1,720,985 research outputs found
Characterization of Chronic Kidney Disease Progression in Patients with Diabetes via Group-Based Multi-Trajectory Modeling
No full textPatients suffering from chronic kidney disease (CKD) show a reduction in kidney functionality. Uncontrolled diabetes is among the causes of CKD. As patients with diabetes attend periodic visits, relevant amounts of data end up being available within EHR systems. This information could be exploited to extract insights into disease progression and provide clinicians with tools to better understand the expected disease course. In this work, we applied multi-trajectory group-based trajectory models (GBTM) to identify and characterize groups of patients with similar progression patterns of CKD and diabetes. Specifically, we studied a population of 7,000 patients with diabetes and an initial diagnosis of CKD stage III followed at diabetes outpatient clinics spread across the Veneto Region (Italy). GBTM analysis led to the identification of 6 unique groups of patients with differing CKD and diabetes progression trajectories. Our results suggest that multi-trajectory modeling via GBTM can shed light on the progression of CKD and its interaction with glycemic control, as well as provide clinicians with tools to preemptively identify patients expected to experience significant CKD worsening
Machine learning for healthcare behavioural or: addressing waiting time perceptions in emergency care
Full text linkRecent research has highlighted the need to improve patient satisfaction by reducing perceived waiting times in hospitals. This study examines factors that are associated with waiting time estimation behaviour and how to control flow of patients who overestimate waiting times. Using data from more than 250 patients, we test the applicability of machine learning methods to understand under-, correct and overestimation behaviour of waiting times in two emergency department areas. Our attribute ranking and selection methods reveal that actual waiting time, clinical attributes, and the service environment are among the top ranked and selected attributes. The classification methods reveal that the precision to classify a patient to the true outcome of overestimating waiting times reaches almost 70% in the first waiting area. If a patient waits in a treatment room which is the second waiting area under study, this precision level reaches almost 78%. We developed a discrete-event simulation model which we linked with the machine learning models of each waiting area. Our scenario analysis revealed that changing staffing patterns can lead to a substantial drop-off in the number of patients overestimating waiting times. Our results can be employed to control waiting time perceptions and, potentially, increase patient satisfaction
Flexible hospital-wide elective patient scheduling
Full text linkIn this paper, we build on and extend Gartner and Kolisch (2014)’s hospital-wide patient scheduling problem. Their contribution margin maximizing model decides on the patients' discharge date and therefore the length of stay. Decisions such as the allocation of scarce hospital resources along the clinical pathways are taken. Our extensions which are modeled as a mathematical program include admission decisions and flexible patient-to-specialty assignments to account for multi-morbid patients. Another flexibility extension is that one out of multiple surgical teams can be assigned to each patient. Furthermore, we consider overtime availability of human resources such as residents and nurses. Finally, we include these extensions in the rolling-horizon approach and account for lognormal distributed recovery times and remaining resource capacity for elective patients. Our computational study on real-world instances reveals that, if overtime flexibility is allowed, up to 5% increase in contribution margin can be achieved by reducing length of stay by up to 30%. At the same time, allowing for overtime can reduce waiting times by up to 33%. Our model can be applied in and generalized towards other patient scheduling problems, for example in cancer care where patients may follow defined cancer pathways
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Mathematical Modelling and Cluster Analysis in Healthcare Analytics - The Case of Length of Stay Management
Full text linkLength of Stay (LOS) is an important metric of care quality and efficiency in hospitals that has been studied for decades. Longer stays lead to increased costs and higher burdens on patients, caregivers, clinicians and facilities. Understanding characteristics of LOS outliers is important for developing actionable steps to address LOS reduction. Our study examines clustering of inpatients using key clinical and demographic attributes to identify LOS outliers and investigates the opportunity to reduce their LOS by comparing order sequences with similar non-outliers in the same cluster. Learning from retrospective data, we develop a mathematical model and a two-stage heuristic algorithm. Results indicate that switching orders in homogeneous inpatient sub-populations within the limits of clinical guidelines may be a promising decision support strategy for LOS management. These novel data-driven insights can be offered as suggestions for clinicians to apply new evidence-based, clinical guideline-compliant opportunities for LOS reduction through healthcare analytics
Mathematical modelling and cluster analysis in healthcare analytics - the case of length of stay management
No full textE-HOSPITAL – A digital workbench for hospital operations and services planning using information technology and algebraic languages
Full text linkIn this paper, we describe the development of a unified framework and a digital workbench for the strategic, tactical and operational hospital management plan driven by information technology and analytics. The workbench can be used not only by multiple stakeholders in the healthcare delivery setting, but also for pedagogical purposes on topics such as healthcare analytics, services management, and information systems. This tool combines the three classical hierarchical decision-making levels in one integrated environment. At each level, several decision problems can be chosen. Extensions of mathematical models from the literature are presented and incorporated into the digital platform. In a case study using real-world data, we demonstrate how we used the workbench to inform strategic capacity planning decisions in a multi-hospital, multi-stakeholder setting in the United Kingdom
- …
