Category: Statistics

JAMA Netw Open. 2026 May 1;9(5):e2611644. doi: 10.1001/jamanetworkopen.2026.11644.

ABSTRACT

IMPORTANCE: Attending anesthesiologists are not subject to work hour restrictions. Fatigue from long shifts may plausibly contribute to patient harm.

OBJECTIVE: To examine the association between anesthesiologist extended work periods and patient outcomes.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used data from the Multicenter Perioperative Outcomes Group registry from January 1, 2010, to August 30, 2020, representing all surgical procedures in patients 18 years or older from more than 50 hospitals across 18 US states. Data analyses were conducted from July 1 through November 30, 2025.

EXPOSURE: Anesthesiologist extended work periods, defined as 16 hours or more of continuous intraoperative work.

MAIN OUTCOMES AND MEASURES: The primary outcome was a composite of in-hospital mortality or major complication. Secondary outcomes were individual components of the composite. A within-anesthesiologist approach was used to compare outcomes for patients treated by each anesthesiologist when in an extended work period vs not, adjusting for patient demographics, comorbidities, American Society of Anesthesiologists classification, emergency status, hour of surgery start, and surgery type.

RESULTS: The study cohort comprised 1 648 720 surgical procedures involving 1711 unique anesthesiologists across 36 institutions (patient mean [SD] age, 51.7 [17.8] years; 982 020 [59.6%] female). A total of 11 556 procedures (0.7%) involved anesthesiologists in an extended work period. Unadjusted incidence of the composite outcome was 12.3% (95% CI, 11.7%-12.9%) for extended work periods vs 12.0% (95% CI, 12.0%-12.1%) for standard work periods. Adjusted absolute risk difference in the composite outcome was 0.1% (95% CI, -0.4% to 0.5%; P = .82). Multiple sensitivity analyses resulted in statistical significance with point estimates of similar direction and magnitude. In-hospital mortality was higher for extended (1.3%; 95% CI, 1.1%-1.6%) vs standard (1.0%; 95% CI, 1.0%-1.0%) work periods, with a risk difference of 0.3 (95% CI, 0.1%-0.5%; P = .009). Other secondary outcomes had risk increases of similar magnitude.

CONCLUSIONS AND RELEVANCE: In this cross-sectional study, the primary analysis did not identify a statistically significant difference in composite outcome, although the magnitude and precision of risk estimates across multiple sensitivity analyses suggested a small (<1.0%) increase in risk. Secondary analyses also suggested small absolute increases in outcomes, including in-hospital mortality. These findings raise the possibility that fatigue related to prolonged anesthesiologist work periods may have implications for patient safety.

PMID:42096198 | DOI:10.1001/jamanetworkopen.2026.11644

Expert Rev Anticancer Ther. 2026 May 7. doi: 10.1080/14737140.2026.2671251. Online ahead of print.

ABSTRACT

BACKGROUND: Treatment options for non-small cell lung cancer (NSCLC) patients lacking actionable genetic alterations (non-AGA) remain limited, particularly after first-line (1 L) therapy. This study evaluated real-world treatment patterns and outcomes among second or later-line (2 L+) patients with non-AGA nonsquamous (NSQ) metastatic NSCLC (mNSCLC) at a Health Maintenance Organization in Israel.

METHODS: A retrospective cohort study used the Maccabi Healthcare Services database to identify 2 L+ adult patients with non-AGA NSQ mNSCLC between January 2017 and December 2020. Outcomes included progression-free survival (PFS), overall survival (OS), and time-to-next-treatment or death (TTNTD), analyzed using descriptive statistics and Kaplan-Meier methodology.

RESULTS: Among 176 2 L+ patients, median age was 67 years, 66.5% were male, 51.7% had an Eastern Cooperative Oncology Group (ECOG) performance status of 0-1, and 30.7% received third-line therapy. Anti-programmed death-receptor/ligand 1 (anti-PD-[L]1) was the most common 2 L therapy (n = 74, 42.0%) following 1 L chemotherapy. Median overall TTNTD from 2 L initiation was 3.73 months (docetaxel: 1.78 months; anti-PD-[L]1 combination: 7.05 months). Median overall PFS was 2.56 months (docetaxel: 1.78 months; anti-PD-[L]1 combination: 9.48 months) and median overall OS was 5.51 months (docetaxel: 1.81 months; anti-PD-[L]1 combination: 9.48 months).

CONCLUSIONS: These findings highlight the high unmet need among 2 L+ non-AGA NSQ mNSCLC patients in Israel.

PMID:42096187 | DOI:10.1080/14737140.2026.2671251

Environ Sci Technol. 2026 May 7. doi: 10.1021/acs.est.5c15628. Online ahead of print.

ABSTRACT

Fine particulate matter (PM₂.₅) is a major environmental health risk in China, yet urban-rural disparities in health burden remain insufficiently characterized. Using the high-resolution PM₂.₅ concentration data, Global Burden of Disease estimates, and urban-rural stratified provincial health statistics, this study quantified PM₂.₅-attributable premature deaths in urban and rural areas from 2000 to 2019 and applied decomposition analysis to assess contributions of PM₂.₅ concentrations, population size, age structure, baseline mortality, and urban expansion. Although PM₂.₅ concentrations declined substantially after 2014, the health burden remained substantial and shifted from rural to urban populations. Population aging was the dominant driver, especially in rural areas, where it offset much of the benefit from air-quality improvement. Urban population growth further increased risks, whereas rural depopulation and urban expansion reduced rural burdens. Declining baseline mortality prevented premature deaths but disproportionately benefited urban residents due to healthcare access inequities. These urban-rural trends were consistent across different exposure-response models, underscoring the robustness of our findings. These findings highlight the need for differentiated environmental health policies that integrate demographic changes, regional disparities, and environmental justice, coordinated urban-rural air-quality management, and targeted health interventions for aging populations.

PMID:42096176 | DOI:10.1021/acs.est.5c15628

Int J Comput Assist Radiol Surg. 2026 May 7. doi: 10.1007/s11548-026-03686-0. Online ahead of print.

ABSTRACT

PURPOSE: Robotic-assisted surgery (RAS) generates vast amounts of video and robotic data, presenting opportunities for machine learning. Video-based models, in particular, that can temporally segment frames by ontological categories such as procedure type, phase, steps, actions, etc., are needed. Training separate models for each category neglects statistical dependencies between categories and can yield incompatible predictions. Training large multi-category models may help, but increases complexity while reducing model modularity and interpretability.

METHODS: We present a model fusion alternative: an effectively zero-free-parameter Bayesian model fusion technique. Incorporating the empirical conditional dependencies across categories and time, we combine predictions from multiple segmentation models into one joint Bayesian inference. The result is a Bayes’ optimal distribution over all categories evolving over time with accumulated evidence.

RESULTS: On a large test set of hundreds of surgical cases, of nearly eight million frames of annotated data, we found that fused predictions from the joint Bayesian model provide clear benefits over the individual models, correcting inconsistent and inaccurate predictions, and even forming accurate beliefs when evidence was absent.

CONCLUSION: The model we present is a lightweight, principled alternative to machine learning-based model fusion. A sufficiently complex model could be trained to produce the same results, but would effectively trade explainable predictions with minimal overheard for computational complexity and transparency. We end by discussing how the same approach can be used to encompass larger more sophisticated models within the same conceptual framework.

PMID:42096125 | DOI:10.1007/s11548-026-03686-0

Clin Oral Investig. 2026 May 7;30(6):219. doi: 10.1007/s00784-026-06897-6.

ABSTRACT

OBJECTIVES: To evaluate the efficacy of artificial intelligence (AI)-driven three-dimensional (3D) anatomical models as an adjunct to cone-beam computed tomography (CBCT) for root canal assessment regarding diagnostic accuracy, observer confidence, and time efficiency among undergraduate and postgraduate students.

MATERIALS AND METHODS: In this observational diagnostic study, 26 observers (13 undergraduates and 13 postgraduates) evaluated 22 tooth roots with complex anatomy from nine CBCT scans under two conditions: CBCT alone and CBCT supplemented with AI-generated 3D anatomical models. Observers assessed the number of roots, root canals, and apical foramina, while confidence (5-point Likert scale) and assessment time were recorded. Each observer performed 132 assessments, totaling 3,432 evaluations. A reference standard was established by consensus between two specialists. A significance level was set at 5% (α = 0.05) for all statistical analyses.

RESULTS: Augmenting CBCT with AI-generated 3D models significantly improved diagnostic accuracy for all parameters (p < 0.001). Root detection accuracy reached 100% in both groups. Root canal detection increased from 83% to 94% among undergraduates and from 88% to 99% among postgraduates, while apical foramina detection increased to 99% in both groups. Observer confidence significantly increased (p < 0.001), reaching a median score of 5 (IQR: 5-5). Workflow efficiency also improved (p < 0.001), with median assessment time decreasing from 102 s to 39 s for undergraduates and from 97 s to 24 s for postgraduates.

CONCLUSION: AI-driven 3D anatomical models used with CBCT enhance diagnostic accuracy, observer confidence, and evaluation efficiency in endodontic assessment. However, multi-centre studies with larger, more diverse samples, particularly including cases with pronounced artefacts, would further support generalisability.

CLINICAL RELEVANCE: AI-generated 3D anatomical models derived from CBCT scans may serve as a valuable adjunct for the interpretation of complex root canal anatomy, improving diagnostic accuracy, increasing observer confidence, and reducing assessment time. These findings support their potential role not only in clinical decision-making but also as an effective educational tool for training dental students and clinicians.

PMID:42096111 | DOI:10.1007/s00784-026-06897-6

Eur J Pediatr. 2026 May 7;185(6):358. doi: 10.1007/s00431-026-07003-4.

ABSTRACT

The purpose of this study is to evaluate whether piperacillin-tazobactam therapy, compared with cefepime monotherapy, results in significant difference in treatment success, mortality, and duration outcomes in pediatric patients with febrile neutropenia (FN). A systematic review and meta-analysis of randomized controlled trials (RCTs) was conducted following PRISMA guidelines. PubMed, Embase, Scopus, Web of Science, and Cochrane CENTRAL were searched up to February 3. No starting limits for dates were used. We only included studies of pediatric patients with FN comparing piperacillin-tazobactam to cefepime monotherapy. Risk of bias was assessed using Cochrane RoB 2 tool. Meta-analysis was performed using random-effects model to calculate risk ratios (RR) and mean differences (MD). Five RCTs involving 470 episodes were included. The pooled analysis for treatment success showed no statistically significant difference between groups (RR = 1.02; 95% CI [0.89; 1.18]; P = 0.76) with low heterogeneity (I² = 0.0%). No significant differences for mortality (RR = 2.09; 95% CI [0.62; 7.03]; P = 0.23; I² = 0.0%). Duration of treatment was 0.9 day shorter for cefepime group (MD = 0.9 day; 95% CI [0.2; 1.6]; P < 0.1; I² = 0%).

CONCLUSION: No statistically significant difference was found in treatment success or mortality between the groups. Patients receiving cefepime had a treatment duration 0.9 days shorter than those receiving piperacillin-tazobactam, although this finding should be interpreted with caution. The available evidence remains limited. Further RCTs are needed to elucidate potential differences in treatment success, mortality, and duration between the drugs.

REGISTRATION: PROSPERO (CRD420261296621).

WHAT IS KNOWN: • Cefepime and piperacillin-tazobactam are widely utilized as empirical monotherapy for febrile neutropenia, but evidence regarding their comparative clinical efficacy and safety in children remains inconclusive.

WHAT IS NEW: • This meta-analysis quantifies differences in treatment success, mortality and duration of treatment outcomes in pediatric patients with febrile neutropenia.

PMID:42096109 | DOI:10.1007/s00431-026-07003-4

Hernia. 2026 May 7;30(1):197. doi: 10.1007/s10029-026-03694-8.

ABSTRACT

BACKGROUND: Ventral hernia repair (VHR) is commonly performed in patients with multiple modifiable risk factors. Preoperative rehabilitation programs aim to optimize these risk factors and improve surgical outcomes; however, their implementation and effectiveness remain variable. This systematic review evaluates current evidence on the role of preoperative optimization in patients undergoing VHR.

MATERIALS AND METHODS: Cochrane Central, Embase, and PubMed were searched for studies comparing preoperative interventions versus standard care in patients undergoing VHR. The primary outcome was characterization of rehabilitation strategies. Secondary outcomes included surgical utilization, emergent repair, readmission, surgical site infection (SSI), surgical site occurrence (SSO), recurrence, reoperation, and mortality.

RESULTS: Six studies were included, comprising a total of 3,556 patients, of whom 1,805 (50.7%) underwent preoperative optimization. Preoperative optimization interventions were highly heterogeneous and primarily consisted of multidisciplinary programs focused on exercise, weight loss, and lifestyle modification. Control groups varied widely, precluding formal meta-analysis. Studies requiring predefined optimization targets reported unsuccessful rehabilitation rates of 5.1%-45%. Emergent repair occurred in 6.8% of intervention patients in one study. SSI rates seemed lower and ranged from 4.2% to 11.8% in intervention groups versus 11.0%-16.7% in controls, SSO and reoperation rates were generally lower in optimization groups.

CONCLUSION: Structured preoperative optimization pathways may be associated with improved short-term outcomes in select patients undergoing VHR. However, the current literature is heterogeneous, and standardized protocols and prospective studies are needed to better define optimal implementation strategies and long-term effectiveness.

PMID:42096105 | DOI:10.1007/s10029-026-03694-8

Inflamm Res. 2026 May 7;75(1):111. doi: 10.1007/s00011-026-02257-y.

ABSTRACT

BACKGROUND: Over 25% of patients with ulcerative colitis (UC) develop extraintestinal manifestations (EIMs), resulting in significant systemic morbidity. We define the shared genetic foundation of these manifestations as the UC-associated Multisystem Inflammatory Genetic Architecture (UC-MIGA). This study aims to identify shared genomic drivers and actionable immunosenescence therapeutic targets across the UC-EIM spectrum.

METHODS: We applied genomic structural equation modeling (SEM) to seven European-ancestry GWAS datasets (UC, deep vein thrombosis, ankylosing spondylitis, primary sclerosing cholangitis, pyoderma gangrenosum, interstitial lung disease, and erythema nodosum) to identify a shared latent genetic factor (F1). Post-SEM analyses included FUMA mapping, SuSIE/FINEMAP fine-mapping, FUSION/FOCUS transcriptome-wide studies, MAGMA enrichment, CELLECT deconvolution, LDSC partitioned heritability, and single-cell eQTL Mendelian randomization (MR). UC exhibited the highest standardized factor loading (0.9801) on F1, justifying its use as a representative proxy for UC-MIGA in downstream analyses. UC-telomere relationships were assessed via tissue-specific eQTL/sQTL enrichment across 49 GTEx tissues, spatial transcriptomics (gsMap), single-cell profiling (GSE214695, GSE163974), hdWGCNA, and colocalization analyses (eCAVIAR, fastENLOC).

RESULTS: SEM identified substantial genetic overlap (CFI = 1.0, SRMR = 0.17). Within the UC-MIGA framework, we identified 17,005 SNPs (P ≤ 1 × 10⁻2⁰⁰), 2,622 risk loci, and 152 high-confidence effector genes. Pathways implicated Th17/Treg imbalance and inflammasome signaling. Super-enhancer regions showed exceptional heritability enrichment (80.16%, fold = 4.79, p = 0.0007). MR identified 35 causal immune cell-gene associations. UC-telomere analyses revealed convergence in colon-specific DNA repair-mitochondrial energetics-telomere maintenance pathways, with B cells prioritized as the core cell type. Colocalization identified NKX2-3 and LINC01475 as high-confidence shared candidates. Embryonic intestinal enrichment supported the developmental origins of this systemic axis.

CONCLUSION: UC-MIGA represents a genetically coherent architecture driven by super-enhancer-mediated epigenetic dysregulation, Th17/Treg imbalance, and immunosenescence features, including telomere dysfunction and B-cell exhaustion. The ‘developmental vulnerability-environmental trigger’ model explains the gut-origin inflammatory cascade underlying extraintestinal manifestations, with UC-telomere analysis providing a genomic foundation for systemic therapeutic strategies targeting the inflammation-aging nexus.

PMID:42096094 | DOI:10.1007/s00011-026-02257-y

Orthopadie (Heidelb). 2026 May 7. doi: 10.1007/s00132-026-04841-w. Online ahead of print.

ABSTRACT

BACKGROUND: Different minimally invasive approaches have been established in the management of elective hip arthroplasty. However, there is almost no data on the endoprosthetic treatment of femoral neck fractures in obese patients.

MATERIALS AND METHODS: The monocentric case-control study is based on the retrospective evaluation of a database. Included were patients with a body mass index (BMI) of > 30 kg/m² who received endoprosthetic treatment due to a femoral neck fracture. The cohort was dichotomized based on the surgical approach: direct lateral (DL) and minimally invasive anterolateral (MIV AL). Demographic and perioperative variables were collected. The primary study objective was the comparison of general and specific complications. The secondary study objective was the comparison of survival rate and mobility according to the Parker score.

RESULTS: A total of 155 surgeries conducted between 2011 and 2024 were included, of which 52 procedures were performed using a DL approach and 103 procedures using an MIV AL approach. The mean age of the overall cohort was 77.9 ± 9.4 years, and the mean BMI was 33.4 ± 3.2 kg/m². The demographic variables showed a homogeneous distribution. The overall revision rate was noticeably higher for the MIV AL approach, but not statistically significant (p = 0.18). Dislocations (n = 7) and infections (n = 3) occurred exclusively in the MIV AL group. There was no significant difference in mortality or survival rates. The 30-day and 1‑year mortality rates were 0% and 9.0%, respectively (p = 0.776). The mean survival rate according to Kaplan-Meier was 7.1 years (95% confidence interval, 6.12-8.03) without significant differences regarding the surgical approaches (log rank = 0.656). Mobility according to Parker (n = 78) at the time of follow-up was also comparable (p = 0.60).

CONCLUSION: The MIV AL approach to endoprosthetic treatment of femoral neck fractures in obese patients showed a higher, but not significant, dislocation and infection rate compared to the DL approach. For the overall cohort, mortality rates were low. Recording the surgical approaches and BMI in registry data would be useful to generate larger case numbers.

LEVEL OF EVIDENCE: III; retrospective case-control study.

PMID:42096082 | DOI:10.1007/s00132-026-04841-w

Hernia. 2026 May 7;30(1):198. doi: 10.1007/s10029-026-03703-w.

ABSTRACT

BACKGROUND: Recurrence of hiatal hernia after primary repair is a major concern, particularly early recurrence related to technical factors. Reliable preoperative predictors are essential for optimal surgical strategies. Because the hernial sac area can also affect operative complexity, this study aimed to evaluate early recurrence-associated factors and their relationship with operative time using quantitative computed tomography (CT).

METHODS: We retrospectively analyzed 48 patients who underwent laparoscopic hiatal hernia repair between July 2012 and July 2025. Preoperative CT was used to measure the maximum diameter and area of the hernial sac in the axial and coronal planes. We examined the association between these parameters and surgical outcomes including recurrence.

RESULTS: Early recurrence was observed in four patients (8.3%), all of whom required reoperation within 3 days. Patients with early recurrence had longer operative times and larger axial hernial sac areas (p < 0.05). Receiver operating characteristic (ROC) curve analysis demonstrated that the axial maximum diameter provided the highest predictive performance for early recurrence (area under the curve [AUC] = 0.89), with an optimal cutoff value of 117.6 mm. Other parameters, including axial area, coronal area, coronal diameter, and hiatal width, showed moderate predictive ability but none exceeded axial diameter accuracy.

CONCLUSION: Axial hernial sac measurements are associated with early recurrence. Preoperative CT-based quantitative assessments may provide valuable information for predicting surgical complexity and recurrence, thereby supporting preoperative risk stratification and surgical planning. Therefore, prospective validation using a larger cohort is warranted.

PMID:42096069 | DOI:10.1007/s10029-026-03703-w