Herbicide and degradate flux in the Yazoo River Basin

During 1996-1997, water samples were collected from five sites in the Yazoo River Basin and analysed for 14 herbicides and nine degradates. These included acetochlor, alachlor, atrazine, cyanazine, fluometuron, metolachlor, metribuzin, molinate, norflurazon, prometryn, propanil, propazine, simazine, trifluralin, three degradates of fluometuron, two degradates of atrazine, one degradate of cyanazine, norflurazon, prometryn, and propanil. Fluxes generally were higher in 1997 than in 1996 due to a greater rainfall in 1997 than 1996. Fluxes were much larger from streams in the alluvial plain (an area of very productive farmland) than from the Skuna River in the bluff hills (an area of small farms, pasture, and forest). Adding the flux of the atrazine degradates to the atrazine flux increased the total atrazine flux by an average of 14.5%. The fluometuron degradates added about 10% to the total fluometuron flux, and adding the norflurazon degradate flux to the norflurazon flux increased the flux by 82% in 1996 and by 171% in 1997.     

Model-based clustering of probability density functions

Complex data such as those where each statistical unit under study is described not by a single observation (or vector variable), but by a unit-specific sample of several or even many observations, are becoming more and more popular. Reducing these sample data by summary statistics, like the average or the median, implies that most inherent information (about variability, skewness or multi-modality) gets lost. Full information is preserved only if each unit is described by a whole distribution. This new kind of data, a.k.a. “distribution-valued data”, require the development of adequate statistical methods. This paper presents a method to group a set of probability density functions (pdfs) into homogeneous clusters, provided that the pdfs have to be estimated nonparametrically from the unit-specific data. Since elements belonging to the same cluster are naturally thought of as samples from the same probability model, the idea is to tackle the clustering problem by defining and estimating a proper mixture model on the space of pdfs. The issue of model building is challenging here because of the infinite-dimensionality and the non-Euclidean geometry of the domain space. By adopting a wavelet-based representation for the elements in the space, the task is accomplished by using mixture models for hyper-spherical data. The proposed solution is illustrated through a simulation experiment and on two real data sets.     

Analytical performance of glucometers in a tertiary care hospital

Point-of-care (POC) testing of glucose (glucometers) represents a convenient alternative to monitor glycemia since the measurement procedure is performed without delay after sampling of the capillary blood, thereby avoiding the metabolism by the blood cells of glucose present in plasma. Likely because of sample instability, there is no proficiency test provider in Brazil for this type of POC sample. In this context, this study aimed to evaluate the analytical performance of glucometers used in a tertiary care hospital. The glucometers used were the Accu-Chek Performa^® model from Roche Diagnostics, which use the principle of amperometry. The reference method was the reaction with modified hexokinase/glucose-6-phosphate in a Dimension^® device. The stability evaluation of the control samples showed that it can be performed up to 90 min after the collection of whole blood samples. In the two rounds performed, only one result of the 17 glucometers evaluated was out of the threshold of two standard deviation. Thus, this method for control of glucometers met the expectations and enabled comparing the glucometers in a hospital. Given the current quality guidelines, daily internal quality control of glucometers is recommended, besides at least two annual comparisons between the results of the glucometers and the reference method and one EQA every 3 months.