Search

Background Coronavirus disease 2019 (COVID-19)-related short-term mortality is high in dialysis patients, but longer-term outcomes are largely unknown. We therefore assessed patient recovery in a large cohort of dialysis patients 3 months after their COVID-19 diagnosis. Methods We analyzed data on dialysis patients diagnosed with COVID-19 from 1 February 2020 to 31 March 2021 from the European Renal Association COVID-19 Database (ERACODA). The outcomes studied were patient survival, residence and functional and mental health status (estimated by their treating physician) 3 months after COVID-19 diagnosis. Complete follow-up data were available for 854 surviving patients. Patient characteristics associated with recovery were analyzed using logistic regression. Results In 2449 hemodialysis patients (mean ± SD age 67.5 ± 14.4 years, 62% male), survival probabilities at 3 months after COVID-19 diagnosis were 90% for nonhospitalized patients (n = 1087), 73% for patients admitted to the hospital but not to an intensive care unit (ICU) (n = 1165) and 40% for those admitted to an ICU (n = 197). Patient survival hardly decreased between 28 days and 3 months after COVID-19 diagnosis. At 3 months, 87% functioned at their pre-existent functional and 94% at their pre-existent mental level. Only few of the surviving patients were still admitted to the hospital (0.8–6.3%) or a nursing home (∼5%). A higher age and frailty score at presentation and ICU admission were associated with worse functional outcome. Conclusions Mortality between 28 days and 3 months after COVID-19 diagnosis was low and the majority of patients who survived COVID-19 recovered to their pre-existent functional and mental health level at 3 months after diagnosis.

Background Patients on kidney replacement therapy (KRT) are at very high risk of coronavirus disease 2019 (COVID-19). The triage pathway for KRT patients presenting to hospitals with varying severity of COVID-19 illness remains ill-defined. We studied the clinical characteristics of patients at initial and subsequent hospital presentations and the impact on patient outcomes. Methods The European Renal Association COVID-19 Database (ERACODA) was analysed for clinical and laboratory features of 1423 KRT patients with COVID-19 either hospitalized or non-hospitalized at initial triage and those re-presenting a second time. Predictors of outcomes (hospitalization, 28-day mortality) were then determined for all those not hospitalized at initial triage. Results Among 1423 KRT patients with COVID-19 [haemodialysis (HD), n = 1017; transplant, n = 406), 25% (n = 355) were not hospitalized at first presentation due to mild illness (30% HD, 13% transplant). Of the non-hospitalized patients, only 10% (n = 36) re-presented a second time, with a 5-day median interval between the two presentations (interquartile range 2–7 days). Patients who re-presented had worsening respiratory symptoms, a decrease in oxygen saturation (97% versus 90%) and an increase in C-reactive protein (26 versus 73 mg/L) and were older (72 vs 63 years) compared with those who did not return a second time. The 28-day mortality between early admission (at first presentation) and deferred admission (at second presentation) was not significantly different (29% versus 25%; P = 0.6). Older age, prior smoking history, higher clinical frailty score and self-reported shortness of breath at first presentation were identified as risk predictors of mortality when re-presenting after discharge at initial triage. Conclusions This study provides evidence that KRT patients with COVID-19 and mild illness can be managed effectively with supported outpatient care and with vigilance of respiratory symptoms, especially in those with risk factors for poor outcomes. Our findings support a risk-stratified clinical approach to admissions and discharges of KRT patients presenting with COVID-19 to aid clinical triage and optimize resource utilization during the ongoing pandemic.

This pattern guides the design of software components that exchange large amounts of data in order to deliver a certain decision service. The general objective is to minimize the total (i.e. development, maintenance and runtime) cost of the system to be built. After the introduction of a set of basic design scenarios to gather data, the pattern proposes an optimal combination of those scenarios such that the total cost is minimized. ispartof: pages:565-577 ispartof: Proceedings of the 11th European Conference on Pattern Languages of Programs (EuroPLoP 2006) pages:565-577 ispartof: European Conference on Pattern Languages of Programs (EuroPLoP 2006) location:Irsee, Germany date:5 Jul - 9 Jul 2006 status: published

status: published

Poor quality data can have significant consequences for both businesses and society and therefore there is a need for proper data quality management. However, because the causes of poor quality data are extremely diverse and difficult to locate, the variety of problem areas that obstruct proper data quality management is typically scattered throughout an organisation. To make things worse, the roles that those areas play in preventing proper data quality management are determined by contextual factors and are thus different for each organisation. This dissertation aims to provide insight into two problem areas that are particularly relevant for organisations in the financial industry: data quality measurement communication and manual data entry. For the problem area of measurement communication, in a first study, we took the current terminology in data quality measurement communication as a constraint and investigated what concept is typically communicated by one of the most popular terms to communicate data quality measurements (i.e. data accuracy) by performing a literature review. We found that there are inconsistencies in how data accuracy is defined and that there are multiple ways to measure data accuracy. In a second study, we did not take the current terminology as a constraint and constructed guidelines that should help researchers and practitioners to answer the question "given a certain measurement operation, what terms shall we use to communicate its resulting measurements?". In a third study, we focused on one specific use of measurement information: detecting data quality root causes. To facilitate root cause detection, we proposed an approach (which was adapted from previous work) where we applied dimensionality reduction to transform data quality measurements into visualisations that particularly support this purpose. For the problem area of manual data entry, we conducted several studies to investigate how errors in manually entered data can be prevented from the perspective of the individual who enters the data. In a first study, we contrasted two existing theories that can be used to explain why individuals enter data incorrectly: the theory of planned behaviour and the theory of knowing why. In this study, we concluded that one construct of the theory of planned behaviour, instrumental attitude, is similar to the knowing why construct but provided a better explanation. In a second study, we investigated to what extent psychosocial factors, as identified by the theory of planned behaviour, determine the intention of the data producers to enter data correctly and to what extent their intention explains the actual errors in the entered data. The results showed that the degree of intention of the data producers to enter data correctly explains a certain share of the actual errors in the studied data, but that a larger share could be explained by statistical control variables related to actual behavioural control. In addition, the results indicated that the surveyed data producers' intention towards entering high quality data is mainly determined by two factors: their instrumental and experiential attitude. In a third study, we investigated how the actual behavioural control construct could be replaced by another construct that provides more information to practitioners. We found that the task technology fit construct is conceptually equal to the actual behavioural control construct in a manual data entry context and can thus serve as a replacement. Yet, unlike for actual behavioural control, the antecedents of the task technology fit construct have been identified: a high task technology fit depends on the characteristics of the individual, the task and the technology involved in manual data entry. Two empirical studies supported the replacement and demonstrated that a high task-technology fit is relatively more important than the data producers' intention. In summary, this dissertation reports about research that contributed to two problem areas that are particularly relevant for data quality management in the financial industry: data quality measurement communication and manual data entry. By doing so, it recognises and highlights the role of the human factor in data quality management. status: published

The global topic of this dissertation is nicely reflected by its title - 'a matter of time' -, as we explore the different uses of data, i.e. 'matter', that is collected sequentially in time and how this data type can be integrated and leveraged within business contexts. Time series data represents the changes that a given feature goes through over time and can be collected for any target or explanatory variable. Time series have therefore been part of business applications for decades and have been studied extensively in both academic literature and practical use cases. This work includes both sides of the spectrum and tries to combine both theoretical and practical points of view on the subject in each project. This thesis therefore has two focal points: business-oriented research, which aims to integrate business concepts into time series use cases in general, and application-based research, which investigates how time series data can be leveraged in specific case studies. Within these two sections, we define three main research topics: profit-driven forecasting, model automation and benefiting from time series data in applications. In terms of profit-driven forecasting, we introduce an expected profit function that can be used for sales forecasting applications, which was extensively tested on datasets from The Coca-Cola Company. If forecasting models are not able to produce accurate forecasts, this leads to an increase in costs, e.g. overstock costs. We therefore calculate the expected profit of a forecast by combining forecasting accuracy with a profit margin that can be set by a business expert in order to evaluate a model in an entirely different way. The resulting formula is easily adaptable to any situation and interpretable as the potential monetary value of a forecast. We show that this function can be integrated into both the building and evaluation of a forecasting model, and can be used next to traditional evaluation metrics for the final model selection. The expected profit function therefore emphasizes the value of a forecasting model by connecting a well-known practical concept, i.e. profit, with the evaluation of existing techniques, which also facilitates its integration into forecasting processes that are already in place. The second topic of the business-oriented research looks into the automation of forecasting models specifically and analytics in general. Model automation can be useful for two separate reasons in business contexts. First of all, if there are too many analytical models to train, it could be beneficial to find a data-driven way of parametrising these models based on their input. By automating this process, less analytics experts are needed and the models could be shared more easily across different people or departments. In this thesis, we therefore investigate how to combine the expected profit function with automatically selecting parameters and features for sales forecasting applications. We even try to optimize the search for the correct parameters instead of using the brute force approach by turning to evolutionary algorithms. The second use for automated analytics is especially apparent in cases where there is a complete lack of analytical expertise, e.g. in smaller companies. Analytics-as-a-Service (AaaS) offers a potential solution to this problem by offering web-based platforms that could also be used by business users to implement ready-made predictive models. Our research takes a closer look at the potential users of these AaaS tools and concludes that some novice users are able to achieve sufficient results if they meet certain requirements. The final topic of this dissertation is difficult to generalize, as all of its conclusions are inevitably linked to the application of the work in question. The two projects of this section deal with completely different use cases: blood donation forecasting and churn prediction by time series classification. In the first study, we leverage the hierarchical structure of the time series data to generate better forecasts by using a bottom-up approach. Furthermore, we look into the potential of web traffic data to capture the intention of a donor to donate in the near future. In the second project, we turn to time series as a means to deal with dynamic features for a churn prediction application. In a world where a large amount of data is being created and collected every day, we have to become more aware of the dynamic aspect of all of this data. For example, a lot of analytical tasks take user behaviour into account and assume that the most recent data about a person has the highest predictive value. However, we wanted to examine whether capturing a change in this behaviour does not lead to a better performing model in the end. Concretely, we take a closer look at a classification use case that aims to identify churners for a telecommunication company by looking at changes in their call detail records. In short, this dissertation looks into the different uses of time series data in several business applications. This work shows the possibilities for this type of data, which is already present in a lot of businesses, but has not reached its full potential yet. status: published

The rise of information systems, and especially the internet, has given service providers the opportunity to increase the number of services they can offer their customers. E-commerce platforms, such as Amazon and AliExpress, can offer their customers millions of products because they do not need shelf space to present all their products. Entertainment providers, such as Youtube, Spotify and Netflix, have millions of songs and movies from which the user can choose. Also in the labor market, employment services and job-posting sites, such as LinkedIn, can now present hundreds of thousands of vacancies to job seekers. With so many open vacancies, it can be difficult for job seekers to find jobs they are interested in, provided that they even know what they are interested in. The general problem, in which the large amount of available information hinders users in their decision making, is called the information overload problem. Recommender systems are developed as a solution for this information overload problem. They are the set of software tools and techniques that help users find items that are relevant to them, by predicting which items the user will be interested in. This thesis presents three studies on the properties of recommender systems that recommend vacancies to job seekers, or vice versa, recommend job seekers for vacancies. The first study compares two ways in which job recommender systems can capture job seekers' interest in open vacancies: explicit user interest and implicit user interest. This study shows that there is a large discrepancy between the job types people explicitly state to be interested in (explicit interest), compared to the job types people show implicit interest in. It also shows that job recommendations generated using implicit feedback data resulted in a more diverse set of recommendations that better predict job seekers' job interest and received higher expert appreciation, compared to recommendations generated using explicit interest data. The second study in this thesis zooms in on two topics. First, the fact that for job recommendation to be successful, the job seeker has to be interested in the recommended job, but that this interest also needs to be reciprocated by the job provider. Second, the comparison between the reciprocated interest of vacancies that the user consumes via search, and vacancies that the user consumes via recommendation. The key take-aways from this chapter are the following: Vacancies consumed by traditional recommender systems have much lower reciprocity than items consumed via traditional user search. This means that simply plugging in standard unidirectional recommender systems in a reciprocal recommendation context could have a negative impact on the consumer of the recommended items. However, the reciprocity of recommended items can be increased beyond the reciprocity of user search, with a trade-off in predictive power, by varying the rating matrices used by the recommendation algorithms. In the third study, we present the extension of a methodology that we originally developed to visually inspect the distribution of recommendation quality over all users. The methodology allows allows non-technical interpreters to answer the following questions about the relationship between a feature and a complex dataset: Is the feature distributed randomly, or is it systematically biased towards groupings in the data? If there is an observable bias, what features from the dataset drive this bias? Applied to recommender systems, this methodology can be used to study for which job seekers recommender systems work well, and for which they do not. status: published

status: published