AKAlexandros Katsiferis

Alexandros Katsiferis

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Alexandros Katsiferis has a total of 70 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Development and validation of early warning scores predicting 24 h mortality addressing intercurrent medical interventions in Denmark: a multicentre retrospective cohort study of 2·08 million hospital encounters

    by Alexandros Katsiferis, Neil Scheidwasser, Tri-Long Nguyen, Theis Lange, Mark P Khurana, Pernille B Nielsen, Kasper Karmark Iversen, Christian S Meyhoff, Eske Kvanner Aasvang, Jesper Mølgaard, et al.

  • Open Access

    SARS-CoV-2 reinfections and subsequent risk of hospital-diagnosed post-acute sequelae in Denmark (2020–2022): a nationwide cohort study

    by Mark P. Khurana, Mathilde Marie Brünnich Sloth, Neil Scheidwasser, Jacob Curran-Sebastian, Christian Morgenstern, Nicolas Banholzer, David Thein, Laust H. Mortensen, Morten Rasmussen, Pikka Jokelainen, et al.


  • Open Access

    Transmission Networks and Intervention Effects From SARS-CoV-2 Genomic and Social Network Data in Denmark

    by Jacob Liam Curran-Sebastian, Christian Morgenstern, Jonas Juul, Mark Poulsen Khurana, Neil Alexandre Scheidwasser, Nicolas Banholzer, Alexandros Katsiferis, Frederik Trier Møller, The Danish COVID-19 Genome Consortium, Pikka Jokelainen, et al.

    Decision-making by governments during disease outbreaks is increasingly reliant on large-scale pathogen genomics and detailed individual-level data. We use 293,841 SARS-CoV-2 genomes collected in Denmark, between September 1st 2020 and December 31st 2021, and combine these with comprehensive individual-level data on settings, including households, schools, workplaces and family relationships, to infer plausible transmission pathways. Next, we use the plausible transmission pathways to assess the effectiveness of specific non-pharmaceutical interventions (NPIs) and quantify transmission heterogeneities, providing a more detailed understanding than is possible from aggregate national estimates. School restrictions were associated with reduced community transmission; face coverings and gathering restrictions were associated with reduced transmission across all settings. We find that those vaccinated with two doses, compared to the unvaccinated, had onward transmission reduced by 23.5% (95% highest posterior density interval: 21.5-25.2%). Our proposed approach is pathogen agnostic, and can be used in future outbreaks where genomic data and data on social relationships are available.


  • Open Access

    “Developing machine learning models of self-reported and register-based data to predict eating disorders in adolescence”

    by Alexandros Katsiferis, Andrea Joensen, Liselotte Vogdrup Petersen, Claus Thorn Ekstrøm, Else Marie Olsen, Samir Bhatt, Tri-Long Nguyen, Katrine Strandberg Larsen

    Early detection and prevention of eating disorders (EDs) in adolescence are crucial yet challenging. We developed and validated diagnostic and prognostic models to predict EDs using data from 44,357 Danish National Birth Cohort participants. Models were trained to identify ED presence in early and late adolescence (11- and 18-year follow-up), utilizing approximately 100 predictors from self-reported and registry-based data. The machine learning model demonstrated strong discrimination for both tasks (diagnostic Area Under the receiver operating characteristic Curve = 81.3; prognostic AUC = 76.9), while a logistic regression model using the top 10 predictors achieved comparable performance. Sex, emotional symptoms, peer relationship and conduct problems, stress levels, parental BMI values, body dissatisfaction, and BMI at the 7-year follow-up emerged as key predictors. Our models showed potential utility in supporting clinical risk assessment, particularly for low-risk preventive interventions, though further validation studies are needed to evaluate their effectiveness in real-world clinical settings.


  • Open Access

    “Complex models, marginal benefits--a multi-centre development and validation study of early warning scores across 2·16 million patient admissions addressing intercurrent medical interventions”

    by Alexandros Katsiferis, Neil Scheidwasser, Tri-Long Nguyen, Theis Lange, Mark P Khurana, Pernille B Nielsen, Kasper Karmark Iversen, Christian S Meyhoff, Eske Kvanner Aasvang, Jesper Moelgaard, et al.

    Summary: Background: The National Early Warning Score (NEWS) is a nationally recommended, clinically implemented system, used to prevent patient deterioration. While numerous studies have compared predictive models for clinical deterioration, large-scale evaluations of their potential clinical utility remain undetermined. Here, we compared NEWS’s clinical net benefit against simplified scoring rules and modern machine learning to determine whether simpler approaches are sufficient or if complex models provide meaningful advantages in clinical practice. Methods: We included fifteen Danish hospitals with over 2·16 million patient admissions representing 829 610 unique patients over five years (2018 to 2023). We compared NEWS against both simpler and more complex approaches for predicting 24-hour mortality: NEWS-Light (NEWS without blood pressure and temperature), DEWS (NEWS-Light with age and sex; DEWS denotes the Demographic Early Warning Score), and a model based on eXtreme Gradient Boosting (XGB-EWS) incorporating vital signs, demographics, laboratory markers, plus medical history embeddings extracted using sentence transformers. We used propensity score weighting to mitigate intervention bias and evaluated performance using Area Under the Receiver Operating Characteristic Curve (AUC), calibration, and net benefit. Findings: XGB-EWS achieved the highest discrimination (AUC 0·932, 95% Confidence Interval [0·929-0·936]), followed by DEWS (0·908 [0·904-0·912]), NEWS (0·902, [0·898-0·906]), and NEWS-Light (0·879, [0·873-0·885]). Decision curve analysis showed maximum net benefit differences of 1·8 additional correct mortality identifications per 10 000 patients between XGB-EWS and NEWS, and 1·7 per 10 000 between NEWS and NEWS-Light, across the evaluated risk thresholds. Interpretation: Machine learning approaches provided marginal clinical utility improvements over traditional scoring systems, with NEWS-Light showing small performance decrements compared to full NEWS. The clinical significance of these differences must be weighed against workflow optimization benefits, suggesting healthcare systems should evaluate trade-offs between predictive performance and operational efficiency when selecting early warning approaches. Funding: Novo Nordisk Foundation Research in context: Evidence before this study: Early warning score systems have evolved from single vital sign monitoring to standardized multivariable scores such as the National Early Warning Score (NEWS), to even more sophisticated machine learning and deep learning frameworks utilizing electronic health records data. All these early warning scores are primarily designed to guide clinical decision-making by helping identify patients at risk of clinical deterioration within hospitals. Despite these advances, validation studies predominantly focus on statistical metrics measuring discrimination performance rather than meaningful clinical utility. On Aug 12, 2025, we searched PubMed for English-language studies published in the last 10 years, using terms including “early warning score”, “NEWS”, “NEWS2”, “machine learning”, “artificial intelligence”, “decision curve analysis”, “net benefit”, “clinical utility”, and “24-hour mortality.” While most of the published models, including the more sophisticated machine learning ones, have demonstrated better discrimination compared to traditional early warning scores, we found only one study that combined early warning score validation with clinical utility analysis for short-term clinical deterioration. Additionally, no studies have evaluated clinical utility across multiple patient subgroups while using causal inference methods to address intervention bias in a large-scale healthcare system validation. Added value of this study: The current study represents the largest multi-center early warning score validation to date, encompassing 2·16 million patient encounters across 15 Danish hospitals and nine distinct clinical specialties over a period of five years. We used the predictimand framework with causal inference methods to address intervention bias, a critical and novel methodological advance in the assessment of early warning scores. Unlike previous research that focused primarily on discrimination metrics, we evaluated clinical utility using decision curve analysis, providing evidence that sophisticated machine learning early-warning-score approaches delivered in overall only marginal clinical utility improvements (up to 1.7 additional correctly identified mortality cases per 10 000 patients) despite better discrimination. We find that a simplified version of NEWS (NEWS without blood pressure and temperature components) achieves a comparable clinical net benefit to full NEWS. Additionally, we provide the first quantitative assessment of the healthcare resources implications in this area of research, showing that simplified approaches could potentially redirect 98·1 full-time equivalent positions annually from routine vital sign collection to direct patient care, providing evidence for healthcare administrators to reallocate clinical resources toward patient interaction and care delivery. Implications of all the available evidence: We demonstrated that differences in the ability of the models to identify 24-hour mortality cases correctly were less than 2 per 10 000 patients. Given that often in clinical practice, a proportion of these identifications translates to lives being rescued, the real-world clinical advantage is likely even more modest. These modest gains of machine learning approaches, coupled with NEWS-Light displaying marginal performance decrements compared to the full NEWS, challenge the prevailing emphasis on algorithmic complexity over clinical value. NEWS-Light has the potential to enable workflow optimization, freeing approximately 3 minutes per patient encounter without compromising clinical effectiveness. Future research should conduct prospective cluster-randomized controlled trials comparing NEWS versus NEWS-Light implementation, while the broader prediction modeling research community should adopt utility-based evaluation frameworks to ensure that algorithmic advances translate to tangible improvements in patient care rather than statistical superiority alone.

  • Open Access

    Exploring the Potential and Limitations of Deep Learning and Explainable AI for Longitudinal Life Course Analysis

    by Helen Coupland, Neil Scheidwasser, Alexandros Katsiferis, Megan Davies, Seth Flaxman, Naja Hulvej Rod, Swapnil Mishra, Samir Bhatt, H. Juliette T. Unwin

    Background: Understanding the complex interplay between life course exposures, such as adverse childhood experiences and environmental factors, and disease risk is essential for developing effective public health interventions. Traditional epidemiological methods, such as regression models and risk scoring, are limited in their ability to capture the non-linear and temporally dynamic nature of these relationships. Deep learning (DL) and explainable artificial intelligence (XAI) are increasingly applied within healthcare settings to identify influential risk factors and enable personalised interventions. However, significant gaps remain in understanding their utility and limitations, especially for sparse longitudinal life course data and how the influential patterns identified using explainability are linked to underlying causal mechanisms. Methods: We conducted a controlled simulation study to assess the performance of various state-of-the-art DL architectures including CNNs and (attention-based) RNNs against XGBoost and logistic regression. Input data was simulated to reflect a generic and generalisable scenario with different rules used to generate multiple realistic outcomes based upon epidemiological concepts. Multiple metrics were used to assess model performance in the presence of class imbalance and SHAP values were calculated. Results: We find that DL methods can accurately detect dynamic relationships that baseline linear models and tree-based methods cannot. However, there is no one model that consistently outperforms the others across all scenarios. We further identify the superior performance of DL models in handling sparse feature availability over time compared to traditional machine learning approaches. Additionally, we examine the interpretability provided by SHAP values, demonstrating that these explanations often misalign with causal relationships, despite excellent predictive and calibrative performance. Conclusions: These insights provide a foundation for future research applying DL and XAI to life course data, highlighting the challenges associated with sparse healthcare data, and the critical need for advancing interpretability frameworks in personalised public health.


  • Open Access

    Cause-specific mortality after spousal bereavement in a Danish register-based cohort

    by Mathilde Marie Brünnich Sloth, Johannes Hruza, Laust Hvas Mortensen, Samir Bhatt, Alexandros Katsiferis

    Spousal bereavement is associated with increased all-cause mortality, but less is known for cause-specific mortality. We estimated the average effect of bereavement on cause-specific mortality. Furthermore, we developed and evaluated the performance of prediction models estimating individualized mortality risks. This matched cohort study leveraged nationwide registry data on 223,500 married Danish individuals aged ≥ 65 years. Individuals were followed from bereavement date until death, emigration, or 3-year end of follow-up. G-estimation was used to estimate the average effect of bereavement on cause-specific mortality. Risk prediction models were developed and internally validated using logistic regression and Extreme Gradient Boosting utilizing information on sociodemographic factors and healthcare expenditures prior to bereavement. Among males, bereavement was associated with increased 3-year mortality from cardiovascular disease (CVD) (Risk Difference: 8 deaths/1000 individuals [95% confidence intervals (CI) 3;13]), digestive diseases (3 [95% CI 1;5]), psychiatric diseases/suicide (3 [95% CI 1;5]), and respiratory diseases (4 [95% CI 1;8]) and decreased 3-year mortality for dying from dementia/Parkinson’s disease (-4 [95% CI -6;-2]). Among females, bereavement was only associated with increased 3-year mortality from CVD (5 [95% CI 3;7 ]) and from psychiatric diseases/suicide (1 [95% CI 1;2]). The prediction models developed using sociodemographic information and healthcare expenditures showed limited accuracy in identifying cause-specific mortality risk. On average, spousal bereavement slightly elevated the risk for most causes of death for males and less so for females. Despite incorporating comprehensive healthcare and sociodemographic data, accurately predicting individual cause-specific mortality remains challenging.


  • Open Access

    Machine learning models of healthcare expenditures predicting mortality: A cohort study of spousal bereaved Danish individuals

    by Alexandros Katsiferis, Samir Bhatt, Laust Hvas Mortensen, Swapnil Mishra, Majken Karoline Jensen, Rudi G. J. Westendorp

    Background: The ability to accurately predict survival in older adults is crucial as it guides clinical decision making. The added value of using health care usage for predicting mortality remains unexplored. The aim of this study was to investigate if temporal patterns of healthcare expenditures, can improve the predictive performance for mortality, in spousal bereaved older adults, next to other widely used sociodemographic variables. Methods: This is a population-based cohort study of 48,944 Danish citizens 65 years of age and older suffering bereavement within 2013–2016. Individuals were followed from date of spousal loss until death from all causes or 31st of December 2016, whichever came first. Healthcare expenditures were available on weekly basis for each person during the follow-up and used as predictors for mortality risk in Extreme Gradient Boosting models. The extent to which medical spending trajectories improved mortality predictions compared to models with sociodemographics, was assessed with respect to discrimination (AUC), overall prediction error (Brier score), calibration, and clinical benefit (decision curve analysis). Results: The AUC of age and sex for mortality the year after spousal loss was 70.8% [95% CI 68.8, 72.8]. The addition of sociodemographic variables led to an increase of AUC ranging from 0.9% to 3.1% but did not significantly reduce the overall prediction error. The AUC of the model combining the variables above plus medical spending usage was 80.8% [79.3, 82.4] also exhibiting smaller Brier score and better calibration. Overall, patterns of healthcare expenditures improved mortality predictions the most, also exhibiting the highest clinical benefit among the rest of the models. Conclusion: Temporal patterns of medical spending have the potential to significantly improve our assessment on who is at high risk of dying after suffering spousal loss. The proposed methodology can assist in a more efficient risk profiling and prognosis of bereaved individuals.


  • Open Access

    Predicting mortality risk after a fall in older adults using health care spending patterns: a population-based cohort study

    by Alexandros Katsiferis, Laust Hvas Mortensen, Mark P Khurana, Swapnil Mishra, Majken Karoline Jensen, Samir Bhatt

    To develop a prognostic model of 1-year mortality for individuals aged 65+ presenting at the emergency department (ED) with a fall based on health care spending patterns to guide clinical decision-making. Population-based cohort study (n = 35,997) included with a fall in 2013 and followed 1 year. Health care spending indicators (dynamical indicators of resilience, DIORs) 2 years before admission were evaluated as potential predictors, along with age, sex and other clinical and sociodemographic covariates. Multivariable logistic regression models were developed and internally validated (10-fold cross-validation). Performance was assessed via discrimination (area under the receiver operating characteristic curve, AUC), Brier scores, calibration and decision curve analysis. The AUC of age and sex for mortality was 72.5% [95% confidence interval 71.8 to 73.2]. The best model included age, sex, number of medications and health care spending DIORs. It exhibited high discrimination (AUC: 81.1 [80.5 to 81.6]), good calibration and potential clinical benefit for various threshold probabilities. Overall, health care spending patterns improved predictive accuracy the most while also exhibiting superior performance and clinical benefit. Patterns of health care spending have the potential to significantly improve assessments on who is at high risk of dying following admission to the ED with a fall. The proposed methodology can assist in predicting the prognosis of fallers, emphasising the added predictive value of longitudinal health-related information next to clinical and sociodemographic predictors.


  • Open Access

    Assessing racial bias in type 2 diabetes risk prediction algorithms

    by Héléne T. Cronjé, Alexandros Katsiferis, Leonie K. Elsenburg, Thea O. Andersen, Naja H. Rod, Tri-Long Nguyen, Tibor V. Varga

    Risk prediction models for type 2 diabetes can be useful for the early detection of individuals at high risk. However, models may also bias clinical decision-making processes, for instance by differential risk miscalibration across racial groups. We investigated whether the Prediabetes Risk Test (PRT) issued by the National Diabetes Prevention Program, and two prognostic models, the Framingham Offspring Risk Score, and the ARIC Model, demonstrate racial bias between non-Hispanic Whites and non-Hispanic Blacks. We used National Health and Nutrition Examination Survey (NHANES) data, sampled in six independent two-year batches between 1999 and 2010. A total of 9,987 adults without a prior diagnosis of diabetes and with fasting blood samples available were included. We calculated race- and year-specific average predicted risks of type 2 diabetes according to the risk models. We compared the predicted risks with observed ones extracted from the US Diabetes Surveillance System across racial groups (summary calibration). All investigated models were found to be miscalibrated with regard to race, consistently across the survey years. The Framingham Offspring Risk Score overestimated type 2 diabetes risk for non-Hispanic Whites and underestimated risk for non-Hispanic Blacks. The PRT and the ARIC models overestimated risk for both races, but more so for non-Hispanic Whites. These landmark models overestimated the risk of type 2 diabetes for non-Hispanic Whites more severely than for non-Hispanic Blacks. This may result in a larger proportion of non-Hispanic Whites being prioritized for preventive interventions, but it also increases the risk of overdiagnosis and overtreatment in this group. On the other hand, a larger proportion of non-Hispanic Blacks may be potentially underprioritized and undertreated.