Hasil untuk "Public aspects of medicine"

Felix Stoehr, María Díaz Trigo, Evanthia Hatziminaoglou et al.

The success of an astronomical facility is measured by its scientific impact. A principal metric for this impact is the ensemble of peer-reviewed publications based on the observational data obtained by the facility. We present a comprehensive study of the statistics of the 4,190 refereed publications of the Atacama Large Millimeter/Submillimeter Array (ALMA) in the period from 2012 to 2024. The publications have received 169,985 citations and are based on 2,670 ALMA projects totalling 19,265 hours of 12-m-array-equivalent observing time. Our study analyses publication statistics related to various aspects, e.g. science categories, geographical distribution, archival research, time to publication, publication fraction, and citations. We also look into the community and compare ALMA with other facilities. We find that ALMA is a high-impact observatory with an average of 41 citations per publication, ~70% of observed projects published, ~40% of publications making use of archival data in 2024, more than 9,400 unique authors, and a publication evolution following that of HST and VLT. Currently, the impact factor for ALMA publications is larger than that of all other major astronomical facilities. ALMA also plays a pivotal role in very long baseline interferometry (VLBI), substantially contributing to landmark achievements such as capturing the first image of a black hole shadow.

Jaemu Lee, Kyung Hye Park, Kyung Hye Park et al.

IntroductionHappiness differs according to population groups and cultures. For medical students, more studies have focused on negative emotions than on happiness. This study explored the overall perceptions and standards of medical students to analyze the concept of happiness from various perspectives in the Korean context.MethodsA concept mapping analysis comprising five stages was conducted with medical students at Yonsei University’s Wonju College of Medicine in South Korea. Focus questions were generated in Phase 1, and 23 students participated in individual brainstorming in Phase 2. Fifty statements were confirmed in Phase 3. Sixteen students assigned an importance score to each statement and participated in the individual sorting of statements and naming of categories in Phase 4. Finally, the concept maps were interpreted using multidimensional scaling and hierarchical cluster analysis.ResultsThe medical students’ perception of happiness was divided into two dimensions, “Study–Life” on the X-axis and “Self–Relationship” on the Y-axis, and was expressed in three categories and five sub-categories. The subcategories of “Self-management” and “Quality of life” were grouped under “Personal development,” “Social support” was named as a single category, and “Guaranteed future” and “Academic achievement” were grouped as “Professional fulfillment.” The most important sub-category for medical students was “Social support.” Among the statements generated in these categories, the most important was “When I have a healthy body and stamina,” which belonged to “Quality of life.”DiscussionThis study showed that to enhance the happiness of medical students, a system that supports their social relationships, careers, learning, and individual efforts is required. The results of this study can provide information for the development of student support programs that allow medical educators and institutions to promote medical students’ happiness.

Yuan Zhang, Ya Zhang, Jing Wang et al.

Abstract Objective By conducting TORCH screening and risk assessment analysis on childbearing-age women in different regions of China, the aim is to provide reference for reducing adverse pregnancy outcomes and improving the health status of childbearing-age women. Methods Between February and May 2021, in the eastern, central, and western regions of China (Beijing, Henan, Gansu), a total of 1,942 couples aged 18 to 49, from both urban and rural areas, were included in this cross-sectional study. TORCH screening was conducted on all these women of childbearing-age, and risk assessment was performed based on the examination results. Result In this study, toxoplasmosis, rubella, CMV, HSV, IgM positive rate were 0.2%, 0.4%, 0.3%, 0.3%, respectively, and CT, TP, HBsAg, HCV, HIV, positive rate were 0.1%, 0.2%, 2.1%, 0.3%, 0.0%, respectively. The total TORCH screening identified 63.4% of women as having potential risks, compared to 15.5% of routine ToRCH screening. The distribution of the risk population shows significant differences among provinces, ethnicities, education levels, and age groups (p ≤ 0.001). Trend chi-square tests revealed that as the level of education increased, the proportion of the risk population decreased. Conclusions The TORCH screening utilized in this study demonstrates advantages over ToRCH, as it can identify more women of childbearing age with potential risks before pregnancy, allowing for early interventions. Simultaneously, these findings underscore the necessity for targeted health education, especially for young women in economically underdeveloped areas and those with relatively lower education levels.

Caroline W. Grant, Jerry Li, Swan Lin et al.

ABSTRACT Patient outcomes in advanced renal cell carcinoma (RCC) remain poor, with five‐year survival rates ranging from ~10% to 30%. Early projections of therapeutic outcomes could optimize precision medicine and accelerate drug development. While machine learning (ML) models integrating tumor growth inhibition (TGI) metrics have improved survival predictions over traditional models, their application in RCC remains unexplored. Herein, we used TGI metrics and baseline data to evaluate parametric (PM) and semi‐parametric (SPM) survival models alongside ML approaches for predicting progression‐free (PFS) and overall survival (OS) in 1839 RCC patients from four trials (evaluating sunitinib, axitinib, sorafenib, interferon‐alpha, and avelumab + axitinib). Data were split into training (70%) and testing (30%), and feature selection was used to determine parsimonious and robust models. Bootstrap resampling (n = 100) was employed for models' validation, and performance was assessed using C‐index and Integrated Brier Score. In brief, training data results demonstrated that tree‐based ML models (random survival forest (RSF) and XGBoost) outperformed PM and SPM models in predicting PFS (C‐index: 0.783–0.785 vs. 0.725–0.738 for PM and SPM; p < 0.05) and OS (C‐index: 0.77–0.867 vs. 0.750–0.758 for PM and SPM; p < 0.05), with RSF achieving better prediction of PFS and OS using only 3–5 covariates, compared to 9–35 with other tested methods. Tree‐based methods were also superior in the testing data. SHapley Additive exPlanations revealed nonlinear relationships among top predictors, including TGI metrics, underscoring the ability of tree‐based methods to capture complex prognostic interactions. Further validation is required to confirm models' generalizability to additional therapies and patients with differing tumor severity.

Hanne Marie Rostad, Lucille Xiang, Elizabeth M. White

Abstract Background Health disparities exist across healthcare settings, including nursing homes, contributing to preventable differences in care quality. Health disparities are a global issue, yet most studies on nursing home health disparities have been conducted in the United States. In this scoping review, our objective was to synthesize methods used in U.S. nursing home disparities research to gain insights to inform similar research in other countries. Specifically, we summarized different approaches for conceptualizing and measuring health disparities, available data sources, study designs, and analytic strategies. Methods We employed two parallel search strategies across five databases, targeting specific aspects of health disparities and broader concepts. Study selection was conducted independently by two reviewers. Using a numerical and analytic approach, we categorized and summarized the results. Results The search yielded 6,817 records, with 82 unique studies meeting the inclusion criteria. All studies used quantitative methods, with only two incorporating mixed methods. Most were observational cross-sectional studies (n = 60), while 21 were longitudinal studies, and 1 was a randomized controlled trial. Most studies used administrative data (n = 62). The majority (n = 65) measured differences in health outcomes across nursing homes. A significant number of studies (n = 71) focused on racial and/or ethnic health disparities, and a few studied clinical conditions (n = 7), rural–urban location of the nursing home (n = 4), socioeconomic factors (n = 4), age (n = 1), and sex (n = 1) as characteristics to measure disparity. Outcomes were grouped into five domains: 1) Quality of care measures (n = 54), 2) Infection and infection prevention (n = 22), 3) Transitions and acute care utilization (n = 19), 4) Behavioral and mental health (n = 18), and 5) Palliative care, end-of-life and death (n = 13). Across the five domains, the most prevalent outcome category studied was ‘Hospitalization and emergency room use’ (n = 15). Conclusion This review highlights key issues for future research on health disparities in nursing homes, including the need to: 1) clarify concepts of health disparities and health equity; 2) move beyond mere descriptions of disparities to identify underlying factors contributing to those disparities; 3) broaden examination of disparities beyond a single axis such as race or sex; 4) integrate more qualitative data to capture nuances that cannot be measured from quantitative data; and 5) specify whether within or across nursing home differences are studied.

Teddy Mekonnen

This paper analyzes the strategic interactions between a profit-maximizing monopolist and a free, capacity-constrained public option. By restricting its own supply, the monopolist intentionally congests the public option and induces rationing, which increases consumers' willingness to pay for guaranteed access. Counterintuitively, expanding the public option's capacity may raise the monopoly price and lower consumer welfare. However, I derive conditions under which all buyer types benefit from a capacity expansion, and extend these results to a setting where an oligopoly competes with a public option. These findings have implications for mixed public-private markets, such as housing, education, and healthcare.

Mahmoud Alwakeel, Aditya Nagori, Vijay Krishnamoorthy et al.

Objectives: To evaluate the current limitations of large language models (LLMs) in medical question answering, focusing on the quality of datasets used for their evaluation. Materials and Methods: Widely-used benchmark datasets, including MedQA, MedMCQA, PubMedQA, and MMLU, were reviewed for their rigor, transparency, and relevance to clinical scenarios. Alternatives, such as challenge questions in medical journals, were also analyzed to identify their potential as unbiased evaluation tools. Results: Most existing datasets lack clinical realism, transparency, and robust validation processes. Publicly available challenge questions offer some benefits but are limited by their small size, narrow scope, and exposure to LLM training. These gaps highlight the need for secure, comprehensive, and representative datasets. Conclusion: A standardized framework is critical for evaluating LLMs in medicine. Collaborative efforts among institutions and policymakers are needed to ensure datasets and methodologies are rigorous, unbiased, and reflective of clinical complexities.

Shlomi Hod, Lucas Rosenblatt, Julia Stoyanovich

Differentially private (DP) machine learning often relies on the availability of public data for tasks like privacy-utility trade-off estimation, hyperparameter tuning, and pretraining. While public data assumptions may be reasonable in text and image domains, they are less likely to hold for tabular data due to tabular data heterogeneity across domains. We propose leveraging powerful priors to address this limitation; specifically, we synthesize realistic tabular data directly from schema-level specifications - such as variable names, types, and permissible ranges - without ever accessing sensitive records. To that end, this work introduces the notion of "surrogate" public data - datasets generated independently of sensitive data, which consume no privacy loss budget and are constructed solely from publicly available schema or metadata. Surrogate public data are intended to encode plausible statistical assumptions (informed by publicly available information) into a dataset with many downstream uses in private mechanisms. We automate the process of generating surrogate public data with large language models (LLMs); in particular, we propose two methods: direct record generation as CSV files, and automated structural causal model (SCM) construction for sampling records. Through extensive experiments, we demonstrate that surrogate public tabular data can effectively replace traditional public data when pretraining differentially private tabular classifiers. To a lesser extent, surrogate public data are also useful for hyperparameter tuning of DP synthetic data generators, and for estimating the privacy-utility tradeoff.

Marc Brinner, Sina Zarrieß

We propose SemCSE-Multi, a novel unsupervised framework for generating multifaceted embeddings of scientific abstracts, evaluated in the domains of invasion biology and medicine. These embeddings capture distinct, individually specifiable aspects in isolation, thus enabling fine-grained and controllable similarity assessments as well as adaptive, user-driven visualizations of scientific domains. Our approach relies on an unsupervised procedure that produces aspect-specific summarizing sentences and trains embedding models to map semantically related summaries to nearby positions in the embedding space. We then distill these aspect-specific embedding capabilities into a unified embedding model that directly predicts multiple aspect embeddings from a scientific abstract in a single, efficient forward pass. In addition, we introduce an embedding decoding pipeline that decodes embeddings back into natural language descriptions of their associated aspects. Notably, we show that this decoding remains effective even for unoccupied regions in low-dimensional visualizations, thus offering vastly improved interpretability in user-centric settings.

Gina D'Angelo, Xiaowen Tian, Chuyu Deng et al.

Precision medicine is an evolving area in the medical field and rely on biomarkers to make patient enrichment decisions, thereby providing drug development direction. A traditional statistical approach is to find the cut-off that leads to the minimum p-value of the interaction between the biomarker dichotomized at that cut-off and treatment. Such an approach does not incorporate clinical significance and the biomarker is not evaluated on a continuous scale. We are proposing to evaluate the biomarker in a continuous manner from a predicted risk standpoint, based on the model that includes the interaction between the biomarker and treatment. The predicted risk can be graphically displayed to explain the relationship between the outcome and biomarker, whereby suggesting a cut-off for biomarker positive/negative groups. We adapt the TreatmentSelection approach and extend it to account for covariates via G-computation. Other features include biomarker comparisons using net gain summary measures and calibration to assess the model fit. The PRIME (Predictive biomarker graphical approach) approach is flexible in the type of outcome and covariates considered. A R package is available and examples will be demonstrated.

Mohammad Amaan Sayeed, Mohammed Talha Alam, Raza Imam et al.

Centuries-old Islamic medical texts like Avicenna's Canon of Medicine and the Prophetic Tibb-e-Nabawi encode a wealth of preventive care, nutrition, and holistic therapies, yet remain inaccessible to many and underutilized in modern AI systems. Existing language-model benchmarks focus narrowly on factual recall or user preference, leaving a gap in validating culturally grounded medical guidance at scale. We propose a unified evaluation pipeline, Tibbe-AG, that aligns 30 carefully curated Prophetic-medicine questions with human-verified remedies and compares three LLMs (LLaMA-3, Mistral-7B, Qwen2-7B) under three configurations: direct generation, retrieval-augmented generation, and a scientific self-critique filter. Each answer is then assessed by a secondary LLM serving as an agentic judge, yielding a single 3C3H quality score. Retrieval improves factual accuracy by 13%, while the agentic prompt adds another 10% improvement through deeper mechanistic insight and safety considerations. Our results demonstrate that blending classical Islamic texts with retrieval and self-evaluation enables reliable, culturally sensitive medical question-answering.

Tomoko Nishimura, Tomoko Nishimura, Tomoko Nishimura et al.

ObjectiveThe implementation of school-based mental health screening offers promise for early detection of mental health issues in children; however, various barriers hinder its widespread adoption. This study aimed to investigate the predictive value of digital data obtained from an established daily health observation scheme in Japanese schools to identify later mental health issues in children.MethodsData for the analysis were obtained from 2,433 students enrolled in five public schools. The data acquisition period spanned 76 school days, from September 1, 2022, to December 23, 2022, and student absences were recorded during this period. Depressive and anxiety symptoms were assessed in January 2023. The students’ daily physical and emotional health status was recorded as “daily health issue” scores and group-based trajectory modeling was employed to classify the long-term trends in these scores. Additionally, rolling z-scores were utilized to capture variability in daily health issue scores, with z-scores above +1 considered unusual responses.ResultsAfter 4 months of daily health observations, students’ response trends were classified into five trajectory groups. The group experiencing the highest number of daily health issues (Group 5; 5.4% of the sample) exhibited more subsequent depressive and anxiety symptoms compared to the group with fewer issues (Group 1; 47.5%) (incident rate ratio [IRR] = 5.17; 95% confidence interval [CI]: 3.82, 6.99). Group 5 also demonstrated significantly more days of absence than Group 1 (IRR = 2.14, 95% CI: 1.19, 3.85). The average daily health issue scores for the entire period were associated with both depressive/anxiety symptoms and the number of days absent from school (IRR = 1.59, 95% CI: 1.45, 1.73; IRR = 1.18, 95% CI: 1.04, 1.35, respectively). Furthermore, a higher number of unusual responses during the entire period was also associated with more depressive/anxiety symptoms (IRR = 1.10, 95% CI: 1.07, 1.12).ConclusionThe current study is the first to demonstrate the predictive capability of a traditional daily health observation scheme to identify mental health issues in children. This study highlights the scheme’s potential to screen and safeguard children’s mental health, emphasizing the importance of digitalization and collaboration with various stakeholders.

Judith Owokuhaisa, Eleanor Turyakira, Frank Ssedyabane et al.

Abstract Background Cervical cancer continues to threaten women’s health, especially in low-resource settings. Regular follow-up after screening and treatment is an effective strategy for monitoring treatment outcomes. Consequently, understanding the factors contributing to patient non-attendance of scheduled follow-up visits is vital to providing high-quality care, reducing morbidity and mortality, and unnecessary healthcare costs in low-resource settings. Methods A descriptive qualitative study was done among healthcare providers and patients who attended the cervical cancer screening clinic at Mbarara Regional Referral Hospital in southwestern Uganda. In-depth interviews were conducted using a semi-structured interview guide. Interviews were audio-recorded, transcribed verbatim, and thematically analysed in line with the social-ecological model to identify barriers and facilitators. Results We conducted 23 in-depth interviews with 5 healthcare providers and 18 patients. Health system barriers included long waiting time at the facility, long turnaround time for laboratory results, congestion and lack of privacy affecting counselling, and healthcare provider training gaps. The most important interpersonal barrier among married women was lacking support from male partners. Individual-level barriers were lack of money for transport, fear of painful procedures, emotional distress, and illiteracy. Inadequate and inaccurate information was a cross-cutting barrier across the individual, interpersonal, and community levels of the socio-ecological model. The facilitators were social support, positive self-perception, and patient counselling. Conclusions Our study revealed barriers to retention in care after cervical cancer screening, including lack of partner support, financial and educational constraints, and inadequate information. It also found facilitators that included social support, positive self-perception, and effective counselling.

Bongsu Kang, Jundong Kim, Tae-Rim Yun et al.

We propose a natural language prompt-based retrieval augmented generation (Prompt-RAG), a novel approach to enhance the performance of generative large language models (LLMs) in niche domains. Conventional RAG methods mostly require vector embeddings, yet the suitability of generic LLM-based embedding representations for specialized domains remains uncertain. To explore and exemplify this point, we compared vector embeddings from Korean Medicine (KM) and Conventional Medicine (CM) documents, finding that KM document embeddings correlated more with token overlaps and less with human-assessed document relatedness, in contrast to CM embeddings. Prompt-RAG, distinct from conventional RAG models, operates without the need for embedding vectors. Its performance was assessed through a Question-Answering (QA) chatbot application, where responses were evaluated for relevance, readability, and informativeness. The results showed that Prompt-RAG outperformed existing models, including ChatGPT and conventional vector embedding-based RAGs, in terms of relevance and informativeness. Despite challenges like content structuring and response latency, the advancements in LLMs are expected to encourage the use of Prompt-RAG, making it a promising tool for other domains in need of RAG methods.

Candice P. Chu

ChatGPT, the most accessible generative artificial intelligence (AI) tool, offers considerable potential for veterinary medicine, yet a dedicated review of its specific applications is lacking. This review concisely synthesizes the latest research and practical applications of ChatGPT within the clinical, educational, and research domains of veterinary medicine. It intends to provide specific guidance and actionable examples of how generative AI can be directly utilized by veterinary professionals without a programming background. For practitioners, ChatGPT can extract patient data, generate progress notes, and potentially assist in diagnosing complex cases. Veterinary educators can create custom GPTs for student support, while students can utilize ChatGPT for exam preparation. ChatGPT can aid in academic writing tasks in research, but veterinary publishers have set specific requirements for authors to follow. Despite its transformative potential, careful use is essential to avoid pitfalls like hallucination. This review addresses ethical considerations, provides learning resources, and offers tangible examples to guide responsible implementation. Carefully selected, up-to-date links to platforms that host large language models are provided for advanced readers with programming capability. A table of key takeaways was provided to summarize this review. By highlighting potential benefits and limitations, this review equips veterinarians, educators, and researchers to harness the power of ChatGPT effectively.

Christoph Hauert, Alex McAvoy

When individuals interact in groups, the evolution of cooperation is traditionally modeled using the framework of public goods games. These models often assume that the return of the public good depends linearly on the fraction of contributors. In contrast, in real life public goods interactions, the return can depend on the size of the investor pool as well. Here, we consider a model in which the multiplication factor (marginal per capita return) for the public good depends linearly on how many contribute, which results in a nonlinear model of public goods. This simple model breaks the curse of dominant defection found in linear public goods interactions and gives rise to richer dynamical outcomes in evolutionary settings. We provide an in-depth analysis of the more varied decisions by the classical rational player in nonlinear public goods interactions as well as a mechanistic, microscopic derivation of the evolutionary outcomes for the stochastic dynamics in finite populations and in the deterministic limit of infinite populations. This kind of nonlinearity provides a natural way to model public goods with diminishing returns as well as economies of scale.

Renee Reynolds, Melanie Kingsland, Justine Daly et al.

Abstract Background Exclusive breastfeeding to six months of age is a major global public health priority. Several characteristics are known to be associated with early cessation of breastfeeding, however, limited evidence exists regarding whether women’s reported reasons for cessation are associated with maternal, pregnancy and infant characteristics. The aims of this study were to: i) describe women’s reported intention to breastfeed and their subsequent breastfeeding practices; ii) describe women’s reported reasons for breastfeeding cessation prior to the infant being five months of age; and iii) examine associations between these factors and maternal, pregnancy and infant characteristics. Methods Telephone and online surveys were conducted between October 2019 and April 2020 with 536 women who had given birth in the previous eight to 21 weeks at four public maternity services in Australia. Results The majority of women intended to (94%), and did, initiate (95%) breastfeeding. At the time the survey was conducted, 57% of women were exclusively breastfeeding. Women who: had less than University level education, had a pre-pregnancy BMI in the overweight or obese category, and who smoked tobacco at the time of the survey had lower odds of exclusively breastfeeding. The most common self-reported reasons for breastfeeding cessation were breastfeeding challenges (47%) and low milk supply (40%). Women aged 26–35 years and 36 + years had greater odds of reporting breastfeeding cessation due to low milk supply (OR = 2.92, 95% CI: 1.11, 7.66; OR = 5.57, 95% CI: 1.70, 18.29) compared to women aged 18–25 years. While women who had completed a TAFE certificate or diploma had lower odds of reporting this as a reason for breastfeeding cessation (OR = 0.28; 95% CI: 0.11, 0.73) compared to women who had University level education. There were no other significant associations found between characteristics and reasons for ceasing breastfeeding. Conclusions The most common reasons for breastfeeding cessation may be modifiable through the provision of breastfeeding support in the early postpartum period, with such support being tailored to women’s age and level of education. Such support should aim to increase women's self-efficacy in breastfeeding, and be provided from the antenatal period and throughout the first six months postpartum.

Cyril Zakka, Akash Chaurasia, Rohan Shad et al.

Large-language models have recently demonstrated impressive zero-shot capabilities in a variety of natural language tasks such as summarization, dialogue generation, and question-answering. Despite many promising applications in clinical medicine, adoption of these models in real-world settings has been largely limited by their tendency to generate incorrect and sometimes even toxic statements. In this study, we develop Almanac, a large language model framework augmented with retrieval capabilities for medical guideline and treatment recommendations. Performance on a novel dataset of clinical scenarios (n = 130) evaluated by a panel of 5 board-certified and resident physicians demonstrates significant increases in factuality (mean of 18% at p-value < 0.05) across all specialties, with improvements in completeness and safety. Our results demonstrate the potential for large language models to be effective tools in the clinical decision-making process, while also emphasizing the importance of careful testing and deployment to mitigate their shortcomings.

Paul D. W. Kirk, Filippo Pagani, Sylvia Richardson

Clustering is commonly performed as an initial analysis step for uncovering structure in 'omics datasets, e.g. to discover molecular subtypes of disease. The high-throughput, high-dimensional nature of these datasets means that they provide information on a diverse array of different biomolecular processes and pathways. Different groups of variables (e.g. genes or proteins) will be implicated in different biomolecular processes, and hence undertaking analyses that are limited to identifying just a single clustering partition of the whole dataset is therefore liable to conflate the multiple clustering structures that may arise from these distinct processes. To address this, we propose a multi-view Bayesian mixture model that identifies groups of variables (``views"), each of which defines a distinct clustering structure. We consider applications in stratified medicine, for which our principal goal is to identify clusters of patients that define distinct, clinically actionable disease subtypes. We adopt the semi-supervised, outcome-guided mixture modelling approach of Bayesian profile regression that makes use of a response variable in order to guide inference toward the clusterings that are most relevant in a stratified medicine context. We present the model, together with illustrative simulation examples, and examples from pan-cancer proteomics. We demonstrate how the approach can be used to perform integrative clustering, and consider an example in which different 'omics datasets are integrated in the context of breast cancer subtyping.

Sonia Alejandra Pou, Maria Del Pilar Diaz, Leandro Mariano Gonzalez

ABSTRACT: Objective: To assess the impact of the COVID-19 pandemic on mortality in Argentina, considering temporal trends in life expectancy at birth and premature mortality rate during 2010-2020. Methods: Based on demographic projections, this ecological time-series study compares a “normal” versus a “COVID-19” mortality scenario for 2020 over a set of 11 Argentine provinces. Annual life expectancy at birth and age-standardized rates of premature mortality were estimated from 2010 to 2020. Joinpoint regression and multilevel models were used. Results: A potential reduction in life expectancy at birth (a gap between scenarios >1 year) was observed. A significant (negative) point of inflection in temporal trends was identified for the country and most of the provinces, under the COVID-19 mortality scenario. However, our findings reveal disparities between provinces in the estimated life expectancy reduction toward 2020 (values range from -0.63 to -1.85 year in females and up to -2.55 years in males). While men showed more accentuated declines in life expectancy at birth in 2020 (a national gap between scenarios of -1.47 year in men vs. -1.35 year in women), women experienced more unfavorable temporal trends of premature mortality. In the absence of COVID-19, an improvement in both indicators was estimated toward 2020 in both sexes, while a return to levels reported in the past was observed under the COVID-19 scenario. Conclusion: The COVID-19 pandemic might seriously affect the trends of mortality and exacerbate health disadvantages in Argentina. A temporal and contextual perspective of health inequities merits special attention in the COVID-19 research.