One key factor contributing to emergency department (ED) overcrowding is prolonged waiting time for admission inpatient wards, also known as ED boarding time. To gain insights into reducing this time, we study operations in the wards and their interface with ED. We focus on understanding effect of discharge policies other operational time-of-day performance, such fraction patients longer than six hours before being admitted. Based an empirical at a Singaporean hospital, propose novel...

Problem definition: Inpatient beds are usually grouped into several wards, and each ward is assigned to serve patients from certain “primary” specialties. However, when a patient waits excessively long before primary bed becomes available, hospital managers have the option assign her nonprimary bed. although it undesirable. Deciding use such “overflow” difficult in real time under uncertainty. Relevance: To aid decision making, we model inpatient flow as multiclass, multipool parallel-server...

Abstract Background During the 2009 H1N1 influenza pandemic, concerns arose about potential negative effects of mass public gatherings and travel on course pandemic. Better understanding temporal changes in social mixing patterns could help officials determine if when to cancel large or enforce regional restrictions, advisories, surveillance during an epidemic. Methods We develop a computer simulation model using detailed data from state Georgia explore how various contact patterns,...

We performed a descriptive study of operating room (OR) case scheduling within 1 week the day surgery.The data used were from and transaction audit tables hospital's anesthesia OR information management systems. Each change to scheduled in system was captured an table, including date time when made. The timestamps allowed reconstruction elective schedule for each surgery at preceding dates (e.g., 2 workdays ahead). sample size n = 17 consecutive 4-week periods. allocated time, combination...

We analyze a time-varying M peri /Geo 2timeScale /N queueing system. The arrival process is periodic Poisson. service time of customer has components in different scales: length stay (LOS) days and departure (h dis ) hours. This system been used to study patient flows from the emergency department (ED) hospital inpatient wards. In that setting, LOS simply number she spends ward, her h discharge hour on day discharge. develop new analytical framework can perform exact analysis this novel two...

One of the most important decisions a hospitalist makes at intersection cost and quality care is when to discharge patient from hospital. Keeping patients longer (shorter) increases (decreases) overcrowding hospital costs but also decreases (increases) readmission risk. Here long-run average optimization problem for determining on each day who how many developed. The authors combined structural properties model with an analytical solution special structure approximately solve...

BACKGROUND: In previous studies, hospitals' operating room (OR) schedules were influenced markedly by decisions made within a few days of surgery. At an academic hospital, 46% ORs had their last case scheduled or changed 1 working day surgery, and private hospital 64%. Many these changes for patients who admitted before surgery (i.e., inpatient cases). this study, we investigate the impact on OR productivity how cases are METHODS: We consider case-scheduling choice between 2 ORs. compare 3...

SUMMARY As the 2009 H1N1 influenza pandemic (H1N1) has shown, public health decision-makers may have to predict subsequent course and severity of a pandemic. We developed an agent-based simulation model used data from state Georgia explore influence viral mutation seasonal effects on showed that when begins in April certain conditions can lead second wave autumn (e.g. degree seasonality exceeding 0·30, or daily rate immunity loss 1% per day). Moreover, combinations variables reproduced...

Abstract Background Advanced machine learning models have received wide attention in assisting medical decision making due to the greater accuracy they can achieve. However, their limited interpretability imposes barriers for practitioners adopt them. Recent advancements interpretable tools allow us look inside black box of advanced prediction methods extract while maintaining similar accuracy, but few studies investigated specific hospital readmission problem with this spirit. Methods Our...

When having access to demand forecasts, a crucial question is how effectively use this information make better resource allocation decisions, especially during surges like the COVID-19 pandemic. Despite emergence of various advanced prediction models for hospital resources, there has been lack prescriptive solutions managers seeking concrete decision support, example, guidance on whether allocate beds from other specialties meet surge patients by postponing elective surgeries. In their paper...

This document presents a comprehensive empirical study on the inpatient flow management in Singaporean hospital. The uses high resolution patient data from 2008 to 2010. details statistics of waiting times for patients admitted emergence department (ED) wards, bed occupancy rate and overflow proportions (proportions that are non-primary ward). also reports various related arrival, discharge, length stay (LOS), pre- post-allocation delays incurred during bed-assignment process.

Inpatient flow management plays a critical role in care delivery, patient outcomes, and hospital operational financial costs. Modeling performance analysis of inpatient present unique features challenges that differ from operations other service industries. In this study, we review recent modeling analytical advances the setting management, with particular focus on time models motivated observations discharges. We first compare two new developed to capture time‐of‐day dynamics, reveal...

This paper considers stochastic linear bandits with general nonlinear constraints. The objective is to maximize the expected cumulative reward over horizon $T$ subject a set of constraints in each round $\tau\leq T$. We propose pessimistic-optimistic algorithm for this problem, which efficient two aspects. First, yields $\tilde{\cal O}\left(\left(\frac{K^{0.75}}{\delta}+d\right)\sqrt{\tau}\right)$ (pseudo) regret T,$ where $K$ number constraints, $d$ dimension feature space, and $\delta$...

The fairness in machine learning is getting increasing attention, as its applications different fields continue to expand and diversify. To mitigate the discriminated model behaviors between demographic groups, we introduce a novel post-processing method optimize over multiple constraints through group-aware threshold adaptation. We propose learn adaptive classification thresholds for each group by optimizing confusion matrix estimated from probability distribution of output. As only need an...

Time-series generation has crucial practical significance for decision-making under uncertainty. Existing methods have various limitations like accumulating errors over time, significantly impacting downstream tasks. We develop a novel method, DT-VAE, that incorporates generalizable domain knowledge, is mathematically justified, and outperforms existing by mitigating error accumulation through cumulative difference learning mechanism. evaluate the performance of DT-VAE on several tasks using...

At the onset of COVID‐19 pandemic, hospitals were in dire need data‐driven analytics to provide support for critical, expensive, and complex decisions. Yet, majority being developed targeted at state‐ national‐level policy decisions, with little availability actionable information tactical operational decision‐making execution hospital level. To fill this gap, we a multi‐method framework leveraging parsimonious design philosophy that allows rapid deployment high‐impact predictive...

Introduction When the hospital census is high, perioperative medical directors or operating room (OR) managers may need to consider postponing some surgical cases scheduled be performed within next three workdays. This scenario has arisen at hospitals in regions with large increases admissions due coronavirus disease 2019 (COVID-19). We compare summary measures for length of stay (LOS) guide OR manager having decide which postponed ensure a sufficient reserve available inpatient beds....

This paper introduces the Delta Coverage (DC) analytics program, an innovative solution to address nursing shortage crisis. The program designs a new flexible role supported by accompanying model for dynamic nurse staffing. Using advanced data analytics, DC dynamically allocates nurses across multiple hospitals in response geographical and temporal demand variability. suite integrates nurse-demand forecast using deep generative stochastic optimization optimal on-call deployment decisions. A...

When patients leave the hospital for lower levels of care, they experience a risk adverse events on daily basis. The advent value-based purchasing among other major initiatives has led to an increasing emphasis reducing occurrences these post-discharge events. This spurred development new prediction technologies identify which are at event as well actions mitigate those risks. Those include pre-discharge and interventions reduce risk. However, traditional models have been developed support...

The gap between medical research on diagnostic testing and clinical workflow can lead to rejection of valuable in a busy environment due increased workloads, or the laboratory that may be practice misunderstanding system‐level benefits new test. This has implications for organizations, test manufacturers, hospital managers among others. To bridge this gap, we develop Markov decision process (MDP) from which create “adoption regions” specify combination characteristics must achieve feasible...

Inpatient ward beds play a central role in hospital operations. To better facilitate coordination of care, the are usually grouped into different specialized units, with each unit designated to serve patients certain primary specialties. However, inpatient wards often associated high level bed utilization and large variability demand. When waiting time is excessively long before becomes available, patient may be assigned non-primary ward. This referred as off-service placement. In this...

Abstract In this article, we analyze a discrete‐time queue that is motivated from studying hospital inpatient flow management, where the customer count process captures midnight census. The stationary distribution of has no explicit form and difficult to compute in certain parameter regimes. Using Stein's method framework, identify continuous random variable approximate steady‐state count. corresponds diffusion with state‐dependent coefficients. We characterize error bounds approximation...