Category: Statistics

Cornea. 2026 Jun 3. doi: 10.1097/ICO.0000000000004080. Online ahead of print.

ABSTRACT

PURPOSE: The aim of this study was to compare the outcomes of 2-piece mushroom keratoplasty (MK) between eyes at high risk and eyes at low risk of immunologic rejection.

METHODS: This is a retrospective interventional case series including 475 eyes from 475 patients who underwent 2-piece MK, of whom 146 (30.7%) were classified as high risk of rejection (group 1) and 329 (69.3%) as low risk (group 2). Outcome measures included corrected distance visual acuity, endothelial cell count, endothelial cell loss, graft survival rate, and postoperative complications.

RESULTS: There was no significant difference in endothelial cell count between groups at 10 years after transplantation (754.7 ± 316.7 [95% CI: 695.1-956.6] vs. 892.3 ± 298.8 [95% CI: 757.8-969.4] cells/mm2, P = 0.088, respectively). Although corrected distance visual acuity was better in the low-risk group at 1 year (0.37 ± 0.30 [95% CI: 0.32-0.42] vs. 0.25 ± 0.23 [95% CI: 0.23-0.28], P < 0.001) and 5 years (0.20 ± 0.26 [95% CI: 0.15-0.27] vs. 0.14 ± 0.21 [95% CI: 0.10-0.17], P = 0.044), this difference was no longer statistically significant at 10 years (0.25 ± 0.34 [95% CI: 0.13-0.39] vs. 0.18 ± 0.25 [95% CI: 0.12-0.24] logMAR, P = 0.269, respectively). Postoperative complication rates were comparable, except for infections, which occurred more frequently in group 1 (7.5% vs. 1.5%, P < 0.001). Graft survival rates at 10 years were 80.3% for group 1 (n = 24) and 90.7% for group 2 (n = 50) (P = 0.153).

CONCLUSIONS: Two-piece MK provides consistent long-term outcomes in both high-risk and low-risk eyes, showing comparable visual, endothelial, and survival results. These findings highlight the potential of MK to reduce the impact of preoperative immunologic risk and expand its role in complex corneal transplantation.

PMID:42284083 | DOI:10.1097/ICO.0000000000004080

Am Surg. 2026 Jun 12:31348261451718. doi: 10.1177/00031348261451718. Online ahead of print.

ABSTRACT

BackgroundTransportation access, a recently recognized key social determinant of health, has an understudied relationship to gun violence. We hypothesized that poor transportation access would be associated with increased firearm violence.MethodsThis cross-sectional analysis examined indices of transportation access using University of Minnesota Access Across America (UMAAA) data, which assigns separate rankings based on transit, auto, biking, and walking access scores. Violent firearm injuries and deaths were obtained from the Gun Violence Archive (GVA). Data for educational attainment, income, employment, and housing characteristics came from the US Census’s American Community Survey. Data was aggregated from 2019 to 2021 by the 56 largest metropolitan statistical areas (MSA) in the US. Negative binomial Poisson regression models (univariate and multivariate) were used to examine the association between transportation indices and violent firearm injuries and deaths.ResultsNegative Binomial Poisson regression analysis found that lowered overall transit and auto access scores were associated with slightly lower risk of violent firearm injury and death. This relationship persisted even when adjusted for Gini and ICE. Poverty, unemployment, Gini, and the percentage of persons without a vehicle were directly associated and had higher effects compared to AAA score.DiscussionThis study does not support a strong protective effect of transportation access on firearm violence rates, instead suggesting a relatively small, inverse relationship. These findings underscore the complex nature of urban violence, which is influenced by multiple socioeconomic and community factors. Further studies are needed to determine how injury preventive strategies can target important determinants of firearm injury.

PMID:42284064 | DOI:10.1177/00031348261451718

JAMA Netw Open. 2026 Jun 1;9(6):e2618232. doi: 10.1001/jamanetworkopen.2026.18232.

ABSTRACT

IMPORTANCE: Guidelines suggest administering at least 30 mL/kg of initial fluid to patients with sepsis-induced hypoperfusion. However, there is uncertainty regarding the benefits of fluid resuscitation in patients with severe cardiac or kidney comorbidities or intermediate elevation of lactate level (18.0-36.0 mg/dL).

OBJECTIVE: To evaluate the association of 30 mL/kg or more of fluid administered within 6 hours of hospital arrival with 30-day mortality across key target populations with community-onset sepsis.

DESIGN, SETTING, AND PARTICIPANTS: This cohort study included adults hospitalized for community-onset sepsis in 67 hospitals in the Michigan Hospital Medicine Safety Consortium (discharge dates from December 2021 to January 2025) who had an indication for fluid resuscitation (ie, hypotension or lactate level of 18.0 mg/dL or greater) within 3 hours of hospital arrival. Data were analyzed from November 26, 2024, to November 16, 2025.

EXPOSURE: Receipt of at least 30 mL/kg vs less than 30 mL/kg fluid in the first 6 hours after hospital arrival. Fluid volume included all crystalloid fluid and blood products.

MAIN OUTCOMES AND MEASURES: Association between administration of 30 mL/kg or more of fluid within 6 hours of hospital arrival and 30-day mortality using weighted regression models adjusted for patient characteristics. Target populations were defined by (1) fluid indication: hypoperfusion (hypotension or lactate level >36.0 mg/dL) vs intermediate lactate elevation (18.0-36.0 mg/dL) and (2) presence of severe comorbidities that might increase risk of fluid overload (left ventricular ejection fraction <40%, severe-to-critical aortic stenosis, or end-stage kidney disease). Secondary analyses used adjusted logistic regression models with restricted cubic spline terms to evaluate associations of fluid volume administered with mortality.

RESULTS: Among 43 321 patients hospitalized for community-onset sepsis, 25 481 (58.8%) had an indication for fluid resuscitation and were included in the study (median age, 71 years [IQR, 61-80 years]; 50.5% male; 37.0% with body mass index >30.0, calculated as weight in kilograms divided by height in meters squared). A total of 12 943 (50.8%) had hypoperfusion without severe comorbidities; 1741 (6.8%), hypoperfusion with severe comorbidities; 9974 (39.1%), intermediate lactate elevation without severe comorbidities; and 823 (3.2%), intermediate lactate elevation with severe comorbidities. Administration of 30 mL/kg or more of fluid vs less than 30 mL/kg was associated with lower adjusted 30-day mortality rates in patients with hypoperfusion without severe comorbidities (26.0% [95% CI, 24.9%-27.2%] vs 30.4% [95% CI, 28.8%-32.0%]; adjusted absolute difference [diff], -4.4 percentage points [pp] [95% CI, -6.1 to -2.7 pp]) and intermediate lactate elevation without severe comorbidities (12.0% [95% CI, 10.6%-13.5%] vs 13.9% [95% CI, 12.9%-14.8%]; diff, -1.8 pp [95% CI, -3.6 to -0.1 pp]). For patients with hypoperfusion and severe cardiac or kidney comorbidities, the association between 30-day adjusted mortality and receiving 30 mL/kg or more of fluid vs less than 30 mL/kg was not statistically significant (34.7% [95% CI, 30.8%-38.6%] vs 38.8% [95% CI, 35.8%-41.8%]; diff, -4.1 pp [95% CI, -9.0 to 0.8 pp]), although spline models indicated decreasing mortality with fluid resuscitation of 30 mL/kg or more of fluid.

CONCLUSIONS AND RELEVANCE: In this cohort study of patients with community-onset sepsis, initial administration of 30 mL/kg or more of fluid was associated with lower 30-day mortality among patients who had either hypoperfusion or intermediate lactate elevation without severe cardiac or kidney comorbidities. The findings suggest that broader application of at least 30 mL/kg of initial fluid resuscitation for sepsis in patients with hypoperfusion and cardiac or kidney comorbidities or intermediate lactate elevation may reduce sepsis-related mortality.

PMID:42284053 | DOI:10.1001/jamanetworkopen.2026.18232

JAMA Netw Open. 2026 Jun 1;9(6):e2618248. doi: 10.1001/jamanetworkopen.2026.18248.

ABSTRACT

IMPORTANCE: Evidence continues to mount that ultraprocessed food (UPFs) may contribute to diet-related chronic diseases, including cardiovascular disease, obesity, diabetes, and some cancers. In the US, these conditions account for over 1 million deaths each year and cost the economy over $1.1 trillion in health care costs and lost productivity. This disease burden has intensified interest about the potential health impacts of UPF.

OBJECTIVE: To characterize the portfolio of research on UPF supported by the National Institutes of Health (NIH).

EVIDENCE REVIEW: All NIH-supported research projects included in this Special Communication were awarded between fiscal years 2016 and 2025. An initial 520 projects were identified using iSearch and were manually reviewed for a focus on UPF. All projects that met the inclusion criteria were coded for study design, health conditions, and UPF-specific topics. The research projects were collected in June 2025, and a review and analysis of the data using descriptive statistics were conducted between June 1 and December 31, 2025.

FINDINGS: A total of 81 NIH-supported UPF-specific projects were identified. Of these 81 projects, 58 studied various health conditions in relation to UPF consumption and 29 were interventional research, some of which examined the effects of food policies or lifestyle interventions on health outcomes. Forty-nine projects were coded as basic research, 32 of which examined human biological mechanisms of action from UPF exposure, and 5 studies included hybrids of human and vertebrate animal studies. The total number of NIH-supported projects has increased from 8 projects in fiscal year 2016 to 22 projects in fiscal year 2025, and funding for UPF-specific research increased from $3.3 million in fiscal year 2016 to $12 million in fiscal year 2025.

CONCLUSIONS AND RELEVANCE: NIH support for UPF-specific research has grown since fiscal year 2016. Most research projects have focused on the connection between obesity and UPF intake; fewer have used interventions to address food access and/or insecurity and food environment. This is a key research gap given the role of food environments as a driver of UPF consumption in different populations and their importance in improving health outcomes. Another critical research gap is mechanistic research to elucidate causal pathways linking UPF consumption to adverse health outcomes and validated biomarkers of UPF exposure and disease.

PMID:42284052 | DOI:10.1001/jamanetworkopen.2026.18248

Fam Med. 2026 May;58(5):359-362. doi: 10.22454/FamMed.2026.979674.

ABSTRACT

BACKGROUND AND OBJECTIVES: Increases in graduate medical education application volume led to the introduction of preference signaling-a tool within the Electronic Residency Application Service that allows applicants to signal their sincere interest to a limited number of programs. This study aims to evaluate how family medicine program directors used preference signals during the 2023-2024 recruitment season, the first year this tool was available to family medicine programs. Understanding program director perspectives on preference signal utility is crucial for determining the tool’s impact on the residency selection process.

METHODS: Data were collected through the Council of Academic Family Medicine Educational Research Alliance survey, which was distributed to Accreditation Council for Graduate Medical Education-accredited US family medicine residency program directors and included questions on preference signaling. We performed statistical analysis using χ2 testing and multivariable logistic regression to assess the association between preference signal use and aspects of resident recruitment.

RESULTS: The overall response rate to the question set was 308/767 (40.2%). Program directors generally used preference signals as a component of holistic review, which did not supersede factors such as applicant rotation experiences and geographic location preferences as influential to application review. Overall, preference signals did not significantly influence interview offers or applicant ranking.

CONCLUSIONS: Preference signals have been incorporated into the family medicine residency application review process but did not become a primary determinant in applicant selection during family medicine’s first year of utilization. Traditional factors such as rotation performance and geographic preference remain highly influential. Further research is needed to optimize the use of preference signals in family medicine residency recruitment.

PMID:42284041 | DOI:10.22454/FamMed.2026.979674

Fam Med. 2026 Apr;58(4):280-285. doi: 10.22454/FamMed.2026.317077.

ABSTRACT

BACKGROUND AND OBJECTIVES: The 2022 transition of USMLE Step 1 scoring to pass/fail altered a key metric used by programs to select students to interview for residency. This study explores family medicine clerkship directors’ (FMCDs’) perceptions of how students now distinguish themselves, particularly in relation to clerkship grading methodologies.

METHODS: Ten questions were included in the 2024 Council of Academic Family Medicine Educational Research Alliance survey, distributed to 173 FMCDs. Items investigated perceptions of student distinction, stress, and grading practices. Statistical analyses included descriptive statistics, Kruskal-Wallis tests, and Wilcox signed-rank tests.

RESULTS: Of the 83 respondents (48% response rate), 59% reported using pass/fail grading in the preclerkship phase, while only 22% used pass/fail grading in the clerkship phase. A majority (58%) indicated no changes to clerkship grading systems post-2022, though 20% had changed and 22% were considering changes, predominantly toward less-tiered methodologies. Regarding the impact of Step 1 changes on the students’ ability to distinguish themselves, 37% perceived harm, 14% benefit, and 48% neutrality. Despite this finding, 78% of FMCDs perceived that students were more stressed about distinguishing themselves. No significant associations were found between grading methodology and perceptions of distinction or stress.

CONCLUSIONS: FMCDs perceived increased student stress following the Step 1 pass/fail transition, yet largely believe that students still can distinguish themselves. Neither tiered nor pass/fail grading was viewed as a definitive solution. These findings underscore the need for standardized, competency-based assessment and clearer communication of distinguishing features in residency applications.

PMID:42284029 | DOI:10.22454/FamMed.2026.317077

J Med Internet Res. 2026 Jun 12;28:e64070. doi: 10.2196/64070.

ABSTRACT

BACKGROUND: Noncommunicable diseases (NCDs) account for over 70% of global deaths, with hypertension and diabetes serving as major contributors. The COVID-19 pandemic disrupted traditional health care services for NCDs and highlighted telehealth as a crucial alternative. Telehealth-encompassing synchronous and asynchronous electronic communication to deliver clinical services remotely-can overcome geographical barriers and enhance patient engagement. However, telehealth usability among health care professionals (HCPs) remains under-studied across low-, middle-, and high-income countries.

OBJECTIVE: This study aimed to examine which telehealth engagement patterns, technical infrastructure factors, and user profiles were most strongly associated with usability among HCPs and to descriptively compare these across 4 diverse countries: Brazil (high- to middle-income country), Ghana (low- to middle-income country), Honduras (low- to middle-income country), and the United Kingdom (high-income country).

METHODS: A multinational cross-sectional survey was conducted with 290 HCPs across 4 countries. Participants completed the System Usability Scale and provided data on telehealth engagement (eg, frequency, duration, and number of systems used), technical infrastructure (connection stability and support satisfaction), and their user profile (demographics, job role, and training received). Descriptive statistics summarized these patterns and usability scores. Multiple linear regression with bootstrap-based sensitivity analyses identified factors associated with telehealth usability. Given the nonprobability design, no formal inferential comparisons were made between countries. Instead, observed patterns were reported descriptively.

RESULTS: Higher telehealth usability scores were associated with greater connection stability (b=5.06, 95% CI 3.06-7.05), higher satisfaction with online support information (b=5.02, 95% CI 3.27-6.75), more frequent use (b=3.05, 95% CI 1.36-4.73), longer duration of use (b=1.59, 95% CI 0.49-2.68), and being a physician by profession (b=3.82, 95% CI 0.23-7.40). Average usability scores were highest among users in Ghana (mean 79.75, SD 14.19) and the United Kingdom (mean 79.00, SD 14.71), followed by Brazil (mean 72.01, SD 14.62) and Honduras (mean 63.09, SD 15.57). According to System Usability Scale guidelines, scores corresponded to “good” usability for users in Ghana, the United Kingdom, and Brazil and were below the “good” threshold for users in Honduras. While most users in Ghana (97/111, 87.4%), Honduras (31/38, 81.6%), and Brazil (57/80, 70.4%) reported using only 1 telehealth system, two-thirds of UK users (40/60, 66.7%) reported using 2 or more systems. User profiles also varied; prepandemic use was highest in Ghana (84/111, 75.7%) and lowest in Honduras (7/38, 18.4%). Other engagement patterns across countries were reported.

CONCLUSIONS: Telehealth usability is driven by technical infrastructure reliability, a robust online support infrastructure, and an “experience effect” from frequent and long-term engagement. Descriptive differences in engagement patterns and infrastructure highlight the need for tailored strategies to address setting-specific challenges. These are essential to optimize telehealth integration and improve health care outcomes for patients with NCDs worldwide.

PMID:42284018 | DOI:10.2196/64070

J Relig Health. 2026 Jun 12. doi: 10.1007/s10943-026-02699-y. Online ahead of print.

ABSTRACT

Mental health disparities among Black adults remain a critical public health issue; however, few studies have examined gender differences in perceptions of faith-based mental health hub programs designed to reduce these disparities. Using an exploratory survey design, this study assessed gender differences in perceptions of service utilization, access, engagement, cultural humility, and satisfaction with a faith-based mental health hub programs. This program links, refers, and provides psychoeducation to adults with mental health needs through Community Health Workers. A descriptive analysis and a Mann-Whitney U-test (non-parametric test) were used. All program participants were invited to take part in the study, and we achieved a 65% response rate. Among Black respondents (N = 231), 79% were women, 52.1% were college-educated, and the majority were 46-55 years old (26.4%). Results revealed a statistically significant difference between men and women in access to services only (U = 3831.50, Z = -2.79, p = 0.005, r = 0.18). Women reported significantly higher access to services than men. In general, both groups reported high levels of satisfaction with services, perceived cultural humility, and strong engagement in and utilization of the program, with no significant differences noted. Findings underscore generally positive perceptions among Black adults regarding faith-based mental health hub programs. However, the underrepresentation and lower reported perceptions of men and emerging adults highlight an opportunity to expand outreach efforts. Further, the high proportion of participants with college degrees suggests that educational attainment does not eliminate barriers to care.

PMID:42284005 | DOI:10.1007/s10943-026-02699-y

Clin Exp Nephrol. 2026 Jun 12. doi: 10.1007/s10157-026-02896-9. Online ahead of print.

NO ABSTRACT

PMID:42283989 | DOI:10.1007/s10157-026-02896-9

J Neurooncol. 2026 Jun 12;178(2):61. doi: 10.1007/s11060-026-05661-w.

ABSTRACT

PURPOSE: To estimate the pooled incidence of symptomatic radiation necrosis (RN) following intracranial brachytherapy and explore associations between clinical and technical variables and RN risk.

METHODS: PubMed, Embase, Web of Science, and the Cochrane Library were searched from 1954 – 2024 in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Eligible human studies reported symptomatic RN after intracranial brachytherapy. RN ascertainment was secondarily classified as clear, partial, or unclear based on reproducibility of diagnostic criteria. A random-effects generalized linear mixed model was used to pool incidence rates, with subgroup analyses by tumor type, implant technique, isotope, dose rate, and prior radiation. Study quality was assessed using National Institutes of Health tools.

RESULTS: Eighty-three studies encompassing 3,666 patients were included. The pooled incidence of symptomatic RN was 5.67% (95% CI: 3.84%-8.29%; I² = 75.7%). RN ascertainment was clear in 25 studies, partial in 43, and unclear in 15. In sensitivity analyses, pooled symptomatic RN incidence was 8.71% among studies with clear RN definitions and 7.48% among studies with clear or partial definitions. Subgroup RN rates were 3.72% for low-grade gliomas, 8.44% for high-grade gliomas, 2.07% for brain metastases, and 8.98% for meningiomas. Isotope-specific rates were 5.83% for Iodine-125, 7.61% for Iridium-192, and 2.07% for Cesium-131. No statistically significant subgroup differences were observed.

CONCLUSIONS: Symptomatic RN occurs in approximately 5.7% of patients following intracranial brachytherapy, within the range reported for other focal radiation therapies. Subgroup findings were not statistically significant and should be interpreted in the context of clinical and methodological heterogeneity.

PMID:42283986 | DOI:10.1007/s11060-026-05661-w