Opin vísindi

 

Nýlega bætt við

Verk
Mental, physical, and cognitive manifestations of post COVID-19 condition: Evidence from population-based cohorts and national registers
(University of Iceland, School of Health Sciences, Faculty of Medicine, 2026-10) Magnúsdóttir, Ingibjörg; Thor Aspelund; Faculty of Medicine (UI); Læknadeild (HÍ); School of Health Sciences (UI); Heilbrigðisvísindasvið (HÍ)
The global spread of Coronavirus 2019 (COVID-19) has raised concerns about the long-term health consequences among those infected, particularly beyond the acute phase of illness. While a growing number of studies have examined post COVID-19 condition (PCC), much of the existing literature has focused on hospitalised patients or short follow-up periods, often without appropriate comparison groups. Consequently, there remains limited understanding of long-term mental, physical, and cognitive outcomes in the general population and how these outcomes vary according to acute illness severity. The overarching aim of this thesis was to address these gaps by examining long-term health outcomes following COVID-19 across multiple domains using four complementary studies. In Study I, we explored the prevalence of mental health symptoms among individuals diagnosed with COVID-19 compared with those not diagnosed, using multinational cohort data from six countries (Denmark, Estonia, Iceland, Norway, Scotland and Sweden; N=247,249). Individuals with COVID-19 had a higher prevalence of depressive symptoms and poor sleep quality, while no overall increase was observed for anxiety or COVID-19-related distress. A clear gradient by illness severity was identified, with individuals bedridden for seven days or longer showing persistently elevated levels of depressive symptoms, anxiety, and sleep problems up to 16 months after diagnosis. In contrast, individuals never bedridden due to COVID-19 showed similar or lower symptom prevalence compared with those without COVID-19. In Study II, we examined the association between COVID-19 diagnosis and the risk of incident psychiatric disorders and psychotropic medication dispensing, using national register data from Iceland (N=112,159) from February 2020 to February 2022. Individuals diagnosed with COVID-19 had an increased risk of psychiatric disorders and psychotropic medication use compared with those not diagnosed. The associations were strongest for psychotic disorders and substance misuse. Higher risks were observed among individuals that were older, unvaccinated, and with more severe disease. Estimates decreased over time during the study period. In Study III, we investigated the prevalence of long-term physical symptoms following COVID-19 in four Nordic cohorts (N=64,880). Individuals diagnosed with COVID-19 had a higher prevalence of severe physical symptoms compared with those not diagnosed. A strong dose–response relationship was observed, with individuals bedridden for seven days or longer experiencing the highest and most persistent symptom burden up to 27 months after diagnosis. The most commonly reported symptoms included shortness of breath, dizziness, headaches, and low energy or fatigue. In Study IV, we assessed impaired cognitive function among individuals diagnosed with COVID-19 using data from five cohorts (N=153,841). No overall association was observed between COVID-19 diagnosis and impaired cognitive function. However, illness severity was strongly associated with cognitive outcomes, with individuals bedridden for seven days or longer showing a substantially higher prevalence of impaired cognitive function, persisting for up to 32 months after diagnosis. Individuals who were never bedridden showed little to no evidence of increased prevalence. In conclusion, these studies identify long-term adverse outcomes across mental, physical, and cognitive domains, with acute illness severity emerging as a key indicator across all outcomes. Individuals with more severe acute illness experienced a substantial and persistent symptom burden, whereas those in the lowest measured severity categories generally showed little evidence of increased morbidity. However, the results should be interpreted with caution because illness severity was based on self-reported days bedridden in the cohort studies and healthcare utilisation in the register-based study. These findings contribute to a better understanding of PCC and highlight the importance of considering illness severity in future research and clinical practice.
Verk
Trial Protocol : A Randomised, Controlled Cross-Over Trial Evaluating the Digital Sleep Revolution Diagnostic-Therapeutic Pathway for Obstructive Sleep Apnoea
(2026-09-22) the Sleep Revolution Collaborators; Department of Engineering; Department of Computer Science; Department of Psychology
Obstructive sleep apnoea (OSA) is common, but current diagnostic pathways rely on single-night assessment and the apnoea–hypopnoea index. Both are limited for assessment of variability and linking only weakly to symptoms, comorbidity and treatment response. The current protocol addresses this gap by aiming to validate a diagnostic pathway integrating a digital platform, mobile application, multiple-night sleep testing, semiautomated analysis and a novel sleep health questionnaire, labelled as ‘Sleep Revolution’. Study objectives are to compare staff time required for diagnosis between standard care and Sleep Revolution pathway, and to examine associations between multi-night sleep metrics and Positive Airway Pressure (PAP) treatment outcomes, diagnostic concordance, PAP prescription, patient satisfaction, adherence and feasibility. This multicentre, two-arm cross-over study is conducted at 23 sleep centres in Europe within the European Sleep Apnoea Database (ESADA) network. Adults referred for suspected OSA complete both standard care and the Sleep Revolution pathway in randomised order. Standard care includes single-night polygraphy or polysomnography according to local routines. The Sleep Revolution pathway comprises three nights of self-applied polysomnography (NOX-SAS), digital symptom assessment (mobile app) and the Withings ScanWatch (physical activity, physiological signals). Diagnosis and treatment decisions are based solely on the first pathway. Outcomes include staff time (primary), technical performance, diagnostic yield, patient-reported measures, PAP adherence and symptom change. The sample size is calculated as 1000 randomised participants. Ethical approval is obtained at all participating centres. The trial is registered with ISRCTN (ISRCTN17539031). Data acquisition is performed between Q1 2024 and Q2 2025; data cleaning and analyses are ongoing.
Verk
Public Value Management in Digital Transformation : Emergent technology's impact on public services
(Háskólinn í Reykjavík, 2026-10) Hjaltalín, Illugi Torfason; Sigurðarson, Hallur Þór; Candi, Marina; Department of Business and Economics
This thesis addresses an important knowledge gap by examining the government-as-user role, a perspective that has received limited attention in artificial intelligence and digital governance research. Emergent technology, including artificial intelligence (AI) and algorithmic systems, is reshaping public sector digital transformation, raising important questions about how public value is framed, created, and enacted. This PhD research investigates how governance strategies and mechanisms shape the implementation of digital transformation initiatives in the public sector, and how emergent technology influences the framing and distribution of public value. The research approach is pluralistic, combining content and discourse analysis with theory-driven evaluation of digital transformation policies and service delivery programs to examine the discursive and practical dimensions of digital transformation across four papers. Papers I and II apply discourse analysis to digital policy and strategy documents, analyzing how public values and power dynamics are articulated, how users and public sector AI applications are positioned, and how these framings allocate benefits and responsibilities among stakeholders in digital governance. Papers III and IV draw on a longitudinal case of Digital Iceland (Stafrænt Ísland). Paper III systematically examines the adoption of a digital shared-services center, focusing on the emergence of a “center of excellence” and the functional components — organizational, technical, and governance-related — that support technology-enabled reforms in government. These components inform the contextual conditions used in developing the evaluative framework in Paper IV. Paper IV employs a theory-driven evaluation approach, adopting and adapting a context–mechanism–outcome (CMO) framework to evaluate Digital Iceland, and examines under what contextual conditions governance strategies and arrangements iv — in particular mechanisms such as flexible policy, accountability, and collaboration — are configured to generate or constrain public value outcomes. The originality of this approach lies in combining content and discourse analysis of digital policies with a theory-driven CMO logic to evaluate digital transformation programs, and in developing a PVM in Digital Transformation framework that evaluates digital governance arrangements. The thesis treats digital strategies and initiatives as analytically distinct yet interconnected domains, linking discursive positioning in national digital and AI strategies to the configuration of governance mechanisms in concrete programs. The framework argues that digital transformation outcomes are not determined by technology itself, but by how emergent technology applications are enacted within specific institutional arrangements and governance mechanisms, which in turn shapes whether public value is created, distributed, or destroyed in practice. It serves both analytical and practical purposes: analytically, by synthesizing the main findings into core concepts for evaluating digital governance; and practically, by considering how administrators translate public value framings into practice, how context-dependent mechanisms enable or constrain their efforts, and how these configurations yield public value outcomes such as improved service delivery, transparency, and responsiveness. The research shows both enabling and constraining effects of dominant discourses and organizational routines, identifies key barriers to agile digital transformation, and underscores the importance of inclusive governance strategies that align diverse stakeholder interests. Overall, this thesis advances evaluative e-government and digital governance research by offering a critical and realistic account of PVM that traces under what conditions governance mechanisms contribute to public value and when they become constrained in public sector digital transformation. In doing so, it bridges the longstanding gap between v policy framing and program outcomes, providing both a theoretically grounded evaluation framework and an empirically substantiated model for assessing how digital governance initiatives generate public value.
Verk
Longitudinal investigation of sleep and circadian rest-activity rhythms in non-small cell lung cancer patients during immunotherapy and associations with fatigue, depression, and perceived stress
(2025-10) Strøm, Louise; Zachariae, Robert; Wu, Lisa M.; Meldgaard, Peter; Ancoli-Israel, Sonia; Lekander, Mats; Mroczek, Daniel; Amidi, Ali; Department of Psychology
Background: Patients with advanced non-small cell lung cancer (NSCLC) face poor prognosis. However, new therapies like immune checkpoint inhibitors (ICIs) have improved survival. Nevertheless, ICIs act through immune activation – a proposed underlying mechanism of frequently reported symptoms such as sleep and circadian disturbances. The objective of this study was to longitudinally investigate sleep and circadian rest-activity rhythms and associations with psychological symptoms in NSCLC during ICI treatment. Methods: Newly diagnosed NSCLC patients were included in this prospective study, assessing insomnia severity, total sleep time, circadian rest-activity rhythms, and psychological symptoms (fatigue, depression, and stress) during the first five months of ICI treatment, using validated questionnaires, sleep diaries and actigraphy. Results: Forty-nine patients were included. Prior to treatment, patients slept on average 5.8 hours/night and 49 % reported clinical levels of insomnia. Spontaneous improvements in insomnia severity and total sleep time were observed, along with marginal but non-significant improvements in circadian rest-activity rhythm robustness. Insomnia severity was significantly associated with higher levels of fatigue (p = .004), depression (p = .007), and perceived stress (p = .033). Moreover, lower circadian rest-activity robustness was associated with more fatigue (p = .021). At baseline, mean levels of fatigue, depression, and perceived stress were below clinical levels and declined further over time. Conclusion: Sleep disturbances, circadian disruption, and psychological symptom burden in NSCLC patients were greatest prior to and shortly after initiation of ICI treatment, with gradual improvement over the first five months. However, inter-individual differences in sleep and circadian disturbances were evident and linked to psychological symptoms.
Verk
From Farm to Fork : The Role of AI in Advancing Sustainable Business Practices
(2025-07-31) Sigurjonsson, Throstur Olaf; Wendt, Stefan; Department of Business and Economics
This study examines how artificial intelligence (AI) supports environmental and social sustainability in the Icelandic food industry across the full value chain—from production to retail. Using a qualitative design, five semi-structured interviews were conducted with senior executives from food production, processing, and distribution firms. Thematic analysis revealed that AI is primarily applied for automation, quality control, energy efficiency, and workplace safety. Although not initially adopted for sustainability purposes, AI delivers indirect benefits such as reduced material use, lower energy and water consumption, and minimized food waste. For instance, AI-based quality control improved efficiency in meat and fish processing, while predictive tools enhanced inventory accuracy in retail. Importantly, no evidence was found of rebound effects or increased resource use. Instead, AI contributes to sustainability mainly through operational improvements. Its adoption is largely driven by external factors such as labor shortages, regulatory pressures, and consumer expectations. Despite its promise, challenges remain, particularly in data-sharing, implementation costs, and public trust. The study highlights the importance of embedding AI in broader ethical and sustainability frameworks to ensure its long-term contribution to responsible food system transformation.

Flokkar í Opnum vísindum

Veldu flokk til að skoða.

Niðurstöður 1 - 9 af 9

Öll gögn í Opnum vísindum eru vernduð af ákvæðum höfundalaga og með öllum réttindum áskildum, nema annað sé tekið fram.