Search

In this paper we study the impact of competition on hospital adverse health outcomes, using data on patients admitted to hospitals located in the Lombardy region in Italy between 2004 and 2013. We propose an economic framework that incorporates both short and long range forms of competition among hospitals. In a set up where prices are regulated, and under the assumption that hospitals are profit maximisers, hospital managers compete locally in quality to attract more patients. At the same time, managers have an incentive to compete with all other hospitals within the Lombardy region as their relative quality performance will potentially affect their future states. Our empirical model exploits methods from the graphical modelling literature to estimate local rivals, as well as the degree of local and global interdependence among hospitals. Our results show a significant positive degree of short and long range dependence, which suggests the existence of forms of local and global competition among hospitals with relevant implications for the healthcare policy.

© 2019 The Authors. Despite the recognized importance of interfirm financial links in determining a company's performance, only a few studies have incorporated proxies for interfirm links in credit risk models, and none of these use real financial transactions. We estimate a credit risk model for small and medium-sized enterprises, augmented with information on observed interfirm financial transactions. We exploit a novel data set on about 60000 companies based in the UK and their financial transactions over the years 2015 and 2016. We develop several network-augmented credit risk models and compare their prediction performance with that of a conventional credit risk model that includes only a set of financial ratios. We find that augmenting a default risk model with information on the transaction network makes a significant contribution to increasing the default prediction power of risk models built specifically for small and medium-sized enterprises. Our results may help bankers and credit scoring agencies to improve the credit scoring of these companies, ultimately reducing their propensity to apply excessive lending restrictions. Engineering and Physical Sciences Research Council (grant EP/L021250/1).

In health care, multilevel models are typically used to evaluate hospitals' performance and to rank hospitals accordingly. While multilevel models capture the hierarchical structure in the data, such as the grouping of patients into hospitals, these models do not account for additional latent structures. In this paper, we develop a novel multilevel logistic cluster-weighted model which can predict a binary outcome, such as mortality within 30 days of discharge, while accounting both for known and latent structures of the data. We develop an Expectation-Maximization algorithm for parameter estimation and a parametric bootstrap approach for assessing the variability of the estimators. Using a rich data set of the Lombardy (Italy) health care system and focussing on the two wards of cardiosurgery and medicine, we show how the proposed model detects, in both cases, two well-defined clusters within the patient to hospital hierarchical structure of the data. A comparison with standard multilevel and cluster-weighted approaches reveals a better fit of the proposed model and a greater insight into the structure of the data. We show how this can have implications in the resulting league tables and thus how the proposed model can be a useful tool for policy-makers and healthcare managers to conduct hospital evaluations.

Pay-for-performance approaches have been widely adopted in order to drive improvements in the quality of healthcare provision. Previous studies evaluating the impact of these programs are either limited by the number of health outcomes or of medical conditions considered. In this paper, we evaluate the effectiveness of a pay-for-performance program on the basis of five health outcomes and across a wide range of medical conditions. The context of the study is the Lombardy region in Italy, where a rewarding program was introduced in 2012. The policy evaluation is based on a difference-in-differences approach. The model includes multiple dependent outcomes, that allow quantifying the joint effect of the program, and random effects, that account for the heterogeneity of the data at the ward and hospital level. Our results show that the policy had a positive effect on the hospitals' performance in terms of those outcomes that can be more influenced by a managerial activity, namely the number of readmissions, transfers and returns to the surgery room. No significant changes which can be related to the pay-for-performance introduction are observed for the number of voluntary discharges and for mortality. Finally, our study shows evidence that the medical wards have reacted more strongly to the pay-for-performance program than the surgical ones, whereas only limited evidence is found in support of a different policy reaction across different types of hospital ownership., Comment: 16 pages, 7 figures

This chapter reviews graphical modeling techniques for estimating large covariance matrices and their inverse. The chapter provides a selective survey of different models and estimators proposed by the graphical modeling literature and offers some practical examples where these methods could be applied in the area of health economics.

This is the post-print version of the final paper published in Computational Statistics & Data Analysis. The published article is available from the link below. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. Copyright @ 2013 Elsevier B.V. Networks appear in many fields, from finance to medicine, engineering, biology and social science. They often comprise of a very large number of entities, the nodes, and the interest lies in inferring the interactions between these entities, the edges, from relatively limited data. If the underlying network of interactions is sparse, two main statistical approaches are used to retrieve such a structure: covariance modeling approaches with a penalty constraint that encourages sparsity of the network, and nodewise regression approaches with sparse regression methods applied at each node. In the presence of outliers or departures from normality, robust approaches have been developed which relax the assumption of normality. Robust covariance modeling approaches are reviewed and compared with novel nodewise approaches where robust methods are used at each node. For low-dimensional problems, classical deviance tests are also included and compared with penalized likelihood approaches. Overall, copula approaches are found to perform best: they are comparable to the other methods under an assumption of normality or mild departures from this, but they are superior to the other methods when the assumption of normality is strongly violated.

Veronica Vinciotti and Ernst Wit thank the Isaac Newton Institute for Mathematical Sciences, Cambridge, for its hospitality during the programme Theoretical Foundations of Statistical Network Analysis, which was supported by the EPSRC grant EP/K032208/1.

Factorial Gaussian graphical Models (fGGMs) have recently been proposed for inferring dynamic gene regulatory networks from genomic high-throughput data. In the search for true regulatory relationships amongst the vast space of possible networks, these models allow the imposition of certain restrictions on the dynamic nature of these relationships, such as Markov dependencies of low order – some entries of the precision matrix are a priori zeros – or equal dependency strengths across time lags – some entries of the precision matrix are assumed to be equal. The precision matrix is then estimated by l 1-penalized maximum likelihood, imposing a further constraint on the absolute value of its entries, which results in sparse networks. Selecting the optimal sparsity level is a major challenge for this type of approaches. In this paper, we evaluate the performance of a number of model selection criteria for fGGMs by means of two simulated regulatory networks from realistic biological processes. The analysis reveals a good performance of fGGMs in comparison with other methods for inferring dynamic networks and of the KLCV criterion in particular for model selection. Finally, we present an application on a high-resolution time-course microarray data from the Neisseria meningitidis bacterium, a causative agent of life-threatening infections such as meningitis. The methodology described in this paper is implemented in the R package sglasso, freely available at CRAN, http://CRAN.R-project.org/package=sglasso.

The analysis of microarray data from time-series experiments requires specialised algorithms, which take the temporal ordering of the data into account. In this paper we explore a new architecture of Bayesian classifier that can be used to understand how biological mechanisms differ with respect to time. We show that this classifier improves the classification of microarray data and at the same time ensures that the models can easily be analysed by biologists by incorporating time transparently. In this paper we focus on data that has been generated to explore different types of muscular dystrophy.

Identifying genes that direct the mechanism of a disease from expression data is extremely useful in understanding how that mechanism works. This in turn may lead to better diagnoses and potentially could lead to a cure for that disease. This task becomes extremely challenging when the data are characterised by only a small number of samples and a high number of dimensions, as is often the case with gene expression data. Motivated by this challenge, we present a general framework that focuses on simplicity and data perturbation. These are the keys for robust identification of the most predictive features in such data. Within this framework, we propose a simple selective naive Bayes classifier discovered using a global search technique, and combine it with data perturbation to increase its robustness for small sample sizes. An extensive validation of the method was carried out using two applied datasets from the field of microarrays and a simulated dataset, all confounded by small sample sizes and high dimensionality. The method has been shown to be capable of selecting genes known to be associated with prostate cancer and viral infections.

Supervised classification problems in which the class sizes are very different are common. In such cases, nearest neighbour classifiers exhibit a non-monotonic relationship between the number of nearest neighbours and misclassification rate of each of the two classes separately.

It is generally argued that predictive or decision making steps in statistics are separate from the model building or inferential steps. In many problems, however, predictive accuracy matters more in some parts of the data space than in others, and it is appropriate to aim for greater model effectiveness in those regions. If the relevant parts of the space depend on the use to which the model is to be put, then the best model will depend also on this intended use. We illustrate using examples from supervised classification.

Most of the existing methods for the identification of biologically interesting genes in a temporal expression profiling dataset do not fully exploit the temporal ordering in the dataset and are based on normality assumptions for the gene expression.In this paper, we introduce a Bayesian median regression model to detect genes whose temporal profile is significantly different across a number of biological conditions. The regression model is defined by a polynomial function where both time and condition effects as well as interactions between the two are included. MCMC‐based inference returns the posterior distribution of the polynomial coefficients. From this a simple Bayes factor test is proposed to test for significance.The estimation of the median rather than the mean, and within a Bayesian framework, increases the robustness of the method compared to a Hotelling T2‐test previously suggested. This is shown on simulated data and on muscular dystrophy gene expression data.

PURPOSE. To examine the relationship between an anatomic map relating the retinal nerve fiber layer (RNFL) distribution to the optic nerve head and a functional map derived from the interpoint correlation of raw sensitivities in visual field (VF) testing. METHODS. Previously, interpoint correlations were generated for all possible pairs of VF test points in a dataset of 98,821 Humphrey VF test results taken from the Moorfields Eye Hospital archive. The relationship between these correlations and the physical distance between the VF test point pairs was evaluated by Pearson's correlation coefficient and multiple regression analysis. The distance between the pairs of VF test points was calculated in two ways. First, the anatomic map was used to estimate the angular distance at the optic nerve head (ONH), between the RNFL bundles corresponding to the VF test points in each pair (ONHd). Second, the retinal distance between pairs of test points was calculated from the Humphrey VF template (RETd). A best-fit model for predicting functional correlation (FC) from ONHd and RETd was constructed and used to formulate a filter incorporating the anatomic-functional correlation data. RESULTS. All scatterplots showed a negative association between interpoint retinal sensitivity correlation values and distance between points: ONHd (R 2 = 0.60) and RETd (R 2 = 0.33). The raw sensitivity correlation values could be predicted from a multiple regression model using ONHd, RETd, and a combined interaction of ONHd and RETd (R 2 = 0.75, P < 0.00001). The construction of a new filter was based on the equation FC = 0.9325 - (0.0029 ONHd) - (0.0077 ·RETd) + (0.0001 ·ONHd RETd). CONCLUSIONS. A good level of association was observed between the strength of correlation between points in the VF and the relative location of those test points in the peripheral retina and in corresponding RNFL bundles at the ONH. These results help to validate the relationship between structure and function and may be of use in the further refinement of physiologically derived VF filters to reduce measurement noise.

Background The identification of biologically interesting genes in a temporal expression profiling dataset is challenging and complicated by high levels of experimental noise. Most statistical methods used in the literature do not fully exploit the temporal ordering in the dataset and are not suited to the case where temporal profiles are measured for a number of different biological conditions. We present a statistical test that makes explicit use of the temporal order in the data by fitting polynomial functions to the temporal profile of each gene and for each biological condition. A Hotelling T 2-statistic is derived to detect the genes for which the parameters of these polynomials are significantly different from each other. Results We validate the temporal Hotelling T 2-test on muscular gene expression data from four mouse strains which were profiled at different ages: dystrophin-, beta-sarcoglycan and gamma-sarcoglycan deficient mice, and wild-type mice. The first three are animal models for different muscular dystrophies. Extensive biological validation shows that the method is capable of finding genes with temporal profiles significantly different across the four strains, as well as identifying potential biomarkers for each form of the disease. The added value of the temporal test compared to an identical test which does not make use of temporal ordering is demonstrated via a simulation study, and through confirmation of the expression profiles from selected genes by quantitative PCR experiments. The proposed method maximises the detection of the biologically interesting genes, whilst minimising false detections. Conclusion The temporal Hotelling T 2-test is capable of finding relatively small and robust sets of genes that display different temporal profiles between the conditions of interest. The test is simple, it can be used on gene expression data generated from any experimental design and for any number of conditions, and it allows fast interpretation of the temporal behaviour of genes. The R code is available from V.V. The microarray data have been submitted to GEO under series GSE1574 and GSE3523.

Objective: Progressive loss of the field of vision is characteristic of a number of eye diseases such as glaucoma which is a leading cause of irreversible blindness in the world. Recently, there has been an explosion in the amount of data being stored on patients who suffer from visual deterioration including field test data, retinal image data and patient demographic data. However, there has been relatively little work in modelling the spatial and temporal relationships common to such data. In this paper we introduce a novel method for classifying visual field (VF) data that explicitly models these spatial and temporal relationships. Methodology: We carry out an analysis of our proposed spatio-temporal Bayesian classifier and compare it to a number of classifiers from the machine learning and statistical communities. These are all tested on two datasets of VF and clinical data. We investigate the receiver operating characteristics curves, the resulting network structures and also make use of existing anatomical knowledge of the eye in order to validate the discovered models. Results: Results are very encouraging showing that our classifiers are comparable to existing statistical models whilst also facilitating the understanding of underlying spatial and temporal relationships within VF data. The results reveal the potential of using such models for knowledge discovery within ophthalmic databases, such as networks reflecting the 'nasal step', an early indicator of the onset of glaucoma. Conclusion: The results outlined in this paper pave the way for a substantial program of study involving many other spatial and temporal datasets, including retinal image and clinical data.

International audience; Biodiversity datasets are generally stored in different for-mats. This makes it difficult for biologists to combine and integrate them to retrieve useful information for the purpose of, for example, efficiently classify specimens. In this paper, we present BioKET, a data warehouse which is a consolidation of heterogeneous data sources stored in different formats. For the time being, the scopus of BioKET is botanical. We had, among others things, to list all the existing botanical ontologies and re-late terms in BioKET with terms in these ontologies. We demonstrate the usefulness of such a resource by applying FIST, a combined biclus-tering and conceptual association rule extraction method on a dataset extracted from BioKET to analyze the risk status of plants endemic to Laos. Besides, BioKET may be interfaced with other resources, like GeoCAT, to provide a powerful analysis tool for biodiversity data.