Contemporary Datasets Research Articles

Internal mammary node metastasis in breast cancer: Predictive models to determine status & management algorithms

Aim Internal mammary node (IMN) metastases are an important prognostic factor in breast cancer. However due to difficulty of access, most surgeons ignore these nodes, hence adjuvant treatment decisions may be compromised. Through mathematical modeling based on large datasets this study aims to estimate the current rate of IMN and sentinel node metastasis. Methods Models were created to estimate the current rate of axillary and IM sentinel node metastasis. Data from historical extended radical mastectomy series were analyzed to project contemporary rates of IMN metastasis. This information was coupled with derived models and contemporary datasets: a single-institution breast lymphoscintigraphy database (1992–2007) to establish lymphatic anatomy; and the Surveillance, Epidemiology and End-Results (SEER) registries in the US (2000–2003). Results Rates of IMN metastasis and positive sentinel nodes were estimated and models derived to assist with predicting IMN status in patients. If high definition peritumoral lymphatic mapping were available, the predicted rates of positive sentinel nodes in the axilla (AN) and internal mammary chain (IMN) would be equal. We predicted the overall rate of IMN metastasis is ∼39% the rate of positive sentinel AN. Conclusion Simplified models and algorithms can predict IMN status.

Impact of Selection Bias on the Utilization of Adjuvant Therapy for Pancreas Adenocarcinoma

Improved outcomes have been associated with the use of adjuvant therapy after resection of pancreas adenocarcinoma. However, the frequency with which patients receive adjuvant therapy and the factors impacting its use remain largely undefined. We hypothesized that nonutilization of adjuvant therapy was primarily associated with patient comorbidity and onset of postoperative complications. A prospectively maintained database was reviewed to identify patients who underwent potentially curative resection of histologically confirmed pancreas adenocarcinoma at our institution from January 1996 to May 2007. Clinicopathological data and postoperative treatment history were collected to identify variables associated with receipt of adjuvant therapy. Of 119 patients, 33% did not receive adjuvant therapy. The frequency with which patients underwent adjuvant therapy did not change over time. On multivariate analysis, patient age 70 years or greater, major postoperative complications, distal pancreatectomy, absence of nodal metastases, and absence of perineural invasion were associated with decreased utilization of adjuvant therapy. One-third of patients in this contemporary dataset of patients did not go on to receive adjuvant therapy. The likelihood of receiving adjuvant treatment is negatively impacted by the course of postoperative recovery. Moreover, the fact that adjuvant therapy was undertaken less often for older patients and patients with favorable pathological features highlights the selection bias impacting the decision to pursue postoperative therapy for this disease. This selective utilization of postoperative therapy for patients with adverse oncological characteristics is likely to bias any retrospective analysis attempting to measure the efficacy of adjuvant treatment for pancreas adenocarcinoma.

Universal features of surname distribution in a subsample of a growing population

We examine the problem of family size statistics (the number of individuals carrying the same surname, or the same DNA sequence) in a given size subsample of an exponentially growing population. We approach the problem from two directions. In the first, we construct the family size distribution for the subsample from the stable distribution for the full population. This latter distribution is calculated for an arbitrary growth process in the limit of slow growth, and is seen to depend only on the average and variance of the number of children per individual, as well as the mutation rate. The distribution for the subsample is shifted left with respect to the original distribution, tending to eliminate the part of the original distribution reflecting the small families, and thus increasing the mean family size. From the subsample distribution, various bulk quantities such as the average family size and the percentage of singleton families are calculated. In the second approach, we study the past time development of these bulk quantities, deriving the statistics of the genealogical tree of the subsample. This approach reproduces that of the first when the current statistics of the subsample is considered. Surname statistics for the US in 1790 and 2000 and for Norway in 2008 are analyzed in the light of the theory and show satisfactory agreement, when the time-dependence of the growth rate is taken into account for the two contemporary data sets.

On Consistency and Sparsity for Principal Components Analysis in High Dimensions

Principal components analysis (PCA) is a classic method for the reduction of dimensionality of data in the form of n observations (or cases) of a vector with p variables. Contemporary datasets often have p comparable with or even much larger than n. Our main assertions, in such settings, are (a) that some initial reduction in dimensionality is desirable before applying any PCA-type search for principal modes, and (b) the initial reduction in dimensionality is best achieved by working in a basis in which the signals have a sparse representation. We describe a simple asymptotic model in which the estimate of the leading principal component vector via standard PCA is consistent if and only if p(n)/n → 0. We provide a simple algorithm for selecting a subset of coordinates with largest sample variances, and show that if PCA is done on the selected subset, then consistency is recovered, even if p(n) ≫ n.

A Simulation Study on the Probability of Correct Selection for Large k Populations

An increasing number of contemporary datasets are high dimensional. Applications require these datasets be screened (or filtered) to select a subset for further study. Multiple testing is the standard tool in such applications, although alternatives have begun to be explored. In order to assess the quality of selection in these high-dimensional contexts, Cui and Wilson (2008b) proposed two viable methods of calculating the probability that any such selection is correct (PCS). PCS thereby serves as a measure of the quality of competing statistics used for selection. The first simulation study of this article investigates the two PCS statistics of the above article. It shows that in the high-dimensional case PCS can be accurately estimated and is robust under certain conditions. The second simulation study investigates a nonparametric estimator of PCS.

Assessing historical versus contemporary mercury and lead contamination in Lake Huron sediments

This research utilized surficial sediment core sample data that were collected in 1969/1973 and 2002 from Lake Huron as part of the Environment Canada Great Lakes Sediment Assessment Program. Concentrations for mercury and lead were analyzed due their persistence in the lake ecosystem and their detrimental environmental effects. The analysis area included the main basin of Lake Huron, Georgian Bay, and the North Channel. Comprehending overall pollution levels strictly on the basis of point data is a difficult task, however spatial analysis techniques combined with Geographic Information Systems can be used to gain a better understanding of lake-wide trends. The Geostatistical Analyst extension of the ESRI ArcGIS software was used to carry out ordinary kriging analyses on the datasets. They produced statistically valid concentration estimates with log-normal data transformation procedures occasionally being performed to obtain suitable prediction estimates. Geospatial analysis (including kriging) allows for samples that vary in number and location to be analyzed and compared with each other based on areal estimates. Overall decreases in contamination levels were observed between the historical and contemporary surveys. Mercury has seen a dramatic reduction in concentrations from 1969/1973 to 2002, while the lead results indicate that high levels of contamination (compared to background concentrations) still persist in the contemporary dataset, although they have subsided from historic values. Higher contaminant concentrations were generally found in depositional basins. The interpolated kriging surfaces are more informative than i.e. conventional dot and/or proportional circle maps in the amount of information they present. They also provide an increased understanding of both the spatial distribution and temporal trends in sediment contamination in Lake Huron.

Evaluation of connectedness in the registered population of the Sardinian dairy sheep breed

Connectedness across flocks of the registered population of the Sardinian dairy sheep breed was evaluated Used criteria were :number of sires used in the same contemporary group; number of direct links and the average within and between contemporary group relationship coefficients Dataset including yearlings’ first lactation records of 1990 and 2008 were compared to evaluate the effect of coupling artificial insemination with controlled natural mating on connectedness. An increase of number of sires and direct links were registered between 1990 and 2008. These results confirm that the across years prolonged use of well planned moderate rates of artificial insemination had a strong effect on the overall statistical connectedness across flocks so permitting more accurate estimations of the management and genetic effects As a whole, the decreasing of the average relationship within flock coupled to a marked enhance of the average relationship between contemporary groups, indicates that the genetic connectedness of the Sardinian registered population presents a clearly increasing trend. This fact assures that current genetic evaluations are reliable across flocks.

Sharpening the Toolbox of Computational Chemistry: A New Approximation of Critical F-Values for Multiple Linear Regression

Multiple linear regression is a major tool in computational chemistry. Although it has been used for more than 30 years, it has only recently been noted within the cheminformatics community that the standard F-values used to assess the significance of the resulting models are inappropriate in situations where the variables included in a model are chosen from a large pool of descriptors, due to an effect known in the statistical literature as selection bias. We have used Monte Carlo simulations to estimate the critical F-values for many combinations of sample size (n), model size (p), and descriptor pool size (k), using stepwise regression, one of the methods most commonly used to derive linear models from large sets of molecular descriptors. The values of n, p, and k represent cases appropriate to contemporary cheminformatics data sets. A formula for general n, p, and k values has been developed from the numerical estimates that approximates the critical stepwise F-values at 90%, 95%, and 99% significance levels. This approximation reproduces both the original simulated values and an interpolation test set (within the range of the training values) with an R2 value greater than 0.995. For an extrapolation test set of cases outside the range of the training set, the approximation produced an R2 above 0.93.

Differences in Treatment and Management of Indigenous and Non-Indigenous Patients Presenting with Chest Pain: Results of the Heart Protection Partnership (HPP) Study

Indigenous patients with acute coronary syndromes represent a high-risk group. There are however few contemporary datasets addressing differences in the presentation and management of Indigenous and non-Indigenous patients with chest pain. The Heart Protection Project, is a multicentre retrospective audit of consecutive medical records from patients presenting with chest pain. Patients were identified as Indigenous or non-Indigenous, and time to presentation and cardiac investigations as well as rates of cardiac investigations and procedures were compared between the two groups. Of the 2380 patients included, 199 (8.4%) identified as Indigenous, and 2174 (91.6%) as non-Indigenous. Indigenous patients were younger, had higher rates hyperlipidaemia, diabetes, smoking, known coronary artery disease and a lower rate of prior PCI; and were significantly less likely to have private health insurance, be admitted to an interventional facility or to have a cardiologist as primary physician. Following adjustment for difference in baseline characteristics, Indigenous patients had comparable rates of cardiac investigations and delay times to presentation and investigations. Although the Indigenous population was identified as a high-risk group, in this analysis of selected Australian hospitals there were no significant differences in treatment or management of Indigenous patients in comparison to non-Indigenous.

Interannual variations in fire weather, fire extent, and synoptic-scale circulation patterns in northern California and Oregon

The Mediterranean climate region on the west coast of the United States is characterized by wet winters and dry summers, and by high fire activity. The importance of synoptic-scale circulation patterns (ENSO, PDO, PNA) on fire-climate interactions is evident in contemporary fire data sets and in pre-Euroamerican tree-ring-based fire records. We investigated how interannual variability in two fire weather indices, the Haines index (HI) and the Energy Release Component (ERC), in the Mediterranean region of southern Oregon and northern California is related to atmospheric circulation and fire extent. Years with high and low fire weather index values corresponded to years with a high and low annual area burned, respectively. HI combines atmospheric moisture with atmospheric instability and variation in HI was more strongly associated with interannual variation in wildfire extent than ERC, which is based on moisture alone. The association between fire extent and HI was also higher for fires in southern Oregon than in northern California. In terms of synoptic-scale circulation patterns, years of high fire risk (i.e., increased potential for erratic fire behavior, represented by HI and ERC) were associated with positive winter PNA and PDO conditions, characterized by enhanced regional mid-tropospheric ridging and low atmospheric moisture. The time lag we found between fire risk potential and prior winter circulation patterns could contribute to the development of long-lead fire-climate forecasting.

Management Theory and Occam's Razor

Because organizations are open systems, environmental forces can both benefit the organization and cause it harm. The empirical theory of O'Toole and Meier suggests buffering is related to organizational performance in nonlinear ways by interacting with environmental forces, managerial networking, and organizational structures. This article examines the organization's history to generate an estimate of environmental buffering and then incorporates this measure to study buffering in a more contemporary data set. The authors find buffering is related to a variety of performance indicators and has an especially strong influence on performance in regard to a more disadvantaged clientele. The article then assesses the functional form of the buffering relationship and the potential interaction with other variables. Although the logic of Occam's razor implies that the simple linear relationship best fits the data, significant research needs to be done on other measures of buffering and other data sets before definitive conclusions can be drawn.

Predicting the prognosis of ciliochoroidal uveal melanoma by its clinical characteristics: a European ophthalmic oncology group collaborative study

Abstract Purpose: To assess clinical characteristics independently associated with mortality from uveal melanoma using a large contemporary European dataset in order to improve future classifications. Methods: Data of 8340 patients who had a choroidal and ciliary body melanoma managed by one of five participating ocular oncology centers directed by members of the European Ophthalmic Oncology Group (OOG) were collated. Variables requested included date of treatment and last follow‐up, survival status, largest basal tumour diameter (LBD), tumour thickness, ciliary body involvement and extraocular extension. Data were analysed with the product‐limit method and Cox regression. Results: Valid LBD and thickness was available for for 7922 and 8111 melanomas, and ciliary and extraocular extension for 7418 and 6087 tumours, respectively. Complete data for all 4 variables were available for 5469 tumours (66%). Mean follow‐up time was 4.0 years (interquartile range, 1.5‐7.3). In Cox regression, all variables were significantly (P<0.0001) and independently associated with melanoma‐related mortality. Covariate‐adjusted hazard ratio (HR) was 1.15 for LBD and 1.06 for thickness (each 1 mm increase), and 1.72 for ciliary body involvement and 2.29 for presence of extraocular extension (yes or no). Conclusions: Choroidal and ciliary body melanomas may effectively be categorised by LBD and thickness, and subcategorized by presence of ciliary body and extraocular extension. All four variables are recommended to be included in future clinical classifications of uveal melanoma. How ciliary body and extraocular extension could be collaboratively coded in more detail remains to be discussed.

Pathology of naturally occurring bovine tuberculosis in England and Wales

The aim of this study was to obtain a contemporary data set of pathology in tuberculin reactor and in-contact cattle in England and Wales. Four hundred animals (200 reactors and 200 in-contacts) from 242 farms located in 14 counties in Western England and Wales were examined. The mean number of lymph nodes (LNs) with tuberculosis (TB)-like lesions per TB-confirmed animal was 1.7 in reactors and 1.5 in in-contact animals. Tuberculous lesions in both reactor and in-contact animals were most commonly observed in the LNs of the thorax, followed by the head and abdomen, particularly the mediastinal, retropharyngeal and tracheobronchial LNs. Twenty-five reactors had macroscopic lesions in the palatine tonsils. Among TB-confirmed cattle, 27% of reactors and 9% of in-contact animals had gross TB-like lesions in the lungs, particularly in the caudal lobes. Gross lesions that were not TB-confirmed were parasitic granulomas (45%), bacterial or mycotic club-forming pyogranulomas (27%) and bacterial abscesses (23%). Diagnostic sensitivity was maximised when bacteriology and histopathology were used concurrently. Stage IV granulomas, alone or in combination with other stages, constituted 63% of lesions, while 16% of lesions were stage I/II granulomas. Caseous necrosis and calcification were common features of the granulomas encountered in natural Mycobacterium bovis infections, even with pathology limited to a small number of sites. Granulomas often covered large areas of histological sections and typically contained only small numbers of acid fast bacilli.

Case Volume and Mortality in Pediatric Cardiac Surgery Patients in California, 1998–2003

Previous reports have found an inverse relationship between pediatric cardiac surgery case volume and in-hospital mortality. This association has been noted recently to be decreasing for coronary artery bypass grafting, possibly because of improved training programs, quality improvement activities, or other innovations to improve outcomes. It is unknown whether the volume-mortality association among pediatric cardiac surgery patients is decreasing similarly. We used data from the state of California's patient discharge data set from the years 1998-2003 to replicate 4 previous research studies of pediatric cardiac surgery volume and mortality. The total number of pediatric surgeries varied from 12,801 to 13,971 depending on the selection criteria applied. Using this larger and more contemporary data set, we found a weaker and less consistent volume-mortality relationship than had been reported previously. We also developed a new model, which incorporated elements of the old models, and found a statistically significant relationship with higher volume and lower mortality (odds ratio=0.86 per 100-patient increase in annual volume; 95% CI, 0.81 to 0.92). Post hoc analyses show that this relationship was related to the performance of the single largest-volume hospital. With the use of data from California, the volume-mortality relationship among pediatric cardiac surgery patients has changed since previous research, such that the old models no longer describe a clear or consistent association. With the use of a continuous definition of volume and an updated model, an association is observed but is dependent on highly leveraged covariate patterns found in the largest-volume hospital.

Effects of recent population bottlenecks on reconstructing the demographic history of prairie‐chickens

Current methods of DNA sequence analysis attempt to reconstruct historical patterns of population structure and growth from contemporary samples. However, these techniques may be influenced by recent population bottlenecks, which have the potential to eliminate lineages that reveal past changes in demography. One way to examine the performance of these demographic methods is to compare samples from populations before and after recent bottlenecks. We compared estimates of demographic history from populations of greater prairie-chickens (Tympanuchus cupido) before and after recent bottlenecks using four common methods (nested clade analysis [NCA], Tajima's D, mismatch distribution, and MDIV). We found that NCA did not perform well in the presence of bottleneck events, although it did recover some genetic signals associated with increased isolation and the extinction of intermediate populations. The majority of estimates for Tajima's D, including those from bottlenecked populations, were not significantly different from zero, suggesting our data conformed to neutral expectations. In contrast, mismatch distributions including the raggedness index were more likely to identify recently bottlenecked populations with this data set. Estimates of population mutation rate (theta), population divergence time (t), and time to the most recent common ancestor (TMRCA) from MDIV were similar before and after bottlenecks; however, estimates of gene flow (M) were significantly lower in a few cases following a bottleneck. These results suggest that caution should be used when assessing demographic history from contemporary data sets, as recently fragmented and bottlenecked populations may have lost lineages that affect inferences of their demographic history.

Factors affecting birth weight in sheep: maternal environment

Knowledge of factors affecting variation in birth weight is especially important given the relationship of birth weight to neonatal and adult health. The present study utilises two large contemporary datasets in sheep of differing breeds to explore factors that influence weight at term. For dataset one (Study 1; n=154 Blue-faced Leicester x Swaledale (Mule) and 87 Welsh Mountain ewes, 315 separate cases of birth weight), lamb birth weight as the outcome measure was related to maternal characteristics and individual energy intake of the ewe during specified periods of gestation, i.e. early (1-30 days; term ~147 days gestation), mid (31-80 days) or late (110-147 days) pregnancy. For dataset two (Study 2; n=856 Mule ewes and 5821 cases of birth weight), we investigated using multilevel modelling the influence of ewe weight, parity, barrenness, lamb sex, litter size, lamb mortality and year of birth on lamb birth weight. For a subset of these ewes (n=283), the effect of the ewes' own birth weight was also examined. Interactions between combinations of variables were selectively investigated. Litter size, as expected, had the single greatest influence on birth weight with other significant effects being year of birth, maternal birth weight, maternal nutrition, sex of the lamb, ewe barrenness and maternal body composition at mating. The results of the present study have practical implications not only for sheep husbandry but also for the increased knowledge of factors that significantly influence variation in birth weight; as birth weight itself has become a significant predictor of later health outcomes.

Assessing the acidity of Swedish streams using benthic macroinvertebrates and weighted averaging (WA) regression and calibration

Weighted averaging (WA) regression and calibration methods are commonly used in paleoecology to infer environmental conditions in the past, but few stream studies have used these approaches on contemporary macroinvertebrate data sets. Here we examined if WA regression and calibration methods could better our understanding of the effects of acidification on Swedish stream ecosystems. Macroinvertebrate and water chemistry data from 232 stream sites in the northern and 154 stream sites in the southern parts of the country were used for model calibration. In addition, 42 randomly selected sites were used for model validation (25 in the north and 17 in the south). Inferred WA pH was compared to inferences of pH obtained using four commonly used acidity indices. Performance of the different methods was assessed using the r 2 -values and root mean square errors (RMSE) of regressions between inferred and observed pH. WA approaches were found to be slightly better than the four acidity indices studied here. However, both WA and the acidity indices performed poorly at the test sites in southern Sweden. This study shows that WA approaches could be useful for assessing acidity of Swedish streams, but the overall low r 2 -values indicate that factors other than pH also are affecting the macroinvertebrate assemblages confounding the pH-organism response relationship.

Descriptive statistics for interval-valued observations in the presence of rules

While symbolic data exist in their own right, contemporary datasets can be too large to analyse using traditional statistical methodologies. Aggregation of these large datasets into sets of more managable size perforce produce datasets whose entries are symbolic data. This paper studies the derivation of basic description statistics, in particular, histograms and mean and variances plus joint histograms for interval-valued datasets when logical dependency rules are present. Algorithms for calculating these histograms are also provided.

Detecting climate change induced range shifts: Where and how should we be looking?

Abstract: Global climate warming is expected to cause systematic shifts in the distribution of species and consequently increase extinction risk. Conservation managers must be able to detect, measure and accurately predict range shifts in order to mitigate impacts on biodiversity. However, important responses to climate change may go unnoticed or be dismissed if we fail to collect sufficient baseline data and apply the most sensitive analytical tests. Here we use randomizations of a contemporary data set on rainforest birds of north‐eastern Australia to quantify the sensitivity of three measures for assessing range shifts along altitudinal gradients. We find that smaller range shifts are detectable by analysing change in the mean altitude of presence records rather than upper or lower range boundaries. For a moderate survey effort of 96 surveys, measurements of change in the mean altitude of 34 species have the capacity to provide strong inference for a mean altitudinal range shift as small as 40 m across the species assemblage. We also show that range shifts measured at range boundaries can be potentially misleading when differences in sampling effort between contemporary and historical data sets are not taken into account.

Quantitative evaluation of automated skull‐stripping methods applied to contemporary and legacy images: Effects of diagnosis, bias correction, and slice location

Performance of automated methods to isolate brain from nonbrain tissues in magnetic resonance (MR) structural images may be influenced by MR signal inhomogeneities, type of MR image set, regional anatomy, and age and diagnosis of subjects studied. The present study compared the performance of four methods: Brain Extraction Tool (BET; Smith [2002]: Hum Brain Mapp 17:143-155); 3dIntracranial (Ward [1999] Milwaukee: Biophysics Research Institute, Medical College of Wisconsin; in AFNI); a Hybrid Watershed algorithm (HWA, Segonne et al. [2004] Neuroimage 22:1060-1075; in FreeSurfer); and Brain Surface Extractor (BSE, Sandor and Leahy [1997] IEEE Trans Med Imag 16:41-54; Shattuck et al. [2001] Neuroimage 13:856-876) to manually stripped images. The methods were applied to uncorrected and bias-corrected datasets; Legacy and Contemporary T1-weighted image sets; and four diagnostic groups (depressed, Alzheimer's, young and elderly control). To provide a criterion for outcome assessment, two experts manually stripped six sagittal sections for each dataset in locations where brain and nonbrain tissue are difficult to distinguish. Methods were compared on Jaccard similarity coefficients, Hausdorff distances, and an Expectation-Maximization algorithm. Methods tended to perform better on contemporary datasets; bias correction did not significantly improve method performance. Mesial sections were most difficult for all methods. Although AD image sets were most difficult to strip, HWA and BSE were more robust across diagnostic groups compared with 3dIntracranial and BET. With respect to specificity, BSE tended to perform best across all groups, whereas HWA was more sensitive than other methods. The results of this study may direct users towards a method appropriate to their T1-weighted datasets and improve the efficiency of processing for large, multisite neuroimaging studies.