Category: Statistics

Arthritis Res Ther. 2026 May 1;28(1):102. doi: 10.1186/s13075-026-03822-w.

ABSTRACT

Juvenile idiopathic arthritis (JIA) comprises a heterogeneous group of childhood-onset inflammatory arthritides that remain challenging to treat. An expanding range of biologic and targeted synthetic disease-modifying antirheumatic drugs has improved outcomes, yet major unmet needs persist including predicting treatment response and identifying optimal drug sequences. Traditional frequentist randomised controlled trials (RCTs) are often infeasible in JIA because of small sample sizes and ethical concerns about placebo use in children. Modified RCTs – including randomised withdrawal, placebo-phase, and escape designs – have reduced placebo exposure but have inherent limitations. Here we discuss how innovative trial methodologies may enhance our capacity to generate evidence that can improve outcomes for children and young people with JIA. Adaptive designs, Bayesian methods, sequential multiple assignment randomized trials, and two-stage stop-go designs offer greater flexibility within trials, allowing for study protocols to be adapted based on preliminary data. Insights from multiomic analysis of JIA synovial tissue mean that biologically defined endotypes may fundamentally reshape trial stratification and support biomarker-led precision medicine. Master protocols (basket, umbrella and platform trials) provide opportunities to streamline research by combining data into a single trial. The incorporation of external data can increase statistical power and circumvent challenges in recruiting paediatric controls. Integrating these methodological and biological innovations will be critical for delivering targeted, effective, and age-inclusive therapeutic strategies, ultimately improving outcomes for children and young people living with JIA.

PMID:42067908 | DOI:10.1186/s13075-026-03822-w

J Ovarian Res. 2026 May 1. doi: 10.1186/s13048-026-02110-0. Online ahead of print.

ABSTRACT

BACKGROUND: Characterisation of CT detected ovarian masses is challenging with overlapping imaging features, unreliable biomarker or clinical presentation. We proposed a two-staged CT-based radiomics model to identify early-stage ovarian carcinoma (ES-OC) and sub-classify different types of benign ovarian masses (BOM).

METHODS: Patients with histologically confirmed BOM or ES-OC (FIGO I-II) were retrospectively recruited from 5 centres. Radiomics features were derived from CT images using PyRadiomics (v3.0.1), which intrinsically resampled volumes to isotropic 1 mm³ voxels. To reduce feature redundancy, features with high correlation (Spearman’s ρ ≥ 0.85) were excluded. Two-staged feature selection was applied. First, elastic-net regression with repeated 5-fold stratified cross-validation (100 iterations) was performed to identify highly repeatable features, followed by Mann-Whitney U testing for statistical significance. Second, Boruta algorithm with Random Forest (RF) estimator was employed over 500 iterations to robustly select features by comparing their importance to randomized shadow features. Several machine learning (ML) classifiers were evaluated using stratified 10‑fold GridSearch cross-validation with area under the curve (AUC) as tuning metric. The optimal model from each stage with highest cross-validated AUC was then evaluated on the respective test set. The AUC, calibration plot, and decision curve analysis (DCA) were employed to assess the performance and clinical utility of models.

RESULTS: The study enrolled 483 patients with 529 lesions (ES-OC: 192 patients, 192 lesions; BOM: 291 patients, 337 lesions). In the first-stage, logistic regression (LR) algorithm was selected with high sensitivity (0.870), moderate specificity (0.719) and high AUC (0.859) in the test set. In the second-stage, support vector machines (SVM) had high diagnostic accuracy with sensitivity 0.750, specificity 0.839 and AUC 0.918. DCA identified the highest benefit at 0.20 risk threshold probability in determining ES-OC.

CONCLUSION: The two-staged CT-based radiomics model incorporating LR and SVM algorithms had high diagnostic efficiency in characterising ES-OC and BOM, potentially in triaging disease and personalising care.

PMID:42067904 | DOI:10.1186/s13048-026-02110-0

BMC Psychol. 2026 May 1;14(1):643. doi: 10.1186/s40359-026-04627-x.

ABSTRACT

Stigmatizing attitudes toward individuals with mental disorders represent a major barrier to treatment, recovery, and social inclusion. The present research introduces and psychometrically evaluates the German-language Stigma Toward People with Mental Disorders scale (SToP-MD) across three independent studies with distinct samples.In study 1 (N = 266), an initial item pool was developed and refined based on theoretical frameworks and exploratory factor analysis. In study 2 (N = 448), confirmatory factor analysis supported a two-factor structure comprising prejudiced stigmatization (SToP-MD-PS) and assumption of problems (SToP-MD-AP). The model demonstrated adequate fit according to conventional indices (CFI = 0.97, TLI = 0.96, SRMR = 0.07), although robust indices indicated only moderate fit (robust CFI = 0.91, robust RMSEA = 0.13). Internal consistency was good for the PS subscale (ω = 0.83) but limited for the AP subscale (ω = 0.51). In study 3 (N = 266), the scale’s sensitivity to short-term change was examined following exposure to differently framed media content.As hypothesized, the SToP-MD subscales were positively associated with depression stigma (DSS) and social distance (SDI), and negatively correlated with openness and agreeableness (NEO-FFI), supporting convergent validity. Discriminant validity was partially confirmed by low or non-significant correlations with attitudes toward physically disabled individuals (ATDP), suicide-related cognitions (CCSS), and socially desirable responding (BIDR).Across all three studies, the SToP-MD demonstrated preliminary yet consistent evidence of structural and construct validity, as well as change sensitivity. It captures both overt prejudices and implicit burden assumptions, offering a nuanced assessment of public stigma toward mental disorders. The scale can serve as a valuable tool in stigma research, public health monitoring, and evaluation of interventions. Future research should extend validation to more diverse samples and test predictive and longitudinal utility.

PMID:42067898 | DOI:10.1186/s40359-026-04627-x

BMC Musculoskelet Disord. 2026 May 1. doi: 10.1186/s12891-026-09802-0. Online ahead of print.

ABSTRACT

BACKGROUND: Traumeel (Tr14) and diclofenac gels are commonly prescribed topical treatments for Grade I-II acute lateral ankle sprain (LAS). A meta-analysis of two multi-center trials was performed to compare their clinical efficacy and supplement the evidence in support of topical treatments.

METHODS: Data from two Phase III prospective, randomised, double-blind, reference-controlled clinical trials, TAASS (2013) and TRAUMED (2024) were examined. An Individual Patient Data (IPD) meta-analysis was performed on those with an initial pain VAS (Visual Analogue Scale) score ≥ 50 mm, using the statistical analysis plan from the latter to ensure methodological consistency.

RESULTS: A total of 628 patients underwent meta-analysis (mean baseline pain VAS = 72.3, mean age = 32.5). The primary endpoint of the IPD meta-analysis, Area Under the Curve (AUC) of pain VAS scores, indicated beneficial effects of Tr14 gel compared with diclofenac gel, statistically significant on Days 4, 7, and 14 (Day 4 mean difference MD_AUC-META = -6.9 [95%CI -13.7 to -0.1], P_AUC-META = 0.05; Day 7 MD_AUC-META = -26.3 [95%CI -40.4 to -12.2], P_AUC-META = 0.0003; Day 14 MD_AUC-META = -63.6 [95%CI -92.3 to -34.8]), P_AUC-META < 0.0001). A robust Wilcoxon-Mann-Whitney (WMW) analysis revealed percent changes from baseline in pain and function that were significantly superior for Tr14 gel in the pre-planned method of synthesis, compared to diclofenac gel, on all days (all PMETA < 0.05). Finally, translational effect sizes indicated less than “small” superiority on Day 4 (SMD 0.15, OR 0.78, NNT 23), and “small-medium-sized” superiority on Days 7 and 14 (Day 7: SMD 0.31, OR 0.59, NNT 12; Day 14: SMD 0.37, OR 0.53, NNT 10) for pain measured by VAS, as compared to diclofenac gel.

CONCLUSION: The IPD meta-analysis of the TAASS and TRAUMED trials confirms the efficacy of Tr14 gel. In-depth analysis reveals it is a statistically favourable treatment with respect to pain, and function, compared to the gold-standard diclofenac gel. Taking this together, Tr14 gel represents a statistically favourable treatment option for Grade I-II ankle sprain within the context of the analysed endpoints.

PMID:42067889 | DOI:10.1186/s12891-026-09802-0

BMC Surg. 2026 May 1. doi: 10.1186/s12893-026-03799-y. Online ahead of print.

ABSTRACT

OBJECTIVE: This study aimed to systematically compare robot-assisted nipple-sparing mastectomy (R-NSM) with conventional nipple-sparing mastectomy (C-NSM) in terms of perioperative safety and oncological outcomes through a meta-analysis.

METHODS: A comprehensive literature search was conducted in PubMed/MEDLINE, Web of Science Core Collection, Embase, and the Cochrane Central Register of Controlled Trials from database inception to February 27, 2026. Controlled studies comparing robotic nipple-sparing mastectomy (R-NSM) and conventional nipple-sparing mastectomy (C-NSM) in women with early-stage breast cancer (stage 0-II, including a small proportion of risk-reducing mastectomy cases) were included. Study selection and data extraction were independently performed by two reviewers, with discrepancies resolved by consensus. Statistical analyses were conducted using Review Manager (RevMan) version 5.4. The primary outcomes were overall postoperative complications, major complications (Clavien-Dindo grade ≥ III), positive surgical margin, and local recurrence. Secondary outcomes included operative time, estimated intraoperative blood loss, length of hospital stay, reoperation rate, and individual postoperative complications (including nipple-areolar complex necrosis, skin-flap necrosis, surgical-site infection, hematoma, seroma, delayed wound healing, and implant loss). Perioperative mortality was not reported in the included studies.

RESULTS: A total of 12 studies involving 2,312 patients (1 RCT and 11 non-randomized studies) were included in the quantitative synthesis. The reported follow-up duration across included studies ranged from 3 to approximately 51 months, with most studies reporting a median or mean follow-up of around 18-30 months. Compared with C-NSM, R-NSM was associated with a lower incidence of overall postoperative complications (RR = 0.82, 95% CI: 0.68-0.99, P = 0.04) and major complications defined as Clavien-Dindo grade ≥ III (RR = 0.44, 95% CI: 0.22-0.86, P = 0.02). R-NSM was also associated with a reduced risk of delayed wound healing (RR = 0.51, 95% CI: 0.26-0.98, P = 0.04). However, no statistically significant differences were observed between the two approaches with respect to nipple-areolar complex necrosis, skin flap necrosis, postoperative infection, hematoma, seroma, implant loss, reoperation rate, positive surgical margin, or local recurrence. R-NSM was associated with longer operative time and a modest reduction in intraoperative blood loss. However, TSA suggested that the cumulative evidence remains insufficient.

CONCLUSION: Current evidence suggests that, under strict patient selection, R-NSM does not appear to compromise short-term perioperative safety or oncological outcomes when compared with the conventional approach. Although a reduction in composite postoperative complications was observed, no significant differences were identified in most individual major complications, and operative time remains longer. Given the predominance of retrospective studies, limited information size, and relatively short follow-up, further high-quality prospective research is required to more definitively establish the role of R-NSM in clinical practice.

PMID:42067887 | DOI:10.1186/s12893-026-03799-y

BMC Musculoskelet Disord. 2026 May 1. doi: 10.1186/s12891-026-09873-z. Online ahead of print.

ABSTRACT

INTRODUCTION: Chronic low back pain (CLBP) is a leading cause of disability worldwide. Physiotherapy is a common treatment, but its effect on physical functioning is generally modest, particularly for patients with severe complaints (i.e., high levels of disability and pain). Virtual Reality (VR) may complement physiotherapy, yet evidence for its effectiveness remains limited. The aim of this study was to assess the effectiveness and feasibility of a VR intervention integrated within physiotherapy for people with severe CLBP.

METHODS: A cluster-RCT across Dutch physiotherapy practices was conducted. Patients in the control group received 12 weeks of usual care following physiotherapy guidelines. Patients in the intervention group received similar usual care, enhanced with integrated, home-based VR consisting of pain education, exercise therapy, and relaxation modules. The primary outcome was physical functioning at three months. Secondary outcomes included feasibility, pain intensity, and catastrophizing. Analyses were conducted using linear mixed-effect models accounting for clustering by physiotherapy practice.

RESULTS: Twenty-five patients participated in the intervention group and seven in the VARIETY control group, instead of the planned sample size of 120 participants. Due to poor recruitment (n = 7), we supplemented the VARIETY control group with 18 historical controls from two comparable trials (total control n = 25), effectiveness analyses are therefore exploratory. Between-group differences were neither statistically significant nor clinically relevant for all outcome measures, compared to the VARIETY control group (e.g., ODI mean difference at three months: -4.80, 95%CI: -17.78;8.18), or the total control group (-8.80, 95%CI: -20.17;2.57). The intervention group showed greater improvements from baseline in physical functioning (42%), compared to the VARIETY control group (26%). The intervention was considered feasible and safe to use in practice.

DISCUSSION: This study found limited support for the use of VR as an adjunct to physiotherapy for people with severe CLBP. Given the use of external control data and the limited sample size, the effectiveness results should be interpreted cautiously. Further well-powered trials should reconsider study procedures to optimize patient recruitment and corroborate VR’s clinical effectiveness in physiotherapy.

TRIAL REGISTRATION: The study was registered with ClinicalTrials.gov on 2022-12-02 (reference number: NCT05701891).

PMID:42067884 | DOI:10.1186/s12891-026-09873-z

BMC Oral Health. 2026 May 1. doi: 10.1186/s12903-026-08497-y. Online ahead of print.

ABSTRACT

OBJECTIVE: The aim of this study was to investigate the effects of post-polymerization protocols with and without glycerin on surface roughness, microhardness, and color change of permanent restorative resins fabricated using stereolithography (SLA) and digital light processing (DLP) technologies under different storage media.

MATERIALS AND METHODS: In this in vitro study, a total of 128 permanent restorative resin specimens fabricated using SLA (Permanent Crown, Formlabs) and DLP (Crowntec, Saremco) technologies were evaluated. The specimens were divided into post-polymerization protocols with and without glycerin. Finishing and polishing procedures were performed using a standardized diamond polishing system. Surface roughness (Ra, Sa) was measured using a confocal microscope, microhardness was assessed by the Vickers hardness test (VHN), and color change was evaluated using the CIEDE2000 (ΔE₀₀) method. The specimens were stored in distilled water, tea, coffee, and cola, and measurements were repeated at baseline and after 7 and 30 days.

RESULTS: No statistically significant difference in surface roughness was observed between post-polymerization protocols with and without glycerin. In contrast, specimens post-polymerized with glycerin exhibited significantly higher microhardness values for both fabrication techniques. With respect to color change, lower ΔE₀₀ values were observed in the glycerin-treated groups, particularly in the coffee storage medium.

CONCLUSIONS: Post-polymerization with glycerin increased microhardness without adversely affecting surface roughness and supported color stability under specific staining storage media.

CLINICAL RELEVANCE: In permanent restorative resins fabricated by 3D printing, post-polymerization with glycerin can be considered a complementary protocol that is easily applicable in clinical practice and may positively support the color stability and microhardness of restorations.

PMID:42067883 | DOI:10.1186/s12903-026-08497-y

Arthritis Res Ther. 2026 May 1. doi: 10.1186/s13075-026-03821-x. Online ahead of print.

ABSTRACT

INTRODUCTION: The 6-minute walk distance (6MWD) is commonly used to assess functional capacity in patients with systemic sclerosis (SSc), but its ability to reflect changes in cardiopulmonary status during follow-up remains uncertain. This study aimed to investigate whether variations in the 6MWD accurately mirror changes in organ involvement over time in SSc patients.

METHODS: We conducted a longitudinal study on a global cohort of 227 consecutive SSc patients, including a subgroup of 88 patients with ILD or PH, over a two-year follow-up period. Statistical analyses were performed to assess associations between variations (Δ) of 6MWD (Δ6MWD) and evolution of organ involvement.

RESULTS: The 6MWD varied only mildly during follow-up (mean Δ6MWD – 11.6 ± 67.5 m), including in patients with ILD or PH (mean Δ6MWD – 16 ± 74 m). Significant associations were noted between Δ6MWD and variations of several SSc parameters, mostly with modifications in functional status (changes in ΔBorg (ΔΔBorg) score, p = 0.002; ΔNYHA, p < 10^{– 3}), PH parameters (Δright atrial area, p = 0.045; persistently elevated Nt-pro-BNP levels, p < 10^{– 3}), chronotropic function (ΔΔheart rate, p = 0.015; Δinitial heart rate, p = 0.024), musculoskeletal involvement (occurrence of joint symptoms, p = 0.006) and perceived quality of life (ΔHAQ-DI score, p = 0.023). In most cases, the strength of these associations was mild to moderate (R² = 0.53, adjusted R² = 0.36 in multivariate analysis), suggesting the influence of additional factors to explain the majority of Δ6MWD.

CONCLUSION: While it remains valuable as a marker of disability and functioning, the variation in the 6MWD does not seem to be a robust surrogate for modifications in organ involvements of SSc patients.

PMID:42067881 | DOI:10.1186/s13075-026-03821-x

BMC Public Health. 2026 May 1. doi: 10.1186/s12889-026-27571-2. Online ahead of print.

ABSTRACT

BACKGROUND: Online antenatal education is increasingly used to improve maternal psychological outcomes and preparedness for childbirth. However, evidence regarding its effectiveness remains inconsistent, particularly due to heterogeneity in intervention types and outcome measures.

METHODS: This systematic review and meta-analysis was conducted in accordance with PRISMA 2020 guidelines. PubMed, EMBASE, Scopus, Web of Science, CINAHL, and the Cochrane Library were searched from inception to March 2026. Randomized controlled trials and observational studies evaluating digital antenatal interventions in pregnant women were included. Outcomes included maternal depression, anxiety, fear of childbirth, self-efficacy, and small-for-gestational-age (SGA) incidence. Random-effects models were used, and standardized mean differences (SMDs) and odds ratios (ORs) were calculated. Heterogeneity was assessed using I² statistics.

RESULTS: Twelve studies involving 4,982 participants were included. No significant effects were observed for depression (SMD = – 0.18; 95% CI: -0.45 to 0.09; I² = 86.5%; p = 0.188), anxiety (SMD = – 0.10; 95% CI: -0.66 to 0.46; I² = 92.8%; p = 0.723), or self-efficacy (SMD = 0.56; 95% CI: -0.11 to 1.23; I² = 90.6%; p = 0.102). A reduction in fear of childbirth did not reach statistical significance (SMD = – 0.53; 95% CI: -1.06 to 0.003; p = 0.051. No significant association was found for SGA (OR = 0.73; 95% CI: 0.17-3.14; p = 0.670). Substantial heterogeneity was present across outcomes, limiting the interpretability of pooled estimates.

CONCLUSION: Current evidence, characterized by substantial heterogeneity, does not demonstrate statistically significant effects of online-based antenatal education on maternal depression, anxiety, or fear, nor on improving self-efficacy or preventing SGA infants. However, the high heterogeneity (I² >85% for most outcomes) indicates that these pooled estimates are exploratory, and clinically meaningful effects for specific intervention types cannot be ruled out. Findings should be interpreted cautiously.

PMID:42067880 | DOI:10.1186/s12889-026-27571-2

BMC Med Educ. 2026 May 1. doi: 10.1186/s12909-026-09359-5. Online ahead of print.

ABSTRACT

BACKGROUND: Multiple Choice Questions (MCQs) remain the most widely used written assessment format in health professions education due to their efficiency and reliability; however, their susceptibility to cueing raises concerns regarding their authenticity in assessing true knowledge. Very Short Answer Questions (VSAQs) have emerged as a promising alternative that reduces cueing effects and requires learners to generate responses with greater cognitive engagement, while retaining the logistical advantages of computer-marked assessments. This systematic review and meta-analysis evaluates the comparative performance, discrimination, reliability, and practice effects of VSAQs versus MCQs.

METHODS: Following PRISMA guidelines (31), a systematic search identified studies comparing VSAQs and MCQs in undergraduate and postgraduate health professions education. Data extraction and risk of bias assessment were conducted independently by two reviewers, with disagreements resolved through consensus. Mean scores, standard deviations, discrimination indices, and reliability coefficients were extracted. Standardized Mean Differences (SMD) were pooled using a random-effects model. Psychometric indices were transformed using Fisher’s z-scores. Heterogeneity was quantified using the I² statistic, and sensitivity and subgroup analyses were performed to explore the impact of outliers.

RESULTS: Six cohorts (n = 1,191) derived from three included studies contributed data for score comparisons. The initial pooled analysis showed no significant difference between VSAQs and MCQs (SMD = – 0.52; 95% CI – 1.34 to 0.30), with extreme heterogeneity (I² = 98%). Sensitivity analysis identified Dhok et al. (2023) as an outlier. Excluding this dataset yielded a significant effect favoring MCQs (SMD = – 0.86; 95% CI – 1.01 to – 0.70; p < 0.00001) with markedly reduced heterogeneity (I² = 4%). Meta-analysis of psychometric properties demonstrated strong discrimination (pooled Fisher’s z = 1.45) and acceptable reliability (pooled Fisher’s z = 0.43) for VSAQs. Evidence on practice effects was mixed, with no consistent advantage for either format.

CONCLUSION: VSAQs exhibit strong psychometric integrity and are designed to reduce the cueing effect observed in MCQs. Although associated with lower student scores, VSAQs likely provide a more accurate reflection of independent knowledge and clinical reasoning ability. Their integration into medical assessment systems may enhance the authenticity and validity of written examinations.

PMID:42067879 | DOI:10.1186/s12909-026-09359-5