportunities to address several primary health care challenges, but health care professionals encounter significant implementation barriers. Therefore, resources need to be developed to facilitate the integration of digital health measurement into daily practice.

OBJECTIVE: We aim to identify the most appropriate format and content for an intervention to support primary health care professionals in adopting digital health measurement. In addition, we describe and reflect on the development process.

METHODS: We used a participatory action research approach as well as user-centered design principles. A total of 19 primary health care professionals from 4 disciplines-physical therapy, occupational therapy, speech and language therapy, and general practitioner practice assistance-participated in intervention development as end users. External experts were consulted to broaden perspectives. Data were collected across 3 iterative stages (concept, design, and testing and trials) between January 2022 and December 2023 during cocreative meetings, individual interviews, focus group discussions, usability testing, and prototype use in daily practice. Data were analyzed using content analysis and descriptive statistics.

RESULTS: A web-based, stepwise intervention combining theoretical information, practical aids, examples, and experiences proved most suitable. Key features were concise content, intuitive and attractive graphic design, and flexible navigation and functionalities. Iterative improvements led to an increase in usability ratings from “okay” to “good to excellent.”

CONCLUSIONS: Different health care disciplines benefit from similar support strategies; yet, this requires a careful balancing of intervention design and content. Combining participatory action research and user-centered design principles was useful to tailor the intervention to end users’ daily routines. The described development process offers a replicable framework for creating support strategies for digital health measurement in various health care settings.

PMID:40957009 | DOI:10.2196/72331

J Med Internet Res. 2025 Sep 16;27:e66667. doi: 10.2196/66667.

ABSTRACT

BACKGROUND: Determining effective nurse staffing levels is crucial for ensuring quality patient care and operational efficiency within hospitals. Traditional workload prediction methods often rely on professional judgment or simple volume-based approaches, which can be inaccurate. Machine learning offers a promising avenue for more data-driven and precise predictions, by using historical nursing workload data to forecast future patient care requirements, which could help with staff planning while also improving patient outcomes and nurse well-being.

OBJECTIVE: This methodological study aimed to use nursing activity data, specifically LEP (Leistungserfassung in der Pflege; “documentation of nursing activities”), to predict the future workload requirements using machine learning techniques.

METHODS: We conducted a retrospective observational study at the University Hospital of Zürich, using nursing workload data for inpatients across eight wards, collected between 2017 and 2021. Data were transformed to represent nursing workload per ward and shift, with 3 shifts per day. Variables used in modeling included historical workload trends, patient characteristics, and upcoming operations. Machine learning models, including linear regression variants and tree-based methods (Random Forest and XGBoost), were trained and tested on this dataset to predict workload 72 hours in advance, on a shift-by-shift basis. Model performance was assessed using mean absolute error and mean absolute percentage error, and results were compared against a baseline of assuming no change in workload from the time of prediction. Prediction accuracy was further evaluated by categorizing future workload changes into decreased, similar, or increased workload relative to current shift levels.

RESULTS: Our findings demonstrate that machine learning models consistently outperform the baseline across all wards. The best-performing model was the lasso regression model, which achieved an average improvement in accuracy of 25.0% compared to the baseline. When used to predict upcoming changes in workload levels, the model achieved strong classification performance, giving an average area under the receiver operating characteristic curve of 0.79 and precision values between 66.2% and 75.3%. Crucially, the model severely misclassified-predicting an upcoming increase as a decrease, and vice versa-in just 0.17% of cases, highlighting potential reliability for using the model in practice. Key variables identified as important for predictions include historical shift workload averages and overall ward workload trends.

CONCLUSIONS: This study suggests the potential of machine learning to enhance nurse workload prediction, while highlighting the need for refinement. Limitations due to potential discrepancies between recorded nursing activities and the actual workload highlight the need for further investigation into data quality. To maximize impact, future research should focus on: (1) using more diverse data, (2) more advanced machine learning architecture that performs time-series modeling, (3) addressing data quality concerns, and (4) conducting controlled trials for real-world evaluation.

PMID:40956986 | DOI:10.2196/66667