Tag: statistics

JAMA Netw Open. 2026 Apr 1;9(4):e263419. doi: 10.1001/jamanetworkopen.2026.3419.

ABSTRACT

IMPORTANCE: While numerous states have enacted laws to reduce access to firearms among high-risk individuals, the evidence regarding the associated outcome of reducing firearm suicide has been mixed, in part due to methodological limitations.

OBJECTIVE: To examine the association between state firearm laws and firearm-related suicide deaths across all 50 US states during the period from 1976 to 2024.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study, conducted in December 2025, used a difference-in-differences fixed-effects panel regression with Prais-Winsten correction applied to annual state-level data on firearm-related suicides from all 50 US states from January 1976 through December 2024. Sample data were obtained from the Centers for Disease Control and Prevention-maintained Web-based Injury Statistics Query and Reporting System.

EXPOSURES: Six firearm laws with prior evidence or theoretical plausibility of affecting risk of suicide: (1) required permits to purchase handguns; (2) waiting periods for firearm purchases; (3) laws requiring permits for concealed carry; (4) minimum age requirements; (5) extreme risk protection order laws; and (6) state permit requirements for gun dealers. Laws were modeled with a 2-year lag.

MAIN OUTCOMES AND MEASURES: The primary outcome was annual, age-adjusted, state-specific firearm suicide rate, and the negative control outcome was nonfirearm suicide rate. Models accounted for serial autocorrelation and heteroskedasticity in the data and adjusted for a range of sociodemographic covariates.

RESULTS: Across the study period, 2450 observations were collected. The mean overall suicide rate was 13.7 deaths per 100 000 with 7.9 deaths per 100 000 for firearm-related suicide and 5.8 deaths per 100 000 for non-firearm-related suicide. Firearm suicide rates varied 8-fold across states in 2024 (1.8 deaths per 100 000 in New York vs 15.1 deaths per 100 000 in Wyoming). Handgun permit laws (-6.7%; 95% CI, -9.7% to -3.7%), waiting periods (-12.5%; 95% CI, -22.1% to -1.7%), and requirements for a license for concealed carry (-8.9%; 95% CI, -13.1% to -4.8%) were significantly associated with decreases in firearm suicide rates but not with nonfirearm suicide rates. States with 1 (-8.1%; 95% CI, -11.4% to -4.7%), 2 (-12.5%; 95% CI, -16.3% to -8.5%), or all 3 (-25.3%; 95% CI, -34.2% to -15.2%) of these laws (handgun permit requirements, waiting periods, and concealed carry permits) had progressively lower firearm suicide rates.

CONCLUSIONS AND RELEVANCE: In this cross-sectional study analyzing data from 49 years and 50 states, permit-to-purchase requirements for handguns, waiting periods, and the requirement for a license for concealed carry were each independently and cumulatively associated with significantly lower firearm suicide rates.

PMID:41920544 | DOI:10.1001/jamanetworkopen.2026.3419

JAMA Netw Open. 2026 Apr 1;9(4):e264098. doi: 10.1001/jamanetworkopen.2026.4098.

ABSTRACT

IMPORTANCE: External beam radiation therapy (EBRT) for left-sided breast cancer historically increased the risk of cardiovascular disease (CVD). Advances in EBRT have reduced mean heart dose, but contemporary population-level data on long-term cardiovascular outcomes remain limited.

OBJECTIVE: To examine whether contemporary left-sided EBRT is associated with different long-term CVD risk than right-sided EBRT among women with breast cancer.

DESIGN, SETTING, AND PARTICIPANTS: This population-based, retrospective cohort study used linked administrative health databases in Ontario, Canada, to identify women who received EBRT after a diagnosis of unilateral breast cancer between April 1, 2002, and December 31, 2017. Patients were followed up through February 28, 2025, for most outcomes and December 31, 2022, for cause-specific mortality. Data analysis was completed in August 2023 (with revisions in August 2025).

EXPOSURE: Tumor laterality (left vs right).

MAIN OUTCOMES AND MEASURES: The primary outcome was hospitalization with a most responsible diagnosis of CVD. Secondary outcomes included all-cause mortality, cardiovascular mortality, hospitalizations for specific CVD diagnoses, coronary revascularization, and new diagnoses of ischemic heart disease, heart failure, and atrial fibrillation. Cumulative incidence functions estimated the 15-year risk of outcomes accounting for the competing risk of death; event rates per 100 person-years captured recurrent events.

RESULTS: Among 76 586 women (mean [SD] age, 59 [12] years; 38 427 [50.2%] with left-sided tumors) followed up for a median (IQR) of 10.9 (7.7-15.2) years, the 15-year cumulative incidence of first CVD hospitalization did not differ by laterality (left: 13.8% [95% CI, 13.4%-14.2%]; right: 13.5% [95% CI, 13.1%-13.9%]; P = .43). In women with preexisting CVD, new diagnoses of heart failure (10.2% [95% CI, 9.9%-10.6%] vs 9.6% [95% CI, 9.2%-10.0%]; P = .01) and ischemic heart disease (13.6% [95% CI, 13.2%-14.0%] vs 12.8% [95% CI, 12.4%-13.2%]; P = .03) were slightly more frequent after left-sided EBRT. The rate of CVD hospitalizations when including recurrent events was modestly higher for left-sided disease (1.72 vs 1.63 per 100 person-years; P = .006). Among women with preexisting CVD, there were no differences in all-cause mortality or recurrent CVD hospitalizations.

CONCLUSIONS AND RELEVANCE: In this cohort study of women treated with EBRT for breast cancer in the past 2 decades, left-sided breast cancer radiation therapy was associated with minimal increases in long-term cardiovascular risk. These findings suggest that contemporary photon-based EBRT techniques have substantially reduced the cardiovascular risk historically associated with left-sided breast cancer radiation therapy.

PMID:41920543 | DOI:10.1001/jamanetworkopen.2026.4098

JAMA Netw Open. 2026 Apr 1;9(4):e264122. doi: 10.1001/jamanetworkopen.2026.4122.

ABSTRACT

IMPORTANCE: During early months of the COVID-19 pandemic, presentations for acute myocardial infarction (AMI) declined significantly, and outcomes worsened. However, the full extent and long-term sequelae of changes in AMI epidemiology during the pandemic remain uncertain, as does whether these patterns differed by rurality.

OBJECTIVE: To describe the epidemiology of AMI-related hospitalizations, interventions, and outcomes among Medicare beneficiaries throughout the COVID-19 pandemic, focusing on differences in urban and rural populations.

DESIGN, SETTING, AND PARTICIPANTS: This retrospective cohort study included all Medicare fee-for-service beneficiaries with AMI between January 1, 2018, and December 31, 2023, in the analysis. Data were analyzed from March 19 to July 9, 2025.

EXPOSURES: Time period (prepandemic [January 1, 2018, to December 31, 2019], pandemic [January 1, 2020, to December 31, 2021], and postpandemic [January 1, 2022, to December 31, 2023]) and beneficiary-level rurality.

MAIN OUTCOMES AND MEASURES: The primary outcome was in-hospital death, defined as death within 1 day of discharge from the index episode of AMI. Secondary outcomes included death within 90 days of the index admission date and postdischarge outcomes. AMI episodes were defined as any emergency department (ED), observational, or inpatient stay with a primary ST-segment elevation myocardial infarction (STEMI) or non-STEMI (NSTEMI) diagnosis or a primary cardiogenic shock and secondary STEMI or NSTEMI diagnosis. Generalized estimating equations clustering on hospitals were used to compare pandemic and postpandemic outcomes with the prepandemic period, adjusting for beneficiary characteristics.

RESULTS: A total of 1 152 851 AMI episodes among 1 032 212 beneficiaries were identified between 2018 and 2023, of which 75.6% were NSTEMI. Most AMI episodes were among male (57.6%) beneficiaries aged 65 to 80 years (56.8%). The unadjusted quarterly incidence of AMI decreased from 17.2 to 13.0 episodes per million beneficiary days at risk (quarter 1 of 2018 to quarter 4 of 2023). In-hospital (adjusted odds ratio [AOR], 1.09; 95% CI, 1.07-1.11]) and 90-day mortality (AOR, 1.10; 95% CI, 1.09-1.12) increased during the pandemic and then returned to baseline or lower (AORs, 0.99 [95% CI, 0.97-1.01] and 0.96 [95% CI, 0.95-0.98], respectively). After the pandemic, beneficiaries were less likely to discharge to a skilled nursing facility (AOR, 0.67; 95% CI, 0.66-0.68), utilize the ED (adjusted incidence rate ratio [AIRR], 0.93; 95% CI, 0.92-0.94), or experience readmission (AIRR, 0.90; 95% CI, 0.90-0.92) within 90 days of their index episode of AMI. Patterns were largely similar by rurality.

CONCLUSIONS AND RELEVANCE: In this retrospective cohort study of fee-for-service Medicare beneficiaries, the incidence of AMI decreased during and after the pandemic. Beneficiaries experienced greater in-hospital and 90-day mortality during the pandemic. After the pandemic, in-hospital and 90-day mortality returned to baseline among micropolitan and rural beneficiaries and was lower than baseline among urban beneficiaries.

PMID:41920542 | DOI:10.1001/jamanetworkopen.2026.4122

JAMA Netw Open. 2026 Apr 1;9(4):e264340. doi: 10.1001/jamanetworkopen.2026.4340.

ABSTRACT

IMPORTANCE: Police pursuits account for approximately 1% of motor vehicle fatalities and expose suspects, officers, and bystanders to preventable harm. Contemporary population-based estimates of fatality trends and context-specific risk remain limited.

OBJECTIVE: To evaluate the characteristics of and trends in fatal vehicle crashes involving police pursuit in the US between 2009 and 2023 and the factors associated with these crashes.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used National Highway Traffic Safety Administration Fatality Analysis Reporting System (FARS) data between January 2009 and December 2023. These nationwide, population-based data included individuals involved in fatal motor vehicle crashes involving police pursuits. The data analysis was performed between August 2025 and February 2026.

EXPOSURES: Police motor vehicle pursuits, as documented in the FARS database.

MAIN OUTCOMES AND MEASURES: The primary outcomes were pursuit-related motor vehicle crash fatalities per year and associated factors, including urban vs rural settings, interstate vs noninterstate roads, US census regions, time of day, and day of the week. A negative binomial regression was used to model fatality rates controlling for population and crash frequency.

RESULTS: Between 2009 and 2023, there were 5425 fatal police pursuit-related crashes involving 8307 vehicles and 14 497 persons, resulting in 6352 deaths (mean [SD], 423 [84] deaths per year). Fatal crashes occurred most often in urban settings (3069 [57%]), at night (3794 [70%]), on noninterstate roads (4825 [89%]), and with documented speeding (4183 [77%]). Nonmotor vehicle occupants represented 270 fatalities (4%). When controlling for population and crashes, fatalities increased by 2% (95% CI, 1%-3%; P < .001) annually. Compared with the Northeast, fatalities were significantly higher in the South (difference, 336% [95% CI, 284%-395%]; P < .001), the Midwest (difference, 110% [95% CI, 84%-140%]; P < .001), and the West (difference, 95% [95% CI, 70%-123%]; P < .001).

CONCLUSIONS AND RELEVANCE: In this cross-sectional study of the FARS database from 2009 to 2023, police pursuit-related fatalities increased over 15 years. Higher rates of fatal crashes were concentrated in identifiable settings including nighttime hours, urban corridors, noninterstate roadways, and southern states. The findings support risk-based restrictions, mandatory national pursuit reporting, and investment in nonpursuit alternatives among US police.

PMID:41920540 | DOI:10.1001/jamanetworkopen.2026.4340

JAMA Dermatol. 2026 Apr 1. doi: 10.1001/jamadermatol.2026.0308. Online ahead of print.

ABSTRACT

IMPORTANCE: Human papillomavirus (HPV) infection has been associated with increased risk of cutaneous squamous cell carcinoma (cSCC), particularly in immunocompromised individuals such as organ transplant recipients. However, most studies have focused on β-HPV, with limited attention to α-HPV and γ-HPV. Moreover, little is known about these associations in Asian patients, and the concordance of HPV genotypes across different tissue sites and the anatomical distribution of HPV-positive cSCC lesions remain underexplored. This study aims to address these gaps.

OBJECTIVE: To characterize the spectrum and viral load of HPV genotypes across multiple tissue sites in an Asian cohort of organ transplant recipients and immunocompetent individuals.

DESIGN, SETTING, AND PARTICIPANTS: This multicenter case-control study spanning January 2018 to December 2024 included organ transplant recipients and immunocompetent patients with cSCC residing in Singapore. Tissue samples were collected from eyebrow hair, cSCC lesions, and cutaneous warts for DNA extraction. The data analysis was conducted from June 2024 to July 2025.

MAIN OUTCOMES AND MEASURES: The Illumina MiSeq platform was used to sequence 193 samples to determine the prevalence of α-HPV, β-HPV, and γ-HPV types in an Asian population and their anatomical distribution in patients with cSCC.

RESULTS: The cohort comprised 37 organ transplant recipients with cSCC and 48 immunocompetent patients with cSCC. The median (IQR) age of participants was 73.2 (67.6-81.0) years; 56 (65.9%) were male, and 29 (34.1%) were female, with 78 Chinese individuals (91.8%), 1 Malay individual (1.1%), 1 Indian individual (1.1%), and 5 (5.8%) grouped as having another race or ethnicity. Organ transplant recipients developed cSCC at a statistically significantly younger age than immunocompetent patients (median [IQR] age, 68.5 [64.0-71.0] years vs 79.5 [72.0-85.0] years, respectively), with a higher prevalence of viral warts (27 [73.0%] vs 14 [29.2%]) and greater wart body surface area involvement (median [IQR], 12% [0.5%-6.5%] vs 0% [0%-5%]). In total, 65 HPV genotypes were identified (38 β-HPV, 15 γ-HPV, 12 α-HPV). Site-specific distribution showed 56 genotypes in eyebrow hair, 30 in cSCC, and 17 in warts, with β-HPV exhibiting the greatest overlap across all 3 sites. β-HPV exhibited the highest viral loads across all tissue sites, compared to α-HPV and γ-HPV. Genotype-specific concordance between eyebrow hair and paired cSCC was low in both groups; among 55 patients, 5 (9.1%) had 1 or more concordant genotype.

CONCLUSIONS AND RELEVANCE: In this case-control study of Asian individuals with cSCC, there were differences in the distribution of HPV genotypes in Asian organ transplant recipients and immunocompetent patients, compared to published studies. Future larger studies mapping the prevalence and types of HPV in Asia will enhance understanding of risk factors and inform targeted prevention efforts in this part of the world.

PMID:41920535 | DOI:10.1001/jamadermatol.2026.0308

JMIR Hum Factors. 2026 Apr 1;13:e82703. doi: 10.2196/82703.

ABSTRACT

BACKGROUND: Alarm fatigue caused by frequent or false alarms poses a persistent threat to patient safety. Despite technological progress, alarm acoustics remain largely unchanged and are often perceived as disruptive. To address this, Philips redesigned its patient monitoring alarm sounds through a user-centered approach aimed at improving priority differentiation and reducing emotional strain.

OBJECTIVE: This study provides insights into human-technology interaction by examining anesthesia providers’ experiences with the original and updated alarms, with a focus on emotional responses, usability, and guidance for the user-centered design of future clinical alarm systems.

METHODS: This single-center qualitative study involved anesthesia providers who completed an online questionnaire before and after the implementation of the updated Philips alarms. Only those who completed the pre-implementation phase participated in the postimplementation phase. The questionnaire included 4 open-ended questions addressing perceptions of the current alarm sounds, suggestions for improvement, design expectations, and attitudes toward an alarm-free operating room. Responses were analyzed using thematic analysis to identify key usability and emotional response themes.

RESULTS: A total of 90 eligible anesthesia providers participated in the preimplementation phase, and 77 (85.6%) participated in the postimplementation phase. Positive emotional responses increased in the postimplementation phase, whereas concerns regarding alarm functionality also became more prominent. Before the introduction of the updated alarm sounds, participants predominantly called for softer sounds. Following implementation, the most frequently expressed concern shifted to the need for clearer prioritization of alarms. Across both phases, the primary expectation remained the alarms’ ability to effectively capture attention. The concept of an alarm-free operating room elicited concerns about increased workload and potential risks to patient safety.

CONCLUSIONS: The redesigned alarm sounds were perceived more positively in terms of emotional acceptance; however, they did not improve the recognition of alarm priority. The modest acoustic changes did not address the broader issue of alarm overload. Suggestions such as visual-only alerts for low-priority alarms show potential but must be balanced with patient safety standards. Future alarm development should combine user feedback with expert-driven and evidence-based approaches to improve both usability and clinical effectiveness.

PMID:41920525 | DOI:10.2196/82703

Interdiscip Sci. 2026 Apr 1. doi: 10.1007/s12539-026-00833-8. Online ahead of print.

ABSTRACT

Alzheimer’s disease (AD) is a progressive neurodegenerative disease of the nervous system, which has become an important public health issue attracting global attention. However, its exact causes and pathogenesis have not been fully elucidated, and the existing treatment methods and intervention measures have limited efficacy. Therefore, how to establish a scientific and efficient risk assessment mechanism has become the key to the prevention and treatment of AD. Aiming at this problem, this paper optimizes the convolutional neural network structure of DenseNet and proposes a Focal-DenseNet model that integrates the focal loss function for the risk assessment and diagnostic prediction of AD. First of all, preprocess the collected data to ensure the data quality for subsequent analysis. Secondly, establish the Focal-DenseNet model. Finally, use the model for training and testing. The test results show that the test set accuracy of the model reaches 98.98%, indicating that the model performs well. In addition, this paper also compares the proposed model with the DenseNet model without using the focal loss function and other common deep learning models (such as VGG16, ResNet, etc.). The results show that the model in this paper exhibits superiority in multiple performance indicators. In particular, it has achieved high scores in key indicators such as accuracy, AUC (the area under the receiver operating characteristic curve), precision, and recall. This study provides an efficient technical support for the early risk assessment of AD, and holds significant clinical application value and academic reference significance for improving the prevention and treatment level of AD and reducing the global public health burden.

PMID:41920507 | DOI:10.1007/s12539-026-00833-8

Spine Deform. 2026 Apr 1. doi: 10.1007/s43390-026-01351-9. Online ahead of print.

ABSTRACT

BACKGROUND: Early (within 30 days) postoperative complications following pediatric scoliosis surgery continue to be of clinical importance; however, reported risk factors demonstrate inconsistency across various studies. We conducted a systematic review and meta-analysis to quantify patient- and procedure-related predictors of 30-day postoperative outcomes complications.

METHODS: A systematic search of PubMed, Embase, Scopus, and Web of Science identified observational studies reporting risk factors for 30-day postoperative complications following scoliosis surgery. Adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were pooled using random-effects models. Heterogeneity was assessed with the I² statistic.

RESULTS: Ten observational studies comprising 113,082 patients met the inclusion criteria; nine reported adjusted estimates. Osteotomy was significantly associated with increased risk (OR = 1.40; 95% CI 1.25-1.55; I² = 0%), as was pelvic fixation (OR = 1.65; 95% CI 1.07-2.55; I² = 0%), with no observed heterogeneity across studies. In contrast, multiple medical and patient-related risk factors including cardiopulmonary, hematologic and metabolic comorbidities were not significantly associated with postoperative morbidity, although point estimates suggested a directionally elevated risk and were characterized by wide confidence intervals. Male sex and neuromuscular disorders were also not associated with a significant increase in risk.

CONCLUSIONS: Among the evaluated factors, complex surgical techniques involving osteotomy and pelvic fixation were the most consistent predictors of early postoperative complications within the contemporary literature. Preoperative patient comorbidities showed variable, generally non-significant associations, underscoring the need for standardized risk definitions and prospective multicenter studies to better predict outcomes in the early postoperative period.

PMID:41920499 | DOI:10.1007/s43390-026-01351-9

Spine Deform. 2026 Apr 1. doi: 10.1007/s43390-026-01341-x. Online ahead of print.

ABSTRACT

PURPOSE: Reduced trunk muscle quantity or quality, commonly assessed using imaging-based metrics, has been associated with adverse outcomes in spine surgery. However, evidence specific to adult spinal deformity (ASD) surgery remains heterogeneous and methodologically limited. This systematic review aimed to synthesize existing muscle metrics and surgical outcomes in patients undergoing ASD surgery.

METHODS: A systematic search of PubMed and EMBASE was performed to identify studies evaluating imaging-based muscle quantity or quality in adult patients (≥ 18 years) undergoing thoracolumbar corrective surgery. Eligible studies employed radiographic measures of muscle status. Outcomes of interest were postoperative complications, reoperation, readmission, length of stay (LOS), and patient-reported outcome measures (PROMs). Data were extracted based on study design, patient demographics, surgical characteristics, and follow-up. Meta-analysis was conducted using Review Manager 5.4, with odds ratios (ORs) for dichotomous outcomes and mean differences for continuous outcomes.

RESULTS: Fifteen studies met the eligibility criteria in this review. Meta-analysis demonstrated substantial heterogeneity across studies, particularly for proximal junctional kyphosis (PJK) and length of stay (LOS), limiting the interpretability of pooled estimates. No statistically significant differences were observed for reoperation, readmission, wound complications, or patient-reported outcomes. A modest increase in LOS was observed in patients with lower muscle metrics, although this finding was highly heterogeneous and may reflect residual confounding.

CONCLUSION: Current evidence examining imaging-based muscle metrics in ASD surgery is characterized by substantial heterogeneity and predominantly retrospective study designs. While reduced muscle quantity or quality may be associated with postoperative outcomes, existing data do not support definitive conclusions or causal inferences. These findings highlight the need for standardized muscle assessment methods and prospective studies incorporating validated sarcopenia criteria to better clarify the clinical relevance of muscle health in ASD surgery.

PMID:41920498 | DOI:10.1007/s43390-026-01341-x

J Comput Neurosci. 2026 Apr 1. doi: 10.1007/s10827-026-00927-8. Online ahead of print.

NO ABSTRACT

PMID:41920486 | DOI:10.1007/s10827-026-00927-8