Annals of Epidemiology最新文献

{"title":"Towards robust real-time surveillance of severe acute respiratory infections: Exploring the potential of an existing national intensive care unit registry","authors":"Christine G.J.I. van Straten ,&nbsp;Jordy Gaspersz ,&nbsp;Nicolette F. de Keizer ,&nbsp;Cornelis H. van Werkhoven ,&nbsp;Arianne B. van Gageldonk-Lafeber ,&nbsp;Dylan W. de Lange ,&nbsp;Dave A. Dongelmans ,&nbsp;Ferishta Bakhshi- Raiez ,&nbsp;Liselotte van Asten","doi":"10.1016/j.annepidem.2025.06.005","DOIUrl":"10.1016/j.annepidem.2025.06.005","url":null,"abstract":"<div><h3>Objectives</h3><div>Although many countries lack Severe Acute Respiratory Infections (SARI) surveillance, establishing such systems is crucial for pandemic preparedness. Adjusting existing registries designed for other purposes may be viable. We investigated the potential of the Dutch National Intensive Care Evaluation (NICE) registry, originally designed to monitor intensive care unit (ICU) care quality, for near real-time SARI surveillance by determining data timeliness.</div></div><div><h3>Study design</h3><div>A descriptive study was performed with retrospective ICU admissions (2012–2019).</div></div><div><h3>Methods</h3><div>The lag time between SARI admissions and their data upload was determined. Incident SARI cases, SARI proportion relative to all medical admissions, and SARI patient mortality were examined at various lag times and compared to the complete data.</div></div><div><h3>Results</h3><div>ICUs uploaded data at varying intervals. The Spearman correlation coefficient between the complete and incomplete ICU incident cases increased the most between 2 and 6 weeks of lag time and increased further with longer lag time. Data uploaded within two weeks fluctuated, with no uploaded data in 13 % of the weeks.</div></div><div><h3>Conclusion</h3><div>The current uploading delays render the NICE registry unsuitable for near real-time SARI surveillance. Investing in timely and preferably automated data exchange mechanisms can enhance existing registry purposes. This may benefit other countries as well.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"108 ","pages":"Pages 47-55"},"PeriodicalIF":3.3,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144263434","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Health inequity learning module: Zadeh et al (2023), Epidemiological approaches to multivariable models of health inequity: A study of race, rurality, and occupation during the COVID-19 pandemic","authors":"Jeb Jones PhD, MPH, MS, Associate Professor","doi":"10.1016/j.annepidem.2025.04.007","DOIUrl":"10.1016/j.annepidem.2025.04.007","url":null,"abstract":"<div><div>Educational Engagement Modules (EEMs) are teaching materials for educators and students that facilitate a deeper understanding of key epidemiological methods and concepts. Each EEM poses a series of questions using a recently published paper in Annals to further understanding of a specific study design and to encourage critical thinking and careful evaluation. This EEM focuses on methods for studying health inequity and references the following article: Zadeh H, Curran M, Del Castillo N, Morales C, Dukes K, Martinez D, Salinas JL, Bryant R, Bojang M, Carvour ML. Epidemiological approaches to multivariable models of health inequity: A study of race, rurality, and occupation during the COVID-19 pandemic. Ann Epidemiol. 2024 Jun;94:42–48. doi: 10.1016/j.annepidem.2024.04.008. Epub 2024 Apr 19. PMID: 38642626; PMCID: PMC11326713.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"107 ","pages":"Pages 68-70"},"PeriodicalIF":3.3,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144254291","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The definition of treatment assignment in observational emulations of target trials – an empirical examination in the Swedish Primary Care Cardiovascular Database","authors":"Anna B.C. Humphreys ,&nbsp;Anthony A. Matthews ,&nbsp;Jessica C. Young ,&nbsp;Anita Berglund ,&nbsp;Bertil Lindahl ,&nbsp;Björn Wettermark ,&nbsp;Issa J. Dahabreh ,&nbsp;Thomas Kahan ,&nbsp;Miguel A. Hernán","doi":"10.1016/j.annepidem.2025.06.003","DOIUrl":"10.1016/j.annepidem.2025.06.003","url":null,"abstract":"<div><h3>Background</h3><div>In randomized trials, the intention-to-treat effect is the effect of assignment to treatment strategies. The concept of assignment may not be clearly defined when using observational data to emulate a target trial.</div></div><div><h3>Aims</h3><div>We aimed to assess the practical implications of using data on prescription versus dispensation as analogues of treatment assignment in observational analyses.</div></div><div><h3>Methods</h3><div>We used the primary care-derived Swedish Primary Care Cardiovascular Database of individuals with newly diagnosed hypertension between 2006 and 2014 and linked registers. We compared the effect of two antihypertensive drug classes on the five-year risk of cancer and ischemic heart disease. Treatment assignment was first mapped using prescription data, and then dispensation data. With unique confounding structures, we sequentially adjusted for different classes of risk factor due to uncertainty over the choice of relevant confounders for prescription vs. dispensation.</div></div><div><h3>Results</h3><div>7770 individuals were eligible when assignment was defined using prescription compared with 5964 when defined using dispensation. For both cancer and ischemic heart disease outcomes, both higher and lower relative risks of the outcome were consistent with our data. Effect estimates did not vary with the choice of prescription or dispensation data as analogues of assignment, nor with sequential adjustment for class of risk factor.</div></div><div><h3>Conclusion</h3><div>The mapping of prescription or dispensation data to treatment assignment influences the size and characteristics of the study population and the structure of confounding. However, we found no clear numerical differences in effect estimates in this study. Further investigation is required in other settings.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"108 ","pages":"Pages 56-62"},"PeriodicalIF":3.3,"publicationDate":"2025-06-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144286953","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Prematurity at birth and risk of self-injury, overdose or death in adolescence and early adulthood: A population-based cohort study","authors":"Joel G. Ray MD MSc ,&nbsp;Neda Pirouzmand BHSc ,&nbsp;Alison L. Park MSc ,&nbsp;Peter C. Austin PhD ,&nbsp;Simone Vigod MD MSC ,&nbsp;Sonia Grandi PhD ,&nbsp;Howard Berger MD ,&nbsp;Eyal Cohen MD MSc","doi":"10.1016/j.annepidem.2025.06.002","DOIUrl":"10.1016/j.annepidem.2025.06.002","url":null,"abstract":"<div><h3>Background</h3><div>Preterm birth may adversely affect cortical brain development, and subsequent cognition in childhood and adolescence. Such abnormalities in brain development are correlated with suicidal and non-suicidal self-injury and persistent substance use, and to unintentional injury as well.</div></div><div><h3>Objective</h3><div>To evaluate the risk of self-injury, overdose, and all-cause mortality among adolescents and young adults – each in relation to preterm birth.</div></div><div><h3>Design</h3><div>Population-based cohort study in Ontario, Canada, where there is universal healthcare for all.</div></div><div><h3>Setting</h3><div>All of Ontario, Canada, where there is universal healthcare.</div></div><div><h3>Participants</h3><div>All youths born in Ontario from 1992 to 2012, and alive at age 10 years.</div></div><div><h3>Exposures</h3><div>Degree of prematurity in 2-week earlier increments, as well as preterm birth &lt; 37 weeks’ gestation.</div></div><div><h3>Outcomes</h3><div>i) Primary composite of an ED visit or hospitalization for intentional self-injury, poisoning or overdose, or all-cause mortality; and ii) co-primary composite of intentional self-injury, poisoning or overdose – each assessed between ages 10–24 years.</div></div><div><h3>Results</h3><div>Among 2352,563 youths in the cohort, 151,182 (6.4 %) were born preterm. Adjusted hazard ratios (aHR) for i) self-injury, overdose, or all-cause mortality was 1.03 (95 % CI 1.01–1.05) per 2-week earlier gestational age at birth; and ii) 1.01 (95 % CI 0.99–1.03) for self-injury or overdose. Compared to term births, preterm birth &lt; 37 weeks’ gestation was associated with a higher aHR for i) self-injury, overdose or all-cause mortality (1.09, 95 % CI 1.05–1.12); and ii) self-injury or overdose (1.06, 95 % CI 1.02–1.10).</div></div><div><h3>Conclusion</h3><div>Preterm birth may be associated with a slightly higher risk of all-cause mortality and, possibly, self-injury and overdose among adolescents and young adults. Ongoing efforts might focus on the health of youths who were born prematurely as they enter their adolescent years.</div></div><div><h3>Primary funding source</h3><div><span>Canadian Institutes of Health Research</span>.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"108 ","pages":"Pages 34-42"},"PeriodicalIF":3.3,"publicationDate":"2025-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144242027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Trends in the epidemiology of extended-spectrum beta-lactamase-producing Enterobacterales – Boulder County, Colorado, 2020–2022","authors":"Alyssa Beck ,&nbsp;Jennifer Driscoll ,&nbsp;Karlie Hoetzer ,&nbsp;Deborah Hindman ,&nbsp;Keith Rawls ,&nbsp;Janell ,&nbsp;Helen Johnston ,&nbsp;Christopher A. Czaja","doi":"10.1016/j.annepidem.2025.06.001","DOIUrl":"10.1016/j.annepidem.2025.06.001","url":null,"abstract":"<div><h3>Purpose</h3><div>Extended-spectrum beta-lactamase-producing Enterobacterales (ESBL-E) are serious antibiotic resistance threats. We describe trends in epidemiological and clinical characteristics of ESBL-E cases in Boulder County, Colorado from 2020 to 2022.</div></div><div><h3>Methods</h3><div>The Colorado Department of Public Health and Environment (CDPHE) and Boulder County Public Health conduct population-based surveillance for ESBL-E as part of the Emerging Infections Program-Healthcare-Associated Infections Community Interface. Trained epidemiologists abstracted data from patient medical records from the year prior to incident specimen collection through 30 days after.</div></div><div><h3>Results</h3><div>There were 1036 ESBL-E cases. Annual case counts increased from 2020 to 2022 (256, 316, and 464, respectively). The rate of ESBL-E infections per 100,000 population increased (77.4 in 2020, 96.6 in 2021, and 141.7 in 2022). Most samples were collected from outpatient settings (n = 812, 92.2 %) and <em>Escherichia coli</em> was the most common organism identified (n = 789, 89.6 %). Most case-patients had a lower urinary tract infection (UTI) (n = 757, 85.9 %), which increased by year (p = 0.03).</div></div><div><h3>Conclusions</h3><div>ESBL-E counts and incidence rates in Boulder County increased from 2020 to 2022. Cases most commonly presented with <em>E. coli</em> urinary tract infections in the outpatient setting. Strategies to combat antibiotic resistance should target ESBL-E in community settings.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"108 ","pages":"Pages 43-46"},"PeriodicalIF":3.3,"publicationDate":"2025-06-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144227482","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Geographical distribution disparities and prediction of health satisfaction among middle-aged and elderly adults in China: An analysis based on national data","authors":"Na Dong ,&nbsp;Xiaohan Yi ,&nbsp;Lijun Mao ,&nbsp;Biying Wang ,&nbsp;Manoj Sharma ,&nbsp;Lei Si ,&nbsp;Guoqun Xie ,&nbsp;Xianglong Xu","doi":"10.1016/j.annepidem.2025.05.012","DOIUrl":"10.1016/j.annepidem.2025.05.012","url":null,"abstract":"<div><h3>Background</h3><div>Health satisfaction among middle-aged and elderly adults has become a critical public health concern in China's aging society. Understanding geographical disparities in health satisfaction and developing prediction models are essential for targeted healthcare interventions and resource allocation.</div></div><div><h3>Methods</h3><div>Our study conducted a cross-sectional analysis of 16,231 participants aged 45 and above from the China Health and Retirement Longitudinal Study 2015. We analysed the provincial spatial distribution of health satisfaction in China. We developed conventional logistic regression (LR), random forest (RF), gradient boosting machine (GBM), XGBoost, and a stacking ensemble model (SEM) to predict health satisfaction and investigate social and biological determinants. We used the SHapley Additive exPlanation (SHAP) method to interpret our machine learning predictive models.</div></div><div><h3>Results</h3><div>Our analysis revealed significant geographical disparities in health satisfaction. The health satisfaction rate was 74.0 %, with regional variations: high in Xinjiang and Shanghai (&gt;80 %), low in some provinces (60–70 %), and moderate in the remaining provinces and municipalities (70–80 %). The AUCs of LR, RF, GBM, XGBoost, and SEM were all around 0.8. SHAP analysis revealed demographics (e.g., age), behavioural factors (e.g., night sleep duration), health-related factors (e.g., troubling body pain, self-expectations of health status, depression, heart problems and stomach or other digestive system diseases) and biological factors (e.g., self-reported distance vision status, self-reported near vision status, self-reported hearing status and MCV) as important predictors of health satisfaction in middle-aged and elderly adults.</div></div><div><h3>Conclusion</h3><div>The findings highlight substantial geographical inequalities in health satisfaction among middle-aged and elderly Chinese adults. The predictive models developed in this study can help policymakers identify high-risk populations, enabling more targeted interventions to improve health satisfaction levels across different regions.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"108 ","pages":"Pages 16-25"},"PeriodicalIF":3.3,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144203882","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Assessing the association between sport participation and suicide ideation and behaviors among middle and high school students in the U.S. between 2007 and 2023","authors":"Philip T. Veliz ,&nbsp;John Jardine ,&nbsp;Ashley Cureton ,&nbsp;Massy Mutumba","doi":"10.1016/j.annepidem.2025.05.017","DOIUrl":"10.1016/j.annepidem.2025.05.017","url":null,"abstract":"<div><h3>Purpose</h3><div>To assess the association between sports participation and suicide ideation/behaviors among both middle and high school students across several national cohorts of adolescents in the U.S. (i.e., 2007–2013, 2015–2019, 2021 and 2023).</div></div><div><h3>Methods</h3><div>Data from the 2007–2023 biennial state <em>Youth Risk Behavior Survey (YRBS)</em> for both middle (n = 326,085) and high schools (n = 508,737) were used to analyze the association between suicide ideation/suicide behaviors and past-year sports participation.</div></div><div><h3>Results</h3><div>Participation in sports dropped for both middle and high school students between 2007 and 2021, while the percent of both middle and high school students who seriously considered attempting suicide, developed a suicide plan and attempted suicide increased between 2007 and 2023. Multivariable models found that the rates of seriously considering attempting suicide (range: aRR=.71 to.81, p &lt; .001), developing a suicide plan (range: aRR=.72 to.80, p &lt; .001) and attempting suicide (range: aRR=.75 to.87, p &lt; .001) were lower for middle and high school students who participated in either one sport or two or more sports when compared to their peers who do not participate in these activities.</div></div><div><h3>Conclusion</h3><div>Participation in sports lowered the risk of suicide ideation/behaviors for both middle and high school students in the U.S. The findings suggest that engaging in sports, particularly multiple sports, serves as an intervention strategy for reducing suicide risks in this population.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"108 ","pages":"Pages 8-15"},"PeriodicalIF":3.3,"publicationDate":"2025-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144210145","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"County-level socioeconomic, race-ethnicity, household, and infrastructure vulnerability associations with bone cancer treatment and prognosis in the US","authors":"David J. Fei-Zhang ,&nbsp;Erik B. Gerlach ,&nbsp;Shyam Chandrasekar ,&nbsp;Mark A. Plantz ,&nbsp;Nicholas C. Arpey ,&nbsp;Peter R. Swiatek ,&nbsp;David R. Christian ,&nbsp;Bejan A. Alvandi ,&nbsp;Corey A. Jones ,&nbsp;Ruohui Chen ,&nbsp;Seth M. Pollack ,&nbsp;Terrance D. Peabody","doi":"10.1016/j.annepidem.2025.05.007","DOIUrl":"10.1016/j.annepidem.2025.05.007","url":null,"abstract":"<div><h3>Importance</h3><div>The impact of social determinants of health (SDoH) on primary bone cancer (PBC) treatment and clinical outcomes has seen little inquiry due to restrictive study designs of narrow histology scopes, limited SDoH-factors assessed, and lacking geographic-span of study populations.</div></div><div><h3>Objectives</h3><div>To assess how varied SDoH impact the surveillance, survival, and treatment of adults with PBCs in the US by quantifying the associations of the summative and individual influences of SDoH-vulnerability with outcome disparities.</div></div><div><h3>Design, setting, participants</h3><div>This retrospective cohort study on adult PBC patients between 1975 and 2017 from the Surveillance-Epidemiology-End Results Program (SEER) database. Data were analyzed from November 2022 to January 2024.</div></div><div><h3>Exposures</h3><div>Overall social vulnerability and its subcomponents comprised of 15 SDoH-factors in 4 themes of socioeconomic status (poverty, unemployment, income, high school education), minority-language (minoritized race/ethnicity, English proficiency), household composition (disability status, family members &lt;18 or 65 + years, single-parent status) and housing-transportation (multiunit-mobile dwelling, overcrowding, vehicle access, group quarters) status and their total composite.</div></div><div><h3>Main outcomes and measures</h3><div>Regression trends were used for SDoH-vulnerability associations with follow-up/surveillance period, survival period, and treatment receipt across PBCs.</div></div><div><h3>Results</h3><div>For 13,664 PBC patients, increasing total social vulnerability was associated with significant decreases in surveillance period for chondrosarcomas, chordomas, Ewing sarcoma, malignant giant cell tumors, NOS Malignant Neoplasms, NOS or other Sarcomas, and osteosarcomas, ranging from 20.62 % to 51.07 % in relative decreases. Survival period decreases were observed with chondrosarcomas (44.08 %) and chordomas (31.30 %) when overall social vulnerability increased. Decreased receipt of recommended surgery (lowest, NOS or other sarcomas: OR 0.86; 95 %CI 0.76–0.97) and radiation therapy (lowest, Ewing Sarcoma: OR 0.89; 95 %CI 0.81–0.98) for several histology types was also observed with increasing total social vulnerability. Socioeconomic status, followed closely by housing-transportation and minority-language status vulnerabilities largely comprised these overall detrimental trends.</div></div><div><h3>Conclusions</h3><div>Overall social vulnerability showed significant decreases in care receipt and worsened prognosis for PBC patients while quantifiably characterizing which factors disproportionately associated with detrimental trends, informing providers which SDoH should be addressed to more effectively combat PBC disparities.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"108 ","pages":"Pages 26-33"},"PeriodicalIF":3.3,"publicationDate":"2025-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144210146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Transfer learning for COVID-19 predictive modeling: A multicenter study of 12 hospitals","authors":"Carine Savalli ,&nbsp;André Henrique Alves Carneiro ,&nbsp;Fabiano Barcellos Filho ,&nbsp;Murilo Afonso Robiati Bigoto ,&nbsp;Roberta Moreira Wichmann ,&nbsp;Alexandre Dias Porto Chiavegatto Filho ,&nbsp;on behalf of IACOV-BR Network","doi":"10.1016/j.annepidem.2025.05.013","DOIUrl":"10.1016/j.annepidem.2025.05.013","url":null,"abstract":"<div><h3>Purpose</h3><div>To use transfer learning to predict Intensive Care Unit admission (ICU) for COVID-19 patients in a multicenter setting.</div></div><div><h3>Methods</h3><div>We trained XGBoost algorithms using demographic and laboratory predictors recorded during hospital admissions across 12 hospitals. Our goal was to identify the best-performing hospital and evaluate its generalization capacity through external validation in the other 11 hospitals. Additionally, this best-performing hospital served as the source domain for transfer learning and was fine-tuned by incorporating new decision trees trained with local data from each of the remaining hospitals.</div></div><div><h3>Results</h3><div>We observed considerable variation in ICU admission outcomes across the 12 hospitals, with local predictive performance (AUC – area under the receiver operating characteristic curve) ranging from 0.6239 to 0.9410. The best-performing model's generalization capacity, evaluated through external validation, resulted in an AUC below 0.7 in 6 hospitals. By applying transfer learning and fine-tuning this best-performing model, we found that its performance stabilized after incorporating 20 new trees trained with data from the target hospitals. This adjustment improved AUC in 9 out of the 11 hospitals, compared to the results from the external validation.</div></div><div><h3>Conclusions</h3><div>The study showed that transfer learning effectively leverages knowledge from the best-performing hospital to initialize and accelerate model training in other hospitals with lower performance or limited sample sizes, thereby enhancing overall adaptability and efficiency.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"108 ","pages":"Pages 1-7"},"PeriodicalIF":3.3,"publicationDate":"2025-05-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144195189","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Benefits of reaching the national 2025 goal for reduction in fatal drug overdoses, AEP-D-25-00208 R2","authors":"David R. Holtgrave PhD ,&nbsp;James V. McDonald MD, MPH","doi":"10.1016/j.annepidem.2025.05.015","DOIUrl":"10.1016/j.annepidem.2025.05.015","url":null,"abstract":"<div><h3>Purpose</h3><div>Centers for Disease Control and Prevention (CDC) provisional data on drug fatalities for year ending December 2024 indicates that the 2025 national goal of reducing drug fatalities to 81,000 or fewer was met one year early. We estimate the 2024 public health impacts of meeting that goal and analyze the impacts of three scenarios for 2025: continued downward trajectory of drug fatalities, stagnation, or regression of progress.</div></div><div><h3>Methods</h3><div>CDC 2024 data was compared to 2023 data to estimate lives saved, years of life saved, and valuation of statistical lives saved. Three scenarios for year ending December 2025 were compared to 2024 data to estimate the impact on the same metrics.</div></div><div><h3>Results</h3><div>Drug fatality progress in 2024 relative to 2023 resulted in 29,646 lives saved, 928,916 years of life saved, and valuation of statistical lives saved ranging from $180 billion to $584 billion. If similar percentage declines in drug fatalities occur in 2025, another 21,659 lives would be saved. However, if drug fatalities flatten or rise in 2025, opportunities to save lives be lost.</div></div><div><h3>Conclusion</h3><div>Meeting the national goal one year early resulted in substantial public health benefits in 2024. Now, programmatic and policy efforts must prevent stagnation or backsliding.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"107 ","pages":"Pages 51-53"},"PeriodicalIF":3.3,"publicationDate":"2025-05-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144178464","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}