Category: Statistics

Ir J Med Sci. 2025 Sep 8. doi: 10.1007/s11845-025-04034-y. Online ahead of print.

ABSTRACT

INTRODUCTION: Information on tertiary hyperparathyroidism (THPTH) among chronic kidney disease (CKD) patients on haemodialysis in developing countries such as India is limited, and the mortality among them remains a query.

MATERIALS AND METHODS: This was a prospective cohort study conducted in at a tertiary care centre from June 2017 to June 2022. The index of suspicion for tertiary hyperparathyroidism was when investigations revealed high serum calcium and high alkaline phosphatase along with new onset of body aches, joint pains, and difficulty in walking. Patients, with above clinical features, were considered for 99 m Tc-Sestamibi scan and high-resolution ultrasound of the neck, when serum parathormone was > 600 pg/mL. Those patients diagnosed with tertiary hyperparathyroidism were followed up for 5 years.

RESULTS: The incidence of tertiary hyperparathyroidism among CKD patients was 13.4%. The mean age of CKD stage 5 patients with tertiary hyperparathyroidism was 55.17 ± 11.1 years. The observation from our study was the mean survival time among patients who underwent parathyroidectomy and among patients who received cinacalcet was almost similar, whereas the mean survival time among patients who received phosphate binders was lower. However, the survival rate among patients on cinacalcet and who underwent parathyroidectomy were not statistically significant.

DISCUSSION: There were no cross-sectional studies on prevalence of tertiary hyperparathyroidism in India as per our knowledge, although the prospective design, large sample size, PTH stratification, and frequent measurements of a comprehensive panel of mineral metabolites are strengths of the current study.

PMID:40920313 | DOI:10.1007/s11845-025-04034-y

Med Biol Eng Comput. 2025 Sep 8. doi: 10.1007/s11517-025-03436-5. Online ahead of print.

ABSTRACT

Fetal standard plane detection is essential in prenatal care, enabling accurate assessment of fetal development and early identification of potential anomalies. Despite significant advancements in machine learning (ML) in this domain, its integration into clinical workflows remains limited-primarily due to the lack of standardized, end-to-end operational frameworks. To address this gap, we introduce FetalMLOps, the first comprehensive MLOps framework specifically designed for fetal ultrasound imaging. Our approach adopts a ten-step MLOps methodology that covers the entire ML lifecycle, with each phase meticulously adapted to clinical needs. From defining the clinical objective to curating and annotating fetal US datasets, every step ensures alignment with real-world medical practice. ETL (extract, transform, load) processes are developed to standardize, anonymize, and harmonize inputs, enhancing data quality. Model development prioritizes architectures that balance accuracy and efficiency, using clinically relevant evaluation metrics to guide selection. The best-performing model is deployed via a RESTful API, following MLOps best practices for continuous integration, delivery, and performance monitoring. Crucially, the framework embeds principles of explainability and environmental sustainability, promoting ethical, transparent, and responsible AI. By operationalizing ML models within a clinically meaningful pipeline, FetalMLOps bridges the gap between algorithmic innovation and real-world application, setting a precedent for trustworthy and scalable AI adoption in prenatal care.

PMID:40920305 | DOI:10.1007/s11517-025-03436-5

J Eval Clin Pract. 2025 Sep;31(6):e70257. doi: 10.1111/jep.70257.

ABSTRACT

AIMS AND OBJECTIVE: The field of medical statistics has experienced significant advancements driven by integrating innovative statistical methodologies. This study aims to conduct a comprehensive analysis to explore current trends, influential research areas, and future directions in medical statistics.

METHODS: This paper maps the evolution of statistical methods used in medical research based on 4,919 relevant publications retrieved from the Web of Science. High-frequency keywords and citation metrics were analyzed to identify research hotspots. A dual-map overlay and document co-citation analysis were performed using CiteSpace to uncover thematic clusters and track knowledge flow between disciplines. Additionally, network metrics, such as betweenness centrality and sigma, were employed to quantify the influence and novelty of publications.

RESULTS: Results identified a strong interdisciplinary exchange between medical statistics and fields such as health, nursing, molecular biology, and computer science, with clinical trials, survival analysis, and predictive modeling emerging as central themes. The influence of artificial intelligence (AI), machine learning (ML), and deep learning (DL) is growing substantially, particularly in areas such as diagnostic imaging, epidemiology, and treatment prediction, highlighting a shift towards more complex, data-driven methodologies. While traditional statistical techniques, such as survival analysis and regression, remain vital, emerging technologies are reshaping research approaches, fostering collaboration, and advancing the field’s capabilities.

CONCLUSION: Future research will likely focus on overcoming challenges related to data privacy, ethical considerations, and the need for continued biostatistics education in healthcare. This study offers a roadmap for ongoing research and highlights opportunities for future interdisciplinary collaborations to address the complexities of modern medical data analysis. This scientometrics study reveals the evolution of statistical methods used in medical research over time, evaluates frequently cited models and thematic changes, and provides implications that can enhance evidence-based decision-making processes regarding methodological choices that guide contemporary clinical practice.

PMID:40916916 | DOI:10.1111/jep.70257

J Magn Reson Imaging. 2025 Sep 8. doi: 10.1002/jmri.70105. Online ahead of print.

ABSTRACT

BACKGROUND: Cerebrovascular reactivity reflects changes in cerebral blood flow in response to an acute stimulus and is reflective of the brain’s ability to match blood flow to demand. Functional MRI with a breath-hold task can be used to elicit this vasoactive response, but data validity hinges on subject compliance. Determining breath-hold compliance often requires external monitoring equipment.

PURPOSE: To develop a non-invasive and data-driven quality filter for breath-hold compliance using only measurements of head motion during imaging.

STUDY TYPE: Prospective cohort.

PARTICIPANTS: Longitudinal data from healthy middle-aged subjects enrolled in the Coronary Artery Risk Development in Young Adults Brain MRI Study, N = 1141, 47.1% female.

FIELD STRENGTH/SEQUENCE: 3.0 Tesla gradient-echo MRI.

ASSESSMENT: Manual labelling of respiratory belt monitored data was used to determine breath hold compliance during MRI scan. A model to estimate the probability of non-compliance with the breath hold task was developed using measures of head motion. The model’s ability to identify scans in which the participant was not performing the breath hold were summarized using performance metrics including sensitivity, specificity, recall, and F1 score. The model was applied to additional unmarked data to assess effects on population measures of CVR.

STATISTICAL TESTS: Sensitivity analysis revealed exclusion of non-compliant scans using the developed model did not affect median cerebrovascular reactivity (Median [q1, q3] = 1.32 [0.96, 1.71]) compared to using manual review of respiratory belt data (1.33 [1.02, 1.74]) while reducing interquartile range.

RESULTS: The final model based on a multi-layer perceptron machine learning classifier estimated non-compliance with an accuracy of 76.9% and an F1 score of 69.5%, indicating a moderate balance between precision and recall for the identification of scans in which the participant was not compliant.

DATA CONCLUSION: The developed model provides the probability of non-compliance with a breath-hold task, which could later be used as a quality filter or included in statistical analyses.

LEVEL OF EVIDENCE: 1: TECHNICAL EFFICACY: Stage 3.

PMID:40916903 | DOI:10.1002/jmri.70105

Zhong Nan Da Xue Xue Bao Yi Xue Ban. 2025 May 28;50(5):757-765. doi: 10.11817/j.issn.1672-7347.2025.240422.

ABSTRACT

OBJECTIVES: Intracranial aneurysm (IA) has an insidious onset, and once ruptured, it carries high rates of mortality and disability. Cardiometabolic factors may be associated with the formation and rupture of IA. This study aims to summarize the application of Mendelian randomization (MR) methods in research on cardiometabolic factors and IA, providing insights for further elucidation of IA etiology and pathogenesis.

METHODS: Literature about MR-based IA studies published up to February 21, 2024, was retrieved from PubMed, Embase, Web of Science, CNKI, and Wanfang. Two researchers independently performed literature screening, data extraction, and quality assessment. A narrative synthesis approach was used to conduct a qualitative systematic review of the included studies.

RESULTS: A total of 11 MR-based studies on IA published between 2017 to 2024 were included, of which 4 were rated as high quality. These studies investigated the associations between blood pressure, blood lipids, blood glucose, obesity-related indicators, and inflammatory cytokines with IA and its subtypes, though issues of duplication were noted. Four MR studies based on the same European population but using different instrumental variable selection criteria, as well as another MR study in a different European cohort, consistently identified blood pressure as a risk factor for IA and its subtypes. Findings for blood lipids, blood glucose, obesity-related indicators, and inflammatory cytokines were inconsistent across MR studies.

CONCLUSIONS: Blood pressure appears to increase the risk of IA and its subtypes. Associations between other cardiometabolic factors and IA/subtypes require further in-depth investigation. Given the inherent limitations of MR studies, causal inferences should be made cautiously in combination with other lines of evidence.

PMID:40916814 | DOI:10.11817/j.issn.1672-7347.2025.240422

J Laparoendosc Adv Surg Tech A. 2025 Sep 8. doi: 10.1177/10926429251376394. Online ahead of print.

ABSTRACT

Background: Robotic-assisted proctectomy (RAP) has been reportedly associated with lower rates of conversion to laparotomy than laparoscopy in several cohort studies. This st0udy aimed to assess the temporal trends in conversion from RAP to laparotomy stratified by patient and treatment-related factors. Methods: This retrospective observational study was undertaken to analyse the temporal trends in unplanned conversion from RAP to laparotomy. Changes in the rates of conversion over time were plotted as line graphs, and the significance of each trend was calculated with the Cochran-Armitage trend test. A case-control analysis of factors associated with conversion to open surgery was conducted. Results: The study included 23,644 patients (62.3% male, median age: 60 years). 1280 (5.4%) patients were converted to laparotomy. There was a significant linear trend of decreased conversion over time (3.9% in 2021 compared with 10.4% in 2010; P < .001). The reduction in conversion rates was significant in all patients except in patients <50 years (P = .838), Black patients (P = .358), patients with a Charlson comorbidity index score >1 (P = .053), patients with governmental insurance other than Medicaid and Medicare (P = .629), and patients undergoing abdominoperineal resection (APR) (P = .129) or pelvic exenteration (PE) (P = .326). The independent predictors for increased conversion were male sex, higher Charlson scores, community cancer programs, comprehensive community cancer programs, household income of <$63,000, tumors ≥5 cm, and PE. Conclusions: Unplanned conversion from RAP to laparotomy showed a linear trend of reduction over time, which was statistically significant except in young patients, Black patients, patients with significant comorbidities, and patients undergoing APR or PE.

PMID:40916784 | DOI:10.1177/10926429251376394

Anticancer Drugs. 2025 Sep 3. doi: 10.1097/CAD.0000000000001764. Online ahead of print.

ABSTRACT

Bortezomib resistance in multiple myeloma (MM) is a significant clinical challenge that limits the long-term effectiveness. Currently, there is a lack of reliable biomarkers to predict bortezomib resistance. Previous studies reported that several proteins regulate bortezomib resistance through targeting ubiquitin-proteasome pathways, including heat shock protein family A member 9 (HSPA9), dickkopf Wnt signaling pathway inhibitor 1 (DKK1), proteasome 26S subunit non-ATPase 14 (PSMD14), and tripartite motif containing 21 (TRIM21). In our study, we aimed to analyze the expression of these proteins in MM patients and evaluate their potential as biomarkers for bortezomib resistance. Our study enrolled 46 newly diagnosed MM patients (38 bortezomib-sensitive and eight bortezomib-resistant patients) and 52 healthy controls, and serum samples were collected from the patients before initial treatments. The levels of HSPA9, DKK1, PSMD14, and TRIM21 proteins in serum samples were measured using ELISA. The diagnostic power of HSPA9 protein for bortezomib resistance was evaluated through receiver operating characteristic curves combined with the area under curve (AUC). The correlation between HSPA9 protein and clinicopathological features was examined using the chi-square test, and Kaplan-Meier method and Cox regression analysis were applied to assess prognostic value. Compared with healthy controls, increased HSPA9 and DKK1, but decreased TRIM21 protein expression, were observed in serum samples from MM patients. There was no statistical difference in PSMD14 protein expression between the two groups. Notably, compared with bortezomib-sensitive patients, only HSPA9 protein was found to be upregulated in bortezomib-resistant patients, whereas no differences were found in the other proteins. Furthermore, the AUC of serum HSPA9 for differentiating MM patients from healthy controls was 0.906 [95% confidence interval (CI): 0.843-0.968]. And serum HSPA9 expression could effectively differentiate bortezomib-resistant MM patients from bortezomib-sensitive MM patients, with an AUC of 0.845 (95% CI: 0.734-0.957). In addition, elevated serum HSPA9 expression positively correlated with advanced International Staging System stage, increased β2-MG, abnormal immunoglobulin, and bortezomib resistance. Higher serum HSPA9 was linked to shorter overall survival rate and independently predicted poor prognosis. Our study demonstrated that elevated serum HSPA9 protein serves as a potential biomarker for bortezomib resistance and poor prognosis in MM patients.

PMID:40916774 | DOI:10.1097/CAD.0000000000001764

Emerg Med Australas. 2025 Oct;37(5):e70131. doi: 10.1111/1742-6723.70131.

ABSTRACT

BACKGROUND: Viscoelastic haemostatic assays (VHAs) guide transfusion decisions in bleeding patients. We assessed testing volumes, clinical indications and patient characteristics in a statewide population in Australia.

METHODS: This retrospective study included all patients who underwent rotational thromboelastometry (ROTEM) or thromboelastography (TEG) across Queensland Health hospitals (1 January 2019 to 15 April 2025), using data from AUSLAB, the statewide laboratory information system and surveyed all hospitals for VHA device availability.

RESULT: Of 39 VHA devices, 31 were transmitting to AUSLAB, with 43,220 tests performed in 21,178 patients, during 18,389 admissions and 6418 ED presentations; 92.0% were ROTEM (n = 39,776) and 8.0% TEG (n = 3444). Most tests occurred during inpatient care (n = 35,527, 82.2%) versus ED (n = 7693, 17.8%). Indications included trauma (n = 23,875, 55.2%), non-variceal gastrointestinal bleeding (n = 4238, 9.8%), obstetrics (n = 3307, 7.7%) and chronic liver disease (CLD) (n = 3853, 8.9%), including 1097 (2.5%) with variceal bleeding. Emergency department (ED) use increased overall (IRR 1.14; 95% CI 1.12-1.15), including trauma (IRR 1.15), CLD (1.16), variceal bleeding (1.12) and non-variceal bleeding (1.12) (all p < 0.001); obstetric use in ED did not change significantly (IRR 0.93; 95% CI 0.86-1.00). Inpatient use also increased (IRR 1.21; 95% CI 1.21-1.22), including trauma (IRR 1.22), CLD (1.16), variceal (1.10), non-variceal bleeding (1.17) and obstetrics (1.07) (all p < 0.001).

CONCLUSION: VHA use increased in both ED and inpatient settings, with prominent use in trauma and CLD. The results indicate growing recognition by clinicians of VHA’s value in guiding haemorrhage management. The need for a consistent, evidence-based approach to testing and interpretation of results is paramount.

PMID:40916731 | DOI:10.1111/1742-6723.70131

Am J Epidemiol. 2025 Sep 8:kwaf199. doi: 10.1093/aje/kwaf199. Online ahead of print.

ABSTRACT

Tree-based scan statistics (TBSS) are data mining methods that screen thousands of hierarchically related health outcomes to detect unsuspected adverse drug effects. TBSS traditionally analyze claims data with outcomes defined via diagnosis codes. TBSS have not been previously applied to rich clinical information in Electronic Health Records (EHR). We developed approaches for integrating EHR data in TBSS analyses, including outcomes derived from natural language processing (NLP) applied to clinical notes and laboratory results, related via multipath hierarchical structures. We consider four settings that sequentially add sources of outcomes to the TBSS tree: 1) diagnosis code, 2) NLP-derived outcomes, 3) binary outcomes from lab results, and 4) continuous lab results. In a comparative cohort study involving second-generation sulfonylureas (SUs) and dipeptidyl peptidase 4 (DPP-4) inhibitors among adults with type-2 diabetes, with an a priori expected signal of hypoglycemia, diagnosis code data showed no statistical alerts for inpatient or emergency department settings. Adding NLP-derived outcomes resulted in an alert for “Headaches” (p=0.047), a nonspecific symptom of hypoglycemia. Progressively adding binary and continuous lab results produced the same alert. Integrating EHR in TBSS can be useful for the detection of safety signals for further investigation.

PMID:40916726 | DOI:10.1093/aje/kwaf199

Am J Epidemiol. 2025 Sep 8:kwaf197. doi: 10.1093/aje/kwaf197. Online ahead of print.

ABSTRACT

In 2016, the NIH designated LGBTQ+ individuals (ie, lesbian, gay, bisexual, transgender, queer, and all sexual and gender minorities) as a health disparities population. The growing interest in studying the health of LGBTQ+ populations merits revisiting the methodological approaches researchers employ. We elucidate how researchers can identify appropriate adjustment sets for causal questions using directed acyclic graphs (DAGs). To illustrate these points, we simulated a simplified example using pregnancy loss as the outcome wherein we generate 1000 datasets with a sample size of 10 000 individuals. We motivate why covariates that are commonly used in LGBTQ+ health disparities research (eg, use of medically assisted reproduction) are mediators, not confounders, and how adjusting for these variables in causal research can induce bias by blocking part of the indirect effect of exposure on the outcome. Next, we illustrate the complexity of mediation analyses with social exposures due to mediator-outcome confounding induced by exposure and compare potential approaches. Then we demonstrate how collider stratification bias can arise from our sample recruitment and selection. Finally, we demonstrate how incorporating heterosexism (ie, stigma and discrimination) as an unobserved node in our DAG can guide decision-making on appropriate adjustment sets.

PMID:40916718 | DOI:10.1093/aje/kwaf197