Tag: statistics

Eur J Neurol. 2025 Jun;32(6):e70241. doi: 10.1111/ene.70241.

ABSTRACT

BACKGROUND: Motor neuron diseases (MND) are heterogeneous and complex neurodegenerative disorders. Biomarkers could facilitate early diagnosis, prognosis determination, and patient stratification. Among the most studied biomarkers are neurofilaments, with peripherin (PRPH), a specific type predominantly expressed in the peripheral nervous system, gaining attention. To date, no studies have evaluated PRPH in human plasma.

METHODS: Sandwich-ELISA was used to quantify plasma peripherin from 120 MND (100 ALS, 4 PMA, 15 PLS), 73 MND-mimics, and 38 healthy-controls (HCs). Plasma was collected at diagnosis or some months earlier. 41 ALS were evaluated longitudinally. ALSFRSr, MRC, spirometry, genetic tests, disease progression rate (PR), blood examinations, and neuropsychological tests were performed. Statistical analyses included Kruskal-Wallis, Mann-Whitney, Cox regression, and Kaplan-Meier curves.

RESULTS: Plasma PRPH levels differed significantly among groups (p < 0.0001), showing higher values in MND participants than MND mimics and HCs. Moreover, PRPH levels were elevated in PLS compared with HSP patients (p = 0.0001). Differences persisted after adjusting for age and sex. ROC curve demonstrated that PRPH discriminated MND from MND mimics (AUC = 0.85). Elevated PRPH correlated positively with ALSFRSr and lower motor neuron index, whereas inversely with disease progression rate. Higher PRPH levels at the beginning of the disease were associated with longer survival.

DISCUSSION: Plasma PRPH is raised in MND, particularly ALS, from the earliest stages, distinguishing MND from mimics and correlating with clinical parameters and survival. This suggests PRPH may reflect an endogenous response of lower motor neuron to injury. Further multicenter studies are required to refine the diagnostic and prognostic utility of PRPH in MND.

PMID:40476320 | DOI:10.1111/ene.70241

Br J Math Stat Psychol. 2025 Jun 6. doi: 10.1111/bmsp.12395. Online ahead of print.

ABSTRACT

The Elo Rating System which originates from competitive chess has been widely utilised in large-scale online educational applications where it is used for on-the-fly estimation of ability, item calibration, and adaptivity. In this paper, we aim to critically analyse the shortcomings of the Elo rating system in an educational context, shedding light on its measurement properties and when these may fall short in accurately capturing student abilities and item difficulties. In a simulation study, we look at the asymptotic properties of the Elo rating system. Our results show that the Elo ratings are generally not unbiased and their variances are context-dependent. Furthermore, in scenarios where items are selected adaptively based on the current ratings and the item difficulties are updated alongside the student abilities, the variance of the ratings across items and students artificially increases over time and as a result the ratings do not converge. We propose a solution to this problem which entails using two parallel chains of ratings which remove the dependence of item selection on the current errors in the ratings.

PMID:40476309 | DOI:10.1111/bmsp.12395

Stat Med. 2025 Jun;44(13-14):e70123. doi: 10.1002/sim.70123.

ABSTRACT

In longitudinal observational studies with time-varying confounders, the generalized computation algorithm formula (g-formula) is a principled tool to estimate the average causal effect of a treatment regimen. However, the standard non-iterative g-formula implementation requires specifying both the conditional distribution of the outcomes and the joint distribution of all time-varying covariates. This process can be cumbersome to implement and is prone to model misspecification bias. As an alternative, the iterative conditional expectation (ICE) g-formula estimator solely depends on a series of nested outcome regressions and avoids the need for specifying the full distribution of all time-varying covariates. This simplicity lends itself to the natural integration of flexible machine learning techniques to develop more robust average causal effect estimators with time-varying treatments. In this work, we introduce a Bayesian approach that includes parametric regressions and Bayesian Additive Regression Trees to flexibly model a series of outcome surfaces. We fit the ICE g-formula and develop a sampling algorithm to obtain samples from the posterior distribution of the final causal effect estimator. We illustrate the performance characteristics of the Bayesian ICE estimator and the associated variations via simulation studies and applications to two real world data examples.

PMID:40476299 | DOI:10.1002/sim.70123

Euro Surveill. 2025 Jun;30(22). doi: 10.2807/1560-7917.ES.2025.30.22.2400594.

ABSTRACT

BackgroundMost Toxoplasma gondii infections in humans are considered foodborne, but the relative importance of the various routes of infection is largely unknown. Consumption of green produce contaminated with T. gondii oocysts has been identified as a possible source.AimWe aimed to estimate the occurrence and prevalence of T. gondii oocysts in commercially available ready-to-eat (RTE) salad mixes in 10 European countries.MethodsA real-time PCR-based method for oocyst detection was developed and optimised by two laboratories and validated in an interlaboratory test. This detection method and a harmonised sampling strategy were applied in a multi-country study. Multivariable logistic regression was used to investigate risk factors for oocyst contamination of RTE salad.ResultsThe real-time PCR method had a detection limit of 10 oocysts per 30 g of salad. We collected 3,329 RTE salad samples (baby leaf and cut leaf mixes) from October 2021 to September 2022. The prevalence of T. gondii oocyst contamination was 4.1% (95% confidence interval (CI): 3.4-4.8%; n = 3,293). In multivariable regression analysis, winter season, sampling and packaging of salad in Northern Europe and production of salad in Western Europe were associated with detection of T. gondii, with no statistically significant differences between salad types.ConclusionWe estimated the prevalence of T. gondii oocysts in RTE leafy green salads using a validated and standardised procedure to assess the potential risk for human infection; highlighting the need to address this risk at each critical point of the salad production chain.

PMID:40476292 | DOI:10.2807/1560-7917.ES.2025.30.22.2400594

Euro Surveill. 2025 Jun;30(22). doi: 10.2807/1560-7917.ES.2025.30.22.2400530.

ABSTRACT

BackgroundDuring the COVID-19 pandemic, Santé publique France (SpF) published incidence (SpFi) rates based on census denominators. Denominators using cell phone connection (CPC) data can better reflect the population present and seasonal mobilities.AimGiven uncertainties regarding the actual number of Île-de-France (IdF) residents present in IdF during summer 2021, we aimed to better approximate true incidence rates from positive SARS-CoV-2 tests in IdF using CPC-derived population denominators.MethodThis longitudinal study used the daily number of positive tests (PCR and Ag) on IdF residents in IdF as the numerator and the estimated resident population present in IdF at midnight as the denominator. We computed the mean corrected incidence rate (MCIR) per moving week between 4 July and 9 September 2021.ResultsThe MCIR showed higher incidence rates than initially estimated, especially during August when residents had left IdF for the holidays. Incidence rates reached a peak on 16 August when the SpFi rate per moving week was 200.9 per 100,000 compared with 315.6 per 100,000 with the MCIR, representing a 57% increase.ConclusionUsing local SARS-CoV-2 testing data and real-time population denominators, we showed that indicators using non-geographically referenced test results and fixed population denominators that ignore seasonal mobility can significantly underestimate incidence rates in IdF. New data sources using CPC data provide the opportunity to calculate more accurate and dynamic incidence rates and to map epidemics more precisely and in real time.

PMID:40476291 | DOI:10.2807/1560-7917.ES.2025.30.22.2400530

J Radiosurg SBRT. 2025;9(3):207-213.

ABSTRACT

PURPOSE: To refer our experience with stereotactic body radiotherapy (SBRT) in patients with pancreatic cancer.

MATERIALS AND METHODS: 45 Patients with unresectable or locally recurrent pancreatic cancer after primary surgery, were submitted to SBRT. Toxicities were graded according to CTCAE version 5. Statistical analysis was performed by the Kaplan-Meier method.

RESULTS: The characteristics of the patients were median age 70 years (range, 46-84 years), median KPS 90% (range, 80-90%). Six patients had recurrent cancer after surgery, the other 39 patients were unresectable. Median radiation dose was 35 Gy (range, 27-40 Gy) delivered in 5 fractions. Simultaneous integrated boost with median dose of 35 Gy (range, 30-45 Gy) was given in 7 patients.After median follow-up of 10 months (range, 3-61 months) median local control was 10 months (range,5-15 months) and 49% (±8%) at 1-year. Median overall survival (OS) was 10 months (range,7-14 months), 38 % (±7%) at 1 year. Type of radiological response statistically significant influenced LC and OS, stage only LC in non-significant way. Clinical response was obtained in 12 of 36 (33%) cases. Median Numeric Rating Scale (NRS) was 7 (range, 4-8) before radiotherapy and 1 (range, 0-5) post SBRT. Acute G1-2 gastrointestinal toxicities were registered in 15% of patients, no late toxicities were found.

CONCLUSION: In our series we obtained a good local palliation with SBRT that is a safe and effective treatment option. Higher doses could be administered in selected patients to obtain better response to treatment that is correlated with LC and OS.

PMID:40476277 | PMC:PMC12136683

Cochrane Evid Synth Methods. 2023 Aug 9;1(6):e12023. doi: 10.1002/cesm.12023. eCollection 2023 Aug.

ABSTRACT

INTRODUCTION: While systematic reviews (SRs) are considered the highest form of evidence in the hierarchy, the quality and standard of reviews varies. Two quality assessment tools have been developed to assess the variation in such standards. This study compared the preference, validity, reliability, and applicability of using A Measurement Tool to Assess Systematic Reviews (AMSTAR-2) and the Risk of Bias in Systematic Reviews (ROBIS) for critically appraising evidence by pharmacy students.

MATERIALS AND METHODS: Students attended eight lectures on evidence-based medicine. Students independently assessed two SRs using AMSTAR-2 and ROBIS. The agreement between both tools were calculated using Spearman’s test while interrater reliability was calculated using Fleiss’ κ statistics.

RESULTS: Students reported a preference for the AMSTAR-2 tool due to its clear and distinct rating criteria as well as guidance provided by the tool’s developer. In comparison, students found the items on the ROBIS tool difficult to judge as it was subjective. A moderate agreement between both tools on the overall domain ratings was noted (Spearman r _s = 0.60). There was slight agreement in the overall confidence using AMSTAR-2 (κ = 0.05; 95% confidence interval [CI]: 0.01-0.12) and the overall domain in ROBIS (κ = 0.09; 95% CI: 0.01-0.16).

CONCLUSION: The AMSTAR-2 tool had a low level of concordance in ratings of review among students. However, the AMSTAR-2 tool was preferred by students due to the clear guidance and ease of use.

PMID:40476276 | PMC:PMC11795889 | DOI:10.1002/cesm.12023

Cochrane Evid Synth Methods. 2024 May 15;2(5):e12057. doi: 10.1002/cesm.12057. eCollection 2024 May.

ABSTRACT

INTRODUCTION: An important mechanism of research waste is inadequate incorporation of, and references to, previous relevant research. Identifying references for a research manuscript can be challenging, in part due to the exponential rise in potentially relevant literature to consider. For large research projects, such as developing or updating reporting guidelines, it may be helpful to construct a supportive topic-specific bibliographic database.

METHODS: In support of updating the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) 2013 and the CONsolidated Standards Of Reporting Trials (CONSORT) 2010, we developed the SPIRIT-CONSORT Evidence Bibliographic database (SCEBdb): a freely available topic-specific bibliographic database of publications providing an evidence foundation for the updates. We searched multiple sources of potential publications and tagged included ones with database-specific keywords. For context, we also formulated 10 core considerations for constructing topic-specific bibliographic databases and identified and described 5 illustrative other databases.

RESULTS: As of April 2024, the SCEBdb included 846 publications. The database proved useful as a supplementary information source for our scoping review of published comments on SPIRIT 2013 and CONSORT 2010, for a supplementary Delphi process, and in the writing phase of the guidance documents. We expect that the database will be useful for future projects within the fields of clinical research methodology, bias, evidence synthesis, and randomized trials.

CONCLUSION: The methods involved in constructing the SCEBdb, and our suggested core considerations for topic-specific bibliographic databases, could be helpful for researchers reflecting on whether, and how, to develop a topic-specific bibliographic database.

PMID:40476263 | PMC:PMC11795946 | DOI:10.1002/cesm.12057

Kans J Med. 2025 Apr 14;18(2):31-34. doi: 10.17161/kjm.vol18.22921. eCollection 2025 Mar-Apr.

ABSTRACT

INTRODUCTION: Aspirin and an oral P2Y12 inhibitor are recommended for one year after percutaneous coronary intervention (PCI) in patients with acute coronary syndromes. While ticagrelor or prasugrel, more potent P2Y12 inhibitors, are preferred over clopidogrel, de-escalation often is based on provider judgment. This study compared cardiovascular outcomes and bleeding risks between patients who remained on ticagrelor or prasugrel (unchanged group) and those de-escalated to clopidogrel within 30 days of PCI.

METHODS: The authors analyzed data from patients admitted between June 2014 and December 2022 for acute coronary syndromes requiring PCI who received an oral P2Y12 inhibitor within 72 hours of admission. The primary outcome was a composite of all-cause mortality, urgent revascularization, stent thrombosis, stroke, and major bleeding at one year. Secondary outcomes included the individual components of the composite outcome. Statistical analyses included chi-square tests, Student’s t-tests, or non-parametric equivalents, as appropriate.

RESULTS: A total of 210 patients met the inclusion criteria, with 149 remaining on unchanged P2Y12 therapy and 61 undergoing de-escalation. There was no statistically significant difference in the composite outcome between the unchanged and de-escalated groups (n [%]: 25 [17] vs. 6 [10]; χ² [1, N = 210] = 1.658, p = 0.198). Additionally, secondary outcomes, including all-cause mortality, urgent revascularization, stent thrombosis, stroke, and major bleeding, did not differ significantly between groups.

CONCLUSIONS: A composite outcome of all-cause mortality, urgent revascularization, stent thrombosis, stroke, and major bleeding at one year was similar between patients who continued ticagrelor or prasugrel and those de-escalated to clopidogrel within 30 days of PCI. Larger studies are needed to confirm these findings and assess the optimal timing for therapy adjustments.

PMID:40476260 | PMC:PMC12135783 | DOI:10.17161/kjm.vol18.22921

IBRO Neurosci Rep. 2025 May 1;18:732-738. doi: 10.1016/j.ibneur.2025.04.018. eCollection 2025 Jun.

ABSTRACT

BACKGROUND: Alzheimer’s disease (AD) is a complex neurodegenerative disorder marked by progressive cognitive decline and disrupted brain network connectivity, particularly within the default mode network (DMN). Neurofilament light chain (NfL) serves as a biomarker for axonal injury, but the role of genetic predisposition, assessed via the Polygenic Hazard Score (PHS), in mediating the association between plasma NfL and DMN connectivity remains unclear. This study investigates whether PHS mediates the association between plasma NfL levels and DMN connectivity in individuals across different cognitive stages, including cognitively normal (CN), mild cognitive impairment (MCI), and AD.

METHODS: Data were extracted from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Plasma NfL concentrations were measured using the Simoa assay, and resting-state fMRI (rs-fMRI), assessed DMN connectivity. The cohort included 102 participants (nCN=28, nMCI=52, and nAD=22). Partial correlation analyses and mediation models were performed, adjusting for age and gender. Statistical significance was set at p < 0.05, after corrections for multiple comparisons.

RESULTS: Plasma NfL levels were significantly higher in AD group compared to CN and MCI groups (p = 0.030). DMN connectivity showed substantial declines in the AD group, particularly in the posterior and ventral regions. Significant negative correlations were observed between plasma NfL and ventral DMN connectivity in AD. However, mediation analysis indicated no significant indirect effect of PHS, suggesting that genetic risk does not mediate the plasma NfL-DMN association.

CONCLUSION: These findings suggest that elevated plasma NfL levels are associated with disrupted ventral DMN connectivity in AD, reflecting neurodegeneration-related network dysfunction. However, the lack of a mediating effect by PHS indicates that this relationship is likely independent of genetic risk burden.

CLINICAL TRIAL NUMBER: not applicable.

PMID:40476223 | PMC:PMC12139489 | DOI:10.1016/j.ibneur.2025.04.018