Category: Statistics

Clin Transl Gastroenterol. 2025 Sep 10. doi: 10.14309/ctg.0000000000000919. Online ahead of print.

ABSTRACT

BACKGROUND: Severe acute pancreatitis (SAP) is a life-threatening condition requiring early risk stratification. While the Bedside Index for Severity in Acute Pancreatitis (BISAP) is widely used, its reliance on complex parameters limits its applicability in resource-constrained settings. This study introduces a decision tree model based on Classification and Regression Tree (CART) analysis, utilizing Neutrophil-to-Lymphocyte Ratio (NLR) and C-reactive Protein (CRP), as a simpler alternative for early SAP prediction.

METHODS: In a prospective cohort of 340 patients at National Hospital, Vietnam (November 2022-September 2023), NLR, CRP, and BISAP scores were assessed upon admission. CART analysis was used to develop a decision tree, and model performance was compared with BISAP using receiver operating characteristic (ROC) curves, decision curve analysis (DCA).

RESULTS: The CART model identified NLR ≥11.4 and CRP ≥173.3 mg/L as optimal thresholds for SAP prediction. The model achieved an area under the curve (AUC) 0.866 in the validation cohort, statistically comparable to BISAP (AUC = 0.900, p = 0.286). The model demonstrated high sensitivity (90.9%), specificity (84.5%), and accuracy (86.25%), confirming its robustness. DCA highlighted similar clinical benefits with BISAP, but the CART-based model offered greater simplicity, making it ideal for resource-limited settings.

CONCLUSION: The CART-derived decision tree using NLR and CRP provides an accessible and reliable tool for early SAP prediction. With performance comparable to BISAP but requiring fewer resources, this model supports rapid, evidence-based decision-making in clinical practice.

PMID:40924813 | DOI:10.14309/ctg.0000000000000919

PLoS One. 2025 Sep 9;20(9):e0331243. doi: 10.1371/journal.pone.0331243. eCollection 2025.

ABSTRACT

This research explores the dynamical properties and solutions of actin filaments, which serve as electrical conduits for ion transport along their lengths. Utilizing the Lie symmetry approach, we identify symmetry reductions that simplify the governing equation by lowering its dimensionality. This process leads to the formulation of a second-order differential equation, which, upon applying a Galilean transformation, is further converted into a system of first-order differential equations. Additionally, we investigate the bifurcation structure and sensitivity of the proposed dynamical system. When subjected to an external force, the system exhibits quasi-periodic behavior, which is detected using chaos analysis tools. Sensitivity analysis is also performed on the unperturbed system under varying initial conditions. Moreover, we establish the conservation laws associated with the equation and conduct a stability analysis of the model. Employing the tanh method, we derive exact solutions and visualize them through 3D and 2D graphical representations to gain deeper insights. These findings offer new perspectives on the studied equation and significantly contribute to the understanding of nonlinear wave dynamics.

PMID:40924812 | DOI:10.1371/journal.pone.0331243

Physiol Int. 2025 Sep 9. doi: 10.1556/2060.2025.00640. Online ahead of print.

ABSTRACT

PURPOSE: Contribution of the gastrocnemii muscles to ankle moment is influenced by the knee joint position because they span the knee and the ankle joint as well. However, limited information is available on the effect of knee joint position on soleus activation under dynamic plantarflexion, hence the aim of this study was to investigate if soleus have a compensatory strategy in fascicle behavior or EMG activity during knee flexed plantarflexion in order to reduce the magnitude of the decrement in ankle moment.

EQUIPMENT AND METHODS: Isokinetic dynamometry with EMG and ultrasound measurements was used to estimate medial gastrocnemius and soleus behavior during knee flexed and extended plantarflexions using three angular velocities. Seventeen healthy males were participated in this study.

RESULTS: Flexed knee plantarflexions resulted in lower peak ankle moments at all ankle angular velocities by 18% (P = 0.1062) at 30°∙s-1, 44% (P < 0.001) at 60°∙s-1 and by 18% (P = 0.0001) at 120°∙s-1. Soleus showed significantly higher EMG activity during knee flexed plantarflexion at 30°∙s-1 (P = 0.0094) and 60°∙s-1 (P = 0.0142). The magnitude of mean shortening of the medial gastrocnemius and soleus show statistically significant difference between knee flexed and knee extended plantarflexion at any contraction velocity.

CONCLUSIONS: Soleus may perform a compensatory EMG activity in knee flexed plantarflexions possibly to counteract the reduced contribution of gastrocnemius to ankle moment at low angular velocity contractions.

PMID:40924490 | DOI:10.1556/2060.2025.00640

JAMA Netw Open. 2025 Sep 2;8(9):e2530730. doi: 10.1001/jamanetworkopen.2025.30730.

ABSTRACT

IMPORTANCE: Long COVID (ie, post-COVID-19 condition) is a substantial public health concern, and its association with health-related social needs, such as food insecurity, remains poorly understood. Identifying modifiable risk factors like food insecurity and interventions like food assistance programs is critical for reducing the health burden of long COVID.

OBJECTIVE: To investigate the association of food insecurity with long COVID and to assess the modifying factors of Supplemental Nutrition Assistance Program (SNAP) participation and employment status.

DESIGN, SETTING, AND PARTICIPANTS: This retrospective, cross-sectional survey study used data from the 2022 to 2023 National Health Interview Survey. Respondents aged 18 years and older who reported prior COVID-19 infection and responded to questions on food insecurity and long COVID were included.

EXPOSURE: Food insecurity, categorized as food secure or food insecure.

MAIN OUTCOMES AND MEASURES: The primary outcome was current long COVID, defined as symptoms lasting 3 or more months after initial COVID-19 infection persisting to time of interview. The secondary outcome was long COVID recovery, indicating history of long COVID without current symptoms. Food insecurity was measured using the validated 10-item National Center for Health Statistics food insecurity scale. Odds ratios (ORs) and 95% CIs for the association of food insecurity with long COVID were calculated using simple and multiple logistic regression.

RESULTS: The study enrolled 21 631 participants (1255 female [weighted percentage, 53%]; 5058 aged 65 years or older [weighted percentage, 16%]), including 19 824 with food security and 1807 with food insecurity. In total, 288 respondents with food insecurity (weighted percentage, 15%) reported current long COVID compared with 1547 (weighted percentage, 7%) without food insecurity. Food insecurity was positively associated with current long COVID (adjusted OR, 1.73; 95% CI, 1.39-2.15) and negatively associated with recovery among adults with prior long COVID (adjusted OR, 0.70; 95% CI, 0.54-0.92). SNAP participation (P for interaction = .04) and unemployment (P for interaction = .04) significantly modified these associations.

CONCLUSIONS AND RELEVANCE: In this survey study of US adults with prior COVID-19 infection, food insecurity was associated with greater odds of long COVID and lower odds of recovery, with SNAP participation and unemployment mitigating these associations. These findings suggest that expanding SNAP eligibility, simplifying enrollment processes, and increasing awareness of food assistance programs may reduce the burden of food insecurity and long COVID and further emphasize the importance of addressing health-related social needs in chronic disease prevention and management.

PMID:40924424 | DOI:10.1001/jamanetworkopen.2025.30730

JAMA Netw Open. 2025 Sep 2;8(9):e2530787. doi: 10.1001/jamanetworkopen.2025.30787.

ABSTRACT

IMPORTANCE: As obesity rates rise in the US, managing associated metabolic comorbidities presents a growing burden to the health care system. While bariatric surgery has shown promise in mitigating established metabolic conditions, no large studies have quantified the risk of developing major obesity-related comorbidities after bariatric surgery.

OBJECTIVE: To identify common metabolic phenotypes for patients eligible for bariatric surgery and to estimate crude and adjusted incidence rates of additional metabolic comorbidities associated with bariatric surgery compared with weight management program (WMP) alone.

DESIGN, SETTING, AND PARTICIPANTS: This retrospective, multicenter cohort study used data from the Veterans Health Administration (VHA) Corporate Data Warehouse, which incorporates data from 128 VHA centers. Participants were adults 18 years or older with a body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) of 30 or higher and at least 1 of 5 metabolic comorbidities or with a BMI of 35 or higher who underwent bariatric surgery (Roux-en-Y gastric bypass or sleeve gastrectomy) or enrolled in the WMP (called MOVE!) between January 1, 2008, and December 31, 2023. Patients with all 5 comorbidities at baseline or missing key data were excluded.

EXPOSURE: Bariatric surgery vs WMP.

MAIN OUTCOME AND MEASURES: Incident diagnosis of any of 5 metabolic comorbidities: type 2 diabetes (T2D), hypertension, hyperlipidemia, obstructive sleep apnea (OSA), and metabolic dysfunction-associated steatotic liver disease (MASLD).

RESULTS: Among 269 470 veterans, 263 657 were enrolled in the WMP and 5813 underwent bariatric surgery. Patients included 232 196 males (87.1%) and had a median (IQR) age of 57 (47-64) years. Median (IQR) follow-up time was 112.9 (79.5-145.4) months. At 5 years, incidence rates per 1000 person-years were 8.89 for hypertension, 9.67 for hyperlipidemia, 4.29 for T2D, 3.99 for OSA, and 2.44 for MASLD in the WMP group. For the bariatric surgery group, incidence rates per 1000 person-years were 3.35 for hypertension, 4.85 for hyperlipidemia, 1.06 for T2D, 3.43 for OSA, and 2.01 for MASLD. Bariatric surgery was associated with a statistically significant lower risk of incident T2D (79.2% lower; hazard ratio [HR], 0.21 [95% CI, 0.18-0.26]), hypertension (58.8% lower; HR, 0.41 [95% CI, 0.33-0.51]), hyperlipidemia (50.5% lower; HR, 0.49 [95% CI, 0.42-0.58]), OSA (56.9% lower; HR, 0.43 [95% CI, 0.35-0.52]), and MASLD (40.4% lower; HR, 0.60 [95% CI, 0.49-0.73]) compared with the WMP. Results were consistent in a subgroup analysis of only female veterans.

CONCLUSIONS AND RELEVANCE: In this cohort study, bariatric surgery was associated with a significantly lower risk of developing major metabolic comorbidities compared with the medical WMP. This finding supports the relevance of bariatric surgery as a durable approach for obesity-related risk mitigation.

PMID:40924423 | DOI:10.1001/jamanetworkopen.2025.30787

JAMA Netw Open. 2025 Sep 2;8(9):e2530980. doi: 10.1001/jamanetworkopen.2025.30980.

ABSTRACT

IMPORTANCE: Patients with advanced cancer frequently receive broad-spectrum antibiotics, but changing use patterns across the end-of-life trajectory remain poorly understood.

OBJECTIVE: To describe the patterns of broad-spectrum antibiotic use across defined end-of-life intervals in patients with advanced cancer.

DESIGN, SETTING, AND PARTICIPANTS: This nationwide, population-based, retrospective cohort study used data from the South Korean National Health Insurance Service database to examine broad-spectrum antibiotic use among patients with advanced cancer who died between July 1, 2002, and December 31, 2021. Data extraction and analysis were conducted between September 2023 and August 2024.

EXPOSURE: A diagnosis of lung cancer, liver cancer, stomach cancer, colorectal cancer, pancreatic cancer, prostate cancer, gallbladder and biliary tract cancer, breast cancer, non-Hodgkin lymphoma, leukemia, or multiple myeloma.

MAIN OUTCOMES AND MEASURES: The use of broad-spectrum antibiotics (ie, antipseudomonal β-lactams, carbapenems, or glycopeptides) was evaluated using 2 metrics according to end-of-life trajectory: (1) prescription proportion (percentage of patients receiving antibiotics) and (2) consumption amount (days of therapy per 1000 patient-days). The end-of-life trajectory was divided into 5 intervals: T1 (6 months to 3 months before death), T2 (3 months to 1 month before death), T3 (1 month to 2 weeks before death), T4 (2 weeks to 1 week before death), and T5 (final week before death). Logistic regression was performed to calculate odds ratios and 95% CIs for antibiotic prescription proportion without adjustment for multiple comparisons, and Poisson regression was used to calculate adjusted relative risks.

RESULTS: Among the 515 366 decedents included, the mean (SD) age was 68.8 (11.7) years, and 347 327 (67.4%) were male. A total of 483 405 patients (93.8%) had solid tumors, with lung cancer (122 142 patients [23.7%]) being the most common type. Overall, 288 151 patients (55.9%) received broad-spectrum antibiotics during the last 6 months of life. The proportion of patients receiving broad-spectrum antibiotics peaked during T2, with 144 920 (28.1%) receiving at least 1 dose, and declined to 68 564 (13.3%) during T5. In contrast, total consumption peaked during T3, reaching 190.0 days of therapy per 1000 patient-days. These patterns were consistent across antibiotic classes and cancer types. During the last week of life, patients with leukemia had the highest exposure to broad-spectrum antibiotics compared with those with lung cancer, both for prescription proportions (crude odds ratio, 1.50; 95% CI, 1.43-1.58) and total consumption (adjusted relative risk, 1.21; 95% CI, 1.19-1.23).

CONCLUSIONS AND RELEVANCE: In this cohort study of patients with advanced cancer, broad-spectrum antibiotic use increased from 3 months to 2 weeks before death, suggesting that this may be a key period for optimizing use and aligning care with patient goals.

PMID:40924422 | DOI:10.1001/jamanetworkopen.2025.30980

JAMA Netw Open. 2025 Sep 2;8(9):e2531166. doi: 10.1001/jamanetworkopen.2025.31166.

ABSTRACT

IMPORTANCE: Caregivers of community-dwelling older adults play a protective role in emergency department (ED) care transitions. When the demands of caregiving result in caregiver burden, ED returns can ensue.

OBJECTIVE: To develop models describing whether caregiver burden is associated with ED revisits and hospital admissions up to 30 days after discharge from an initial ED visit.

DESIGN, SETTING, AND PARTICIPANTS: This prospective cohort study nested within the LEARNING WISDOM longitudinal cohort study included dyads of community-dwelling patients aged 65 years or older and their caregivers. Included patients were triaged to an observation unit stretcher on their index visit to 1 of 4 EDs within an integrated health multisite organization of 4 acute care hospitals in Québec, Canada, between January 1, 2019, and December 21, 2021, and underwent a transition of care when discharged back to the community. Analyses were conducted in May 2024.

EXPOSURE: Caregiver burden, collected using the brief 12-item Quebec French version of the Zarit Burden Interview (ZBI; score range, 0-48, with higher scores indicating higher burden).

MAIN OUTCOMES AND MEASURES: Revisits to the ED, defined as a return to any ED in the 4-hospital network, within 3, 7, or 30 days of the index visit and return visits to the ED resulting in hospitalization within 30 days of the index visit. Moderation of outcomes by wave of the COVID-19 pandemic at the index visit was also assessed.

RESULTS: Among 1409 caregiver-patient dyads, 711 patients (50.5%) and 980 caregivers (69.6%) were women; mean (SD) age was 77.06 (7.39) years for patients and 63.87 (12.04) years for caregivers. The mean (SD) ZBI score was 7.33 (7.11). Caregivers were most often spouses of patients (667 [48.0%]) or children of patients (534 [37.9%]). Among all patients, 75 (5.3%) returned to the ED within 3 days, 133 (9.4%) returned within 7 days, 292 (20.7%) returned within 30 days, and 88 (6.2%) were admitted to the hospital within 30 days. Each point increase on the ZBI scale was associated with an increase in the odds of a 30-day revisit to the ED (odds ratio [OR], 1.03; 95% CI, 1.00-1.05; P = .03), but associations were not found in models with shorter time windows (3 days: OR, 1.01; 95% CI, 0.98-1.04; P = .69; 7 days: OR, 1.01; 95% CI, 0.98-1.03; P = .55) or for revisits with hospital admissions (OR, 1.02; 95% CI, 0.99-1.05; P = .24). Associations between ZBI scores and 30-day ED revisits may have been moderated by the COVID-19 pandemic waves: the first interwave period (between waves 1 and 2) reversed the association (OR, 0.89; 95% CI, 0.78-0.97).

CONCLUSIONS AND RELEVANCE: The findings suggest caregiver burden may be associated with ED revisits within 30 days of discharge from an initial ED visit among community-dwelling older adults. Future studies could enhance the management of ED revisits by demonstrating the longitudinal impact of caregiver burden on ED use in older adults.

PMID:40924417 | DOI:10.1001/jamanetworkopen.2025.31166

Age Ageing. 2025 Aug 29;54(9):afaf247. doi: 10.1093/ageing/afaf247.

ABSTRACT

BACKGROUND: Little is known about how ambulatory care sensitive condition (ACSC)-related readmissions can be reduced in acute care settings.

OBJECTIVE: This study examined the association between transitional care for hospitalised older patients with ACSC and ACSC-related readmissions.

METHODS: This retrospective observational cohort study included patients aged 65 years and older admitted with ACSC as the primary diagnosis from 1 April 2022 to 31 January 2023, using linked data from the Diagnosis Procedure Combination and the medical functions of the hospital beds database. The primary outcomes were cumulative readmissions within 1-7, 1-14, 1-21, 1-30 and 1-60 days, analysed using inverse probability treatment weighting regression models.

RESULTS: Among 85 582 patients from 711 hospitals, 39 916 (46.6%) were female, with a median age of 82 years (interquartile range: 75-88); 57 127 (66.8%) patients received transitional care. The overall readmission rates were 2.9%, 6.0%, 8.7%, 11.4% and 17.5% among total hospitalisations within 7, 14, 21, 30 and 60 days, respectively. Overall, transitional care was associated with reduced odds of ACSC-related readmission, with odds ratios ranging from 0.72 (95% CI: 0.65-0.78) within 7 days to 0.91 (95% CI: 0.87-0.95) within 60 days. The association between transitional care and readmission varied by ACSC category. In chronic ACSC, the association was strongest for 7-day readmission, followed by a downward trend. In acute and vaccine-preventable ACSC, the association was strongest for 7-day readmission but levelled off after 21 days.

CONCLUSIONS: Transitional care in acute care hospitals may be associated with a reduced risk of early readmissions due to ACSC when older patients are hospitalised.

PMID:40924407 | DOI:10.1093/ageing/afaf247

Patient. 2025 Sep 9. doi: 10.1007/s40271-025-00768-0. Online ahead of print.

ABSTRACT

BACKGROUND: Migraine care is often suboptimal owing to undertreatment, variation in clinical outcomes and administration methods among existing treatments, and between- and within-individual heterogeneity in the clinical course of migraine. In response to these challenges, preference studies have been increasingly conducted to inform treatment decision-making and development. However, gaps remain in understanding how treatment preferences have been assessed across different migraine studies.

OBJECTIVE: The aim was to synthesize existing evidence to inform the design and conduct of future preference migraine research. This review examined treatment attributes included in preference studies, focusing on how attributes were developed, framed, and presented; how their values were analyzed and reported; and whether these values differed by respondent characteristics.

METHODS: A systematic review protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO) (CRD42025614690). Embase, MEDLINE, and the Cochrane Library were searched for relevant stated preference studies on migraine treatments (October 2024). Two researchers independently screened studies, and data were extracted using a predefined template. Extracted information included study characteristics, methods for attribute and instrument development, choice task design, attribute framing, and analytical approaches. Narrative synthesis and descriptive statistics were used to summarize findings. Attribute importance was assessed by deriving relative rankings of attributes from marginal utilities or importance scores across studies.

RESULTS: Overall, 18 studies were reviewed from the 186 that were screened. Stated preference methods comprised discrete choice experiment (n = 12), conjoint analysis (n = 1), contingent valuation method (n = 3), thresholding (n = 1), and time trade-off (n = 1). In total, 13 studies reported their attribute development methods, using literature review only (n = 2), expert consultation only (n = 1), and multi-method approaches combining literature reviews with qualitative research and/or expert or payer consultation (n = 10). In addition, 17 studies included at least 1 benefit attribute, resulting in 26 unique attributes grouped into seven overarching concepts. Risk attributes were included in 11 studies, with injection site reactions (n = 5), gastrointestinal effects (n = 4), and cognitive effects (n = 3) as the most common adverse events. Administration-related attributes appeared in ten studies, with mode and/or frequency of administration being the most common (n = 10). Eight studies used visual aids to illustrate attributes. Preference heterogeneity was explored in 14 studies, primarily on the basis of sex (n = 9), monthly migraine days (n = 8), and treatment experience (n = 7).

CONCLUSIONS: This review reveals substantial variation in how treatment attributes were selected, framed, and analyzed across studies. Greater methodological consistency in attribute development, framing, and reporting, along with more robust exploration of preference heterogeneity, is needed to enhance the comparability, validity, and application of future preference research in migraine care.

PMID:40924386 | DOI:10.1007/s40271-025-00768-0

Curr Atheroscler Rep. 2025 Sep 9;27(1):88. doi: 10.1007/s11883-025-01339-2.

ABSTRACT

PURPOSE OF REVIEW: This review examines cardiovascular disease (CVD) risk prediction models relevant to older adults, a rapidly expanding population with elevated CVD risk. It discusses model characteristics, performance metrics, and clinical implications.

RECENT FINDINGS: Some models have been developed specifically for older adults, while several others consider a broader age range, including some older individuals. These models vary in terms of predictors, outcomes, horizon, and statistical approaches, with some accounting for competing risks and considering age-predictor interactions. Discrimination is generally acceptable and more modest in older versus younger individuals. Calibration shows great variation across populations. Accurate CVD risk prediction is essential to guide individualized prevention strategies and support shared decision-making in older adults. CVD risk prediction in this population is challenged by age-related CVD risk heterogeneity, elevated competing risk due to non-CVD mortality, and comorbidities. Further refinement by incorporating geriatric-specific factors may help to enhance discrimination.

PMID:40924377 | DOI:10.1007/s11883-025-01339-2