Health Care Management Science最新文献

Harnessing the synergy between artificial intelligence (AI) and operations research (OR) helps drive efficiency, safety, and innovation in biomanufacturing. AI offers predictive capabilities, while OR represents the pinnacle of prescriptive analytics. AI and OR complement each other by offering structured, interpretable, and verifiable solutions to complex operational challenges. In this commentary, we reflect on how to realize the full potential of AI-OR implementations in biomanufacturing. We elaborate on recent university-industry partnerships demonstrating these benefits and propose a roadmap for AI-OR integration in biomanufacturing.

Contagious diseases severely impact health systems and economies, with close contact leading to further spread and fatalities. This paper examines the effects of government interventions on controlling such diseases. Key interventions include media isolation of susceptible individuals, effective quarantining of infected persons, and vaccination. A system dynamics approach models the complexities of government interventions in coronary conditions. We used the SEIR (Susceptible, Exposed, Infected, and Recovered) model and developed a new model to address its shortcomings for a new virus. Resilience actions were defined and plotted based on the emergency management cycle phases: Prevention, Preparedness, Response, and Recovery. The model can be applied to any contagious disease worldwide. We calibrated the model using data from sources like the World Health Organization (WHO) and Centers for Disease Control (CDC), and validated it against official and historical data. A sensitivity analysis was conducted based on various resilience strategies: Isolation Rate Slope, Isolation Efficiency, Minimum Isolation Rate, Quarantine Portion, Quarantine Transmission, Vaccination Rate, and Media Rate Slope. The study identifies key conditions for controlling outbreaks: achieving rapid isolation with a minimum rate above 50% and efficiency above 95%, rapid detection and quarantine above 90% with efficiency over 92%, and an optimal contact rate below 0.2, achieved with a media rate slope of 0.005 and vaccination rate above 90%. These measures can control the disease within 455 days or less.

In the context of diagnosis of COVID-19, this paper shows how to convert a Causal Network to a Large Language Model (LLM). The Causal Network was converted to the language model using prompts and completions. Prompts were composed from the full-factorial combination of the text associated with statistically significant variables in the Causal Network. Completions were based on the evaluation of the probability of COVID-19 using the Causal Network. The accuracy of the Causal Network and LLM was tested using two databases. The first database was based on a survey of 822 patients, collecting 12 direct (parents on the Markov blanket of COVID-19 diagnosis node), 7 indirect (associated with COVID-19 but not direct cause) symptoms of COVID-19. The second set was based on 80 patients reporting their symptoms in open-ended questions, often reporting some of the direct predictors and rarely reporting any indirect predictors of COVID-19. The accuracy of Causal Network and Markov blanket was tested using Area under the Receiver Operating Curve (AUROC). When indirect information was available, the Causal Network model (AUROC = 0.91) was significantly more accurate than the LLM (AUROC = 0.88), even though LLM model was trained to duplicate predictions of the Causal Network. Where the indirect information was not available, both models had lower accuracy (AUROC of 0.75 and 0.76). The accuracy of LLM depends not only on patterns among direct predictors of the outcome but also data not reported to the LLM. Conversational LLMs need to go beyond information the patient supplies and proactively ask about missing, typically indirect, information.

Motivated by our collaboration with a residency program at an academic health system, we propose new integer programming (IP) approaches for the resident-to-rotation assignment problem (RRAP). Given sets of residents, resident classes, and departments, as well as a block structure for each class, staffing needs, rotation requirements for each class, program rules, and resident vacation requests, the RRAP involves finding a feasible year-long rotation schedule that specifies resident assignments to rotations and vacation times. We first present an IP formulation for the RRAP, which mimics the manual method for generating rotation schedules in practice and can be easily implemented and efficiently solved using off-the-shelf optimization software. However, it can lead to disparities in satisfying vacation requests among residents. To mitigate such disparities, we derive an equity-promoting counterpart that finds an optimal rotation schedule, maximizing the number of satisfied vacation requests while minimizing a measure of disparity in satisfying these requests. Then, we propose a computationally efficient Pareto Search Algorithm capable of finding the complete set of Pareto optimal solutions to the equity-promoting IP within a time that is suitable for practical implementation. Additionally, we present a user-friendly tool that implements the proposed models to automate the generation of the rotation schedule. Finally, we construct diverse RRAP instances based on data from our collaborator and conduct extensive experiments to illustrate the potential practical benefits of our proposed approaches. Our results demonstrate the computational efficiency and implementability of our approaches and underscore their potential to enhance fairness in resident rotation scheduling.

Operating room (OR) scheduling is a critical challenge in healthcare, directly impacting patient outcomes and hospital efficiency. Traditional methods often struggle with the complex, multi-level constraints and uncertainties inherent in OR scheduling, such as resource limitations, variable surgery durations, and emergency cases. This study aims to develop a novel hybrid framework that optimizes OR scheduling by integrating multi-level optimization with reinforcement learning and column generation techniques. The proposed framework decomposes the OR scheduling problem into strategic, tactical, and operational levels, enabling focused optimization at each layer while ensuring cohesive decision-making across the hierarchy. Reinforcement learning guides the column generation process, learning policies that identify promising scheduling options to enhance solution quality and computational efficiency. Robust uncertainty handling mechanisms are incorporated to manage variability in surgery durations and resource availability without compromising tractability. Experiments were conducted using three years of real-world data from Shanxi Provincial People's Hospital, complemented by large-scale synthetic datasets to evaluate scalability and robustness of the framework. The framework demonstrates meaningful improvements in key operational metrics compared to traditional approaches. Analysis of three years of implementation shows consistent enhancements in operational efficiency, including a reduction in average patient waiting time by 15.8% (from 10.1 to 8.5 days), an increase in OR utilization by 5.4 percentage points (from 73.8% to 79.2%), and improved workload balance among surgeons. The framework maintains robust performance under uncertainty, achieving a 92.5% feasibility rate and reducing schedule disruptions by 26.2%. The proposed hybrid framework offers a practical and scalable solution for optimizing OR scheduling, demonstrating improvements in healthcare delivery and operational performance in real hospital environments. By effectively balancing multiple operational objectives while handling practical constraints and uncertainties, the framework provides a viable approach for healthcare systems seeking incremental yet sustainable improvements in efficiency and patient care.

Surgery scheduling is crucial in healthcare management, particularly in hospitals and clinics. This study tackles the elective surgery scheduling problem by integrating affinity and preferences among the surgical team's members. Although these concepts can enhance coordination and improve team performance, they remain understudied in the literature. Affinity is usually quantified as a numerical representation of compatibility between team members, and preferences denote a surgeon's interest in specific surgical resources. Existing approaches have not integrated simultaneously affinity and preferences. In addition, they use mathematical programming models that often incorporate affinity and preferences as constraints or additional objective function terms, adopting a multi-objective approach. The former can significantly reduce the number of surgeries performed, while the latter increases computational complexity. To overcome these limitations, we propose mathematical programming models with a score-based penalty approach that integrates affinity and preferences while maximizing the priority of scheduled surgeries. Our approach is evaluated against two alternative models: a baseline model without affinity or preferences and a constraint-based model that follows conventional literature, incorporating these concepts as hard constraints. We implement these models using integer linear programming and constraint programming. The results show the feasibility of considering affinity and preferences among surgical team members. This can enhance the surgical team's quality with negligible impact on the number of surgeries performed. Therefore, our approach can generate stronger human relationships among surgical team members, which could contribute positively to patient surgical outcomes, as demonstrated by some studies in the literature.

The diagnosis section in hospital discharge summaries plays a critical role in ensuring continuity of care by providing essential diagnostic information and a succinct summary of a patient's condition to subsequent caregivers. However, the lack of standardized structure and content can lead to incomplete, ambiguous, or inaccurate documentation, potentially compromising patient safety. This study takes a foundational step toward standardizing the diagnosis section in German, and potentially international, discharge summaries by developing a taxonomy of structural and content elements and examining the use of standardized terminologies and abbreviations. We conducted a retrospective analysis of 436 de-identified discharge summaries from 112 hospitals across 12 German states. A structured taxonomy development process was applied, supported by natural language processing, to examine structural and content elements as well as the use of standardized terminologies (SNOMED-CT, ICD-10 codes) and abbreviations. The resulting taxonomy for diagnosis sections comprises 87 distinct characteristics across three meta-dimensions: structure, content, and levels of detail. The analysis revealed limited adoption of standardized terminologies; only 8.1% of terms conformed to SNOMED-CT, and only 14.2% of diagnosis sections included ICD-10 codes. Abbreviations appeared in 92% of diagnosis sections, constituting 14.5% of all words, many of which were obscure or infrequently used. These findings underscore the urgent need for a standardized, interoperable, and clinically meaningful diagnosis section to support continuity of care and data-driven healthcare. The proposed taxonomy offers a foundational framework for future standardization efforts by providing structural and content "design options."

Pandemics pose significant challenges, particularly in the early stages when vertical resupply chains are overwhelmed. To mitigate the impact of medical resource shortages, we develop a multi-period optimization model incorporating lateral transshipment and hospital admission to minimize the total number of infected individuals by strategically allocating regional resources in the face of complex dynamics, including endogenous hospital admission rates and pandemic spread. To capture the temporal-spatial nature of pandemics, we extend the Susceptible-Exposed-Infected-Hospitalized-Recovered (SEIHR) model by accounting for population migration. Additionally, we derive threshold-type structures for optimal resource transfers, considering factors such as pandemic dynamics, patient length of stay, and budget constraints. We also demonstrate the effectiveness of our models via numerical experiments. Our research identifies three main findings: 1) Pooling medical resources effectively reduces infections and alleviates shortages in outbreak areas. This strategy is particularly beneficial during pandemics due to self-reinforcing infection dynamics and surging demand. 2) Regions adjacent to the epicenter should exercise caution in contributing resources to avoid exacerbating infections through population migration. 3) While effective in localized outbreaks, widespread resource scarcity can limit the viability of pooling strategies, potentially leading to increased infections and fluctuating resource levels in transferring regions.

Community Health Houses (CHHs) are new entities in the Italian National Health Service that have been envisaged to provide proximity care to an increasingly aging population, and bear some similarities to other facilities in countries that have historically focused on public healthcare. This work proposes an integrated decision support system (DSS) for their planning, envisioned during the aftermath of the COVID-19 pandemic, which highlighted the frailty of the existing system. The DSS is based on an integer linear programming (ILP) model that simultaneously makes location, districting and dimensioning decisions for CHH, and accounts for accessibility and equity requirements. Based on Italian law yet designed in a parametrized way that makes it adaptable to several contexts, the DSS is able to design a hub and spoke network, which considers the provision of both mandatory and additional services. The sizes of the former are determined by directly taking into account the population served, while those of the latter are determined according to the specific demand for these services, accounting for diverse needs arising from different territories. The DSS also uses territorial units that refer to recognizable administrative areas. This ensures that the districting is easily recognized and accepted by the population. In addition to the ILP formulation, three decomposition-based matheuristics are proposed, which allow suitable solutions to be found within a reasonable time also for large and heterogeneous instances, while maintaining the flexibility of the ILP formulation. Computational results on synthetic realistic instances validated the DSS, while its application to a real-life case in a Northern Italian province demonstrated the effectiveness of the heuristic approaches and provided a proof of concept for its practical application.