Decision Support for Healthcare Processes and Services

Permanent URI for this collection


Recent Submissions

Now showing 1 - 5 of 10
  • Item
    Towards a Model for Inclusive Healthcare Access post COVID-19
    ( 2022-01-04) Bhatt, Paras ; Akanfe, Oluwafemi ; Guo, Yuanxiong ; Gong, Yanmin
    The devastation caused by the COVID-19 pandemic has exposed years of cyclic inequalities faced by disadvantaged and minority communities. Unequal access to healthcare and a lack of financial resources further exacerbates their suffering, especially during a pandemic. In such critical conditions, information technology-based healthcare services can be an efficient way of increasing access to healthcare for these communities. In this paper, we put forward a decision model for guiding the distribution of IT-based healthcare services for racial minorities. We augment the Health Belief Model by adding financial and technology beliefs. We posit that financial inclusion of minority populations increases their ability to access technology and, by extension, IT-based healthcare services. Financial inclusion and the use of secure private technologies like federated learning can indeed enable greater access to healthcare services for minorities. Therefore, we incorporate financial, health, and technology tools to develop a model for equitable delivery of healthcare services and test its applicability in different use-case scenarios.
  • Item
    Show Me Your Claims and I'll Tell You Your Offenses: Machine Learning-Based Decision Support for Fraud Detection on Medical Claim Data
    ( 2022-01-04) Matschak, Tizian ; Prinz, Christoph ; Rampold, Florian ; Trang, Simon
    Health insurance claim fraud is a serious issue for the healthcare industry as it drives up costs and inefficiency. Therefore, claim fraud must be effectively detected to provide economical and high-quality healthcare. In practice, however, fraud detection is mainly performed by domain experts resulting in significant cost and resource consumption. This paper presents a novel Convolutional Neural Network-based fraud detection approach that was developed, implemented, and evaluated on Medicare Part B records. The model aids manual fraud detection by classifying potential types of fraud, which can then be specifically analyzed. Our model is the first of its kind for Medicare data, yields an AUC of 0.7 for selected fraud types and provides an applicable method for medical claim fraud detection.
  • Item
    Processing Patient Information Leaflets with Embeddings
    ( 2022-01-04) Stahlmann, Sven ; Hirschmeier, Stefan ; Schoder, Detlef
    As of 2021, more than 100,000 drugs are approved in Germany, 35,000 of which are non-prescriptive over-the-counter drugs. While proven information from medical studies is given in patient information leaflets, patients are often lost when trying to determine which drugs are compatible with their needs or which alternatives are suitable. We show that representing patient information leaflets as dense vectors allows us to extract more valuable medical information than is explicitly stated in the leaflets. Without any explicit insertion of medical knowledge, our embeddings capture concepts of generics, even with respect to the dosage form. Furthermore, the embeddings allow patients to identify drug clusters based on their treatment area and offer suitable alternatives based on analogical reasoning. The carved-out information may not only help patients to explore alternative drugs but also serve pharmacists and patients as a new way to search for drugs tailored to dietary, allergic, or medical needs.
  • Item
    On Intelligence Augmentation and Visual Analytics to Enhance Clinical Decision Support Systems
    ( 2022-01-04) Heart, Tsipi ; Padman, Rema ; Ben-Assuli, Ofir ; Gefen, David ; Klempfner, Robert
    Human-in-the-loop intelligence augmentation (IA) methods combined with visual analytics (VA) have the potential to provide additional functional capability and cognitively driven interpretability to Decision Support Systems (DSS) for health risk assessment and patient-clinician shared decision making. This paper presents some key ideas underlying the synthesis of IA with VA (IA/VA) and the challenges in the design, implementation, and use of IA/VA-enabled clinical decision support systems (CDSS) in the practice of medicine through data driven analytical models. An illustrative IA/VA solution provides a visualization of the distribution of health risk, and the impact of various parameters on the assessment, at the population and individual levels. It also allows the clinician to ask “what-if” questions using interactive visualizations that change actionable risk factors of the patient and visually assess their impact. This approach holds promise in enhancing decision support systems design, deployment and use outside the medical sphere as well.
  • Item
    Monte-Carlo Simulation Based on Patient-Individual Distributions for Supporting Intensive Care Occupancy Management
    ( 2022-01-04) Witteborg, Arne ; Borgstedt, Rainer ; Günther, Markus ; Jansen, Gerrit ; Römer, Michael
    Managing Intensive Care Units (ICUs) in hospitals is a highly challenging endeavor. In particular, decisions such as admitting elective patients and discharging patients from the ICU have to be taken under a high level of uncertainty since the occupancy of ICUs does not only depend on these decisions but also on unknown parameters such emergency patient arrivals and lengths of stay of the patients in the ICU. In this paper, we develop a framework for supporting ICU occupation management by quantifying the impact of admission and discharge decisions on the probability of reaching critical ICU occupancy levels in a given planning horizon. A key component of this framework is the use of data-driven approaches for obtaining probability distributions for the parameters affected by uncertainty. In particular, we use standardized treatment and patient health state data to create patient-specific length-of-stay distributions with a Machine Learning approach. These patient-individual distributions are then validated and/or adjusted by medical experts. The validated distributions form the input to a Monte-Carlo Simulation that is used to approximate the probability distributions of the daily ICU occupancy levels resulting from ICU admission and discharge decisions. We experimentally evaluate our framework in a counterfactual simulation based on one year of historical data from 2019 from a medium-sized ICU in a German hospital. In that evaluation, we use a simple ICU management policy based on the probabilistic occupancy forecasts aiming at reducing the risk of running out of ICU capacity. The results show that following this policy would have avoided hitting critical occupancy levels by around 70% and would have had a smoothing effect on ICU occupancy levels.