Commercial Dataset Research Articles

Neurologist ambulatory care, health care utilization, and costs in a large commercial dataset.

To determine the value of neurologist ambulatory care in chronic neurologic diseases in a large administrative claims dataset detailing costs, adverse events, and health care utilization. The Optum proprietary claims dataset (2010-2012) was examined to describe direct health care costs, as well as specific outcome metrics for a large population of persons with chronic neurologic illnesses. In phase I of the study, we detail neurologist involvement and differences in annualized allowed third--party payments within episode treatment groups (ETGs) for 10 neurologic illnesses. For phase II, we examined health care utilization for ETGs of epilepsy, Parkinson disease (PD), stroke, and multiple sclerosis (MS) with and without neurologist involvement. Reported outcomes were unadjusted differences and odds ratios between treatment groups. For phase I, a total of 1,913,605 ETGs for 10 neurologic conditions were identified, 30.1% meeting criteria for neurologist involvement. All conditions had higher direct costs when neurologists were involved with care, ranging from a 25% increase for Alzheimer dementia to 100% more for MS care. In phase II, fractures, infections, emergent care, and inpatient admission were less with neurologist ambulatory care, while neurologist care was associated with greater utilization of disease-specific treatments (immunotherapies in MS anticoagulation in atrial fibrillation-associated stroke, deep brain stimulation and dopaminergic therapies in PD). Neurologist involvement with care is associated with greater unadjusted allowed payments, but fewer adverse events and less acute care utilization.

What Do Different Commercial Data Sets Tell Us About Private Equity Performance?

This paper examines private equity (both buyout and venture funds) performance around the globe using four data sets from leading commercial sources. For North American funds, our results echo recent research findings: buyout funds have outperformed public equities over long periods of time; in contrast, venture funds saw performance fall after spectacular results for vintages in the 1990s. For funds outside North America, buyout funds show performance similar to those in North America while venture fund performance is weaker than in North America. Venture samples outside North America are, however, relatively small and strong conclusions await further research. The similarity of performance estimates across the data sets strengthens confidence in conclusions about the results of private equity investing.

Variance in age-specific sex composition of Pacific halibut catches, and comparison of statistical and genetic methods for reconstructing sex ratios

Declines in size at age of Pacific halibut Hippoglossus stenolepis, in concert with sexually-dimorphic growth and a constant minimum commercial size limit, have led to the expectation that the sex composition of commercial catches should be increasingly female-biased. Sensitivity analyses suggest that variance in sex composition of landings may be the most influential source of uncertainty affecting current understanding of spawning stock biomass. However, there is no reliable way to determine sex at landing because all halibut are eviscerated at sea. In 2014, a statistical method based on survey data was developed to estimate the probability that fish of any given length at age (LAA) would be female, derived from the fundamental observation that large, young fish are likely female whereas small, old fish have a high probability of being male. Here, we examine variability in age-specific sex composition using at-sea commercial and closed-season survey catches, and compare the accuracy of the survey-based LAA technique to genetic markers for reconstructing the sex composition of catches. Sexing by LAA performed best for summer-collected samples, consistent with the hypothesis that the ability to characterize catches can be influenced by seasonal demographic shifts. Additionally, differences between survey and commercial selectivity that allow fishers to harvest larger fish within cohorts may generate important mismatch between survey and commercial datasets. Length-at-age-based estimates ranged from 4.7% underestimation of female proportion to 12.0% overestimation, with mean error of 5.8±1.5%. Ratios determined by genetics were closer to true sample proportions and displayed less variability; estimation to within <1% of true ratios was limited to genetics. Genetic estimation of female proportions ranged from 4.9% underestimation to 2.5% overestimation, with a mean absolute error of 1.2±1.2%. Males were generally more difficult to assign than females: 6.7% of males and 3.4% of females were incorrectly assigned. Although nuclear microsatellites proved more consistent at partitioning catches by sex, we recommend that SNP assays be developed to allow for rapid, cost-effective, and accurate sex identification.

The Role of Digital Trace Data in Supporting the Collection of Population Statistics – the Case for Smart Metered Electricity Consumption Data

AbstractDebates over the future of the UK's traditional decadal census have led to the exploration of supplementary data sources, which could support the provision of timely and enhanced statistics on population and housing in small areas. This paper reviews the potential value of a number of commercial datasets before focusing on high temporal resolution household electricity load data collected via smart metering. We suggest that such data could provide indicators of household characteristics that could then be aggregated at the census output area level to generate more frequent official small area statistics. These could directly supplement existing census indicators or even enable development of novel small area indicators. The paper explores this potential through preliminary analysis of a ‘smart meter‐like’ dataset, and when set alongside the limited literature to date, the results suggest that aggregated household load profiles may reveal key household and householder characteristics of interest to census users and national statistical organisations. The paper concludes that complete coverage, quasi‐real time reporting, and household level detail of electricity consumption data in particular could support the delivery of population statistics and area‐based social indicators, and we outline a research programme to address these opportunities. © 2015 The Authors. Population, Space and Place published by John Wiley &amp; Sons Ltd.

Examining the impact of Walmart's healthier foods initiative on the nutrient profile of US packaged food purchases

The goal of this study is evaluate whether a healthy foods initiative (HFI) at a major national US food retailer (Walmart) improved the nutrient profile of packaged food and beverage purchases (PFPs). We used a longitudinal, commercial dataset of US household PFPs from Walmart and other chain retailers (OCR) to compare trends in the overall nutrient density and percent volume of key food group in the post-HFI period (2011-2013) to expected trajectories of nutrient profiles based on the pre-HFI (2000-2010) trends (n= 164,315). We controlled for concurrent changes in the social, economic, and food retail environment, and used inverse probability weights to account for potential selectivity of customers into Walmart and OCR stores. While Walmart PFPs showed major declines in energy, sodium, and sugar density from 2000 to 2013, as well as declines in the percent volume of sugary beverages, grain-based desserts, snacks, and candy, these declines were steepest between 2000 to 2007 before the HFI. For most nutrient densities and food groups, and across all race and income subgroups, post-HFI declines were similar to what would be expected based on pre-HFI trends, indicating that the HFI did not have a major impact. Nutrient densities of PFPs from OCR showed only minor declines across the time period. While the nutrient profile of Walmart PFPs improved over time, these trends were already well underway by the time the HFI was initiated, suggesting that the HFI itself did not improve nutrient profile or reducing diet disparities. This work is supported by the Robert Wood Johnson Foundation (Grant 67506, 68793,70017,71837), NIH (R01DK098072) and the Carolina Population Center (R24 HD050924, T32 HD007168).

Fast and global authenticity screening of honey using 1H-NMR profiling

An innovative analytical approach was developed to tackle the most common adulterations and quality deviations in honey. Using proton-NMR profiling coupled to suitable quantification procedures and statistical models, analytical criteria were defined to check the authenticity of both mono- and multi-floral honey. The reference data set used was a worldwide collection of more than 800 honeys, covering most of the economically significant botanical and geographical origins. Typical plant nectar markers can be used to check monofloral honey labeling. Spectral patterns and natural variability were established for multifloral honeys, and marker signals for sugar syrups were identified by statistical comparison with a commercial dataset of ca. 200 honeys. Although the results are qualitative, spiking experiments have confirmed the ability of the method to detect sugar addition down to 10% levels in favorable cases. Within the same NMR experiments, quantification of glucose, fructose, sucrose and 5-HMF (regulated parameters) was performed. Finally markers showing the onset of fermentation are described.

Multivariate exposure modeling of accident risk: Insights from Pay-as-you-drive insurance data

The increasing adoption of in-vehicle data recorders (IVDR) for commercial purposes such as Pay-as-you-drive (PAYD) insurance is generating new opportunities for transportation researchers. An important yet currently underrepresented theme of IVDR-based studies is the relationship between the risk of accident involvement and exposure variables that differentiate various driving conditions. Using an extensive commercial data set, we develop a methodology for the extraction of exposure metrics from location trajectories and estimate a range of multivariate logistic regression models in a case-control study design. We achieve high model fit (Nagelkerke’s R2 0.646, Hosmer–Lemeshow significance 0.848) and gain insights into the non-linear relationship between mileage and accident risk. We validate our results with official accident statistics and outline further research opportunities. We hope this work provides a blueprint supporting a standardized conceptualization of exposure to accident risk in the transportation research community that improves the comparability of future studies on the subject.

Processing and inversion of commercial helicopter time-domain electromagnetic data for environmental assessments and geologic and hydrologic mapping

Helicopter time-domain electromagnetic (HTEM) surveying has historically been used for mineral exploration, but over the past decade it has started to be used in environmental assessments and geologic and hydrologic mapping. Such surveying is a cost-effective means of rapidly acquiring densely spaced data over large regions. At the same time, the quality of HTEM data can suffer from various inaccuracies. We developed an effective strategy for processing and inverting a commercial HTEM data set affected by uncertainties and systematic errors. The delivered data included early time gates contaminated by transmitter currents, noise in late time gates, and amplitude shifts between adjacent flights that appeared as artificial lineations in maps of the data and horizontal slices extracted from inversion models. Multiple processing steps were required to address these issues. Contaminated early time gates and noisy late time gates were semiautomatically identified and eliminated on a record-by-record basis. Timing errors between the transmitter and receiver electronics and inaccuracies in absolute amplitudes were corrected after calibrating selected HTEM data against data simulated from accurate ground-based TEM measurements. After editing and calibration, application of a quasi-3D spatially constrained inversion scheme significantly reduced the artificial lineations. Residual lineations were effectively eliminated after incorporating the transmitter and receiver altitudes and line-to-line amplitude factors in the inversion process. The final inverted model was very different from that generated from the original data provided by the contractor. For example, the average resistivity of the thick surface layer decreased from [Formula: see text] to [Formula: see text], the depths to the layer boundaries were reduced by 15%–23%, and the artificial lineations were practically eliminated. Our processing and inversion strategy is entirely general, such that with minor system-specific modifications it could be applied to any HTEM data set, including those recorded many years ago.

Integrated Multi-Network Modeling Environment for Spectrum Management

We describe a first principles based integrated modeling environment to study urban socio-communication networks which represent not just the physical cellular communication network, but also urban populations carrying digital devices interacting with the cellular network. The modeling environment is designed specifically to understand spectrum demand and dynamic cellular network traffic. One of its key features is its ability to support individual-based models at highly resolved spatial and temporal scales. We have instantiated the modeling environment by developing detailed models of population mobility, device ownership, calling patterns and call network. By composing these models using an appropriate in-built workflow, we obtain an integrated model that represents a dynamic socio-communication network for an entire urban region. In contrast with earlier papers that typically use proprietary data, these models use open source and commercial data sets. The dynamic model represents for a normative day, every individual in an entire region, with detailed demographics, a minute-by-minute schedule of each person's activities, the locations where these activities take place, and calling behavior of every individual. As an illustration of the applicability of the modeling environment, we have developed such a dynamic model for Portland, Oregon comprising of approximately 1.6 million individuals. We highlight the unique features of the models and the modeling environment by describing three realistic case studies.

Patient Adherence to Transdermal Rivastigmine After Switching From Oral Donepezil

To examine patient adherence before and after switching from donepezil to the rivastigmine patch. This retrospective cohort study used the MarketScan Commercial and Medicare data sets (2004 to 2009). Patients with a diagnosis of Alzheimer disease who were new donepezil users and were subsequently switched to the rivastigmine patch were included. The proportion of days covered (PDC) and PDC difference between donepezil and the rivastigmine patch were calculated from the time of initiation to the switch, capped at 1 year after the first respective claim. PDC was calculated as the number of days with drugs available divided by the number of days in the respective follow-up periods. The sample included 772 patients (mean age 77 y; 58% female). The mean time between switching from donepezil to the rivastigmine patch was 579 (SD=317.3) days. The mean PDC for the rivastigmine patch was highest among patients who switched within 3 months (80.4% vs. 90.7%; P=0.04) and within 7 to 9 months (61.3% vs. 71.0%; P=0.05) of initiating donepezil. When adherence was analyzed in increments of 1 year, patients who switched to the rivastigmine patch within the first year of treatment had significantly greater adherence to rivastigmine compared with those who were on donepezil (PDC 69.3% vs. 60.6%; P=0.0004). Switching from donepezil to the rivastigmine patch seems to be associated with increased adherence, especially in patients who switched within the first year of initiating donepezil.

Measurement Error in Studies of the Built Environment: Validating Commercial Data as Objective Measures of Neighborhood Destinations

There is increasing focus on the influence of neighborhood destinations on a variety of health behaviors. Commercial databases, integrated with Geographic Information Systems (GIS), are popular sources of destination information for public health researchers. However, the suitability and accuracy of these data for public health research purposes has been generally unexplored. This study validated the presence and number of a broad range of destination types listed within an Australian-based commercial database (Yellow Pages), thought to be important for encouraging health behaviors, against those identified via field audit. The study was conducted in and around 5 housing developments within the RESIDential Environments project across metropolitan Perth, Western Australia. Overall agreement of the count of destinations listed within the Yellow Pages ranged from 0.29-0.76, depending on the study area, the timing of the data extract and the geocoding methods used. Results also indicated considerable variation between different extracts from the same commercial dataset, and appreciable over- and under-counting of different destination types compared with field audit findings. The choice of database and extraction time and methods, have important implications in the quantification of neighborhood destination mix and robustness of associations with public health behaviors.

Translational Researchers’ Perceptions of Data Management Practices and Data Curation Needs: Findings from a Focus Group in an Academic Health Sciences Library

How translational researchers use data is becoming an important support function for librariesto understand. Libraries’ roles in this increasingly complex area of Web librarianship are often unclearly defined. The authors conducted two focus groups with physicians and researchers at an academic medical center, the UCLA David Geffen School of Medicine, to gain a better understanding of the current state of data use and data management practices among translational researchers. Data management practices varied widely among departments, disciplines, and research centers, but common areas of concern important to Web librarianship included privacy, data access, and data stewardship. The results provided qualitative information about creating, managing, maintaining, and validating data in clinical and research environments, and challenges working with clinical research data in universities. These results inform current practices around data access, data sharing, and privacy of research participant data. Libraries can build on existing Web skills by instructing users how to find open source and commercial data sets, partnering to establish and implement improved data collection and archiving practices, and teaching principles of ethical data use.

Geography of Savings in the German Occupational Pension System

Burger C. Geography of savings in the German occupational pension system, Regional Studies. The German pension reform of 2001 significantly decreased the level of public pensions for future generations, and created instead voluntary supplementary plans. This paper analyses the regional structure of savings amounts in the case of occupational pensions using a unique, commercial dataset containing the records of 286 000 employees joining between the years 2002 and 2009. The analysis applies different location definitions in explaining financial behaviour. The differences in the relationship between metropolitan and non-metropolitan, East and West, and North and South are explained with the geographical nature of decision contexts. Implications are drawn on human behaviour and geography.

Efficient colossal pattern mining in high dimensional datasets

‘Frequent pattern mining’ is considered as an important data mining problem which has been extensively studied over the last decade. There are a large number of algorithms which have been developed for frequent pattern mining on a traditional commercial dataset which usually contains a huge number of transactions besides a small number of items in each transaction. The advent of bioinformatics contributed to the development of new form of datasets – called high dimensional – which are characterized by small number of transactions and large number of items in each transaction. The running time of traditional algorithms increases exponentially with increasing average transaction length, thus these algorithms cannot be suitable for the high dimensional datasets. On the other hand, the mining algorithms on high dimensional datasets create a very large output set as result which includes small and mid-size frequent patterns which do not bear any useful information for scientists. Colossal pattern mining is described as a solution to reduce the amount of output set of mining patterns. Due to ignoring the mining of the small and mid-sized patterns, mining process speed is increased in colossal patterns mining algorithms. Therefore, only very large (colossal) patterns are extracted and mined in this approach. In this paper we represent an efficient vertical bottom up method to conduct mining of frequent colossal patterns in high dimensional datasets. In our algorithm, we use a bit matrix to compress the dataset and make it easy to use in mining process. Our experimental result shows that our algorithm attains very good mining efficiencies on various input datasets. Furthermore, our performance study shows that this algorithm outperforms substantially the best former algorithms.

A robust approach to the share-of-choice product design problem

A critical issue when solving the share-of-choice product design problem is the reliability of the optimal solution in the presence of partworth uncertainty. Existing approaches use point estimates of an individual's partworth utilities as input to the product optimization stage, ignoring within-person variability in estimates. Post-optimality sensitivity analysis is occasionally performed to assess the degree to which a solution is negatively impacted by partworth uncertainty. We propose a robust optimization model that explicitly captures variation in partworth estimates during the optimization process. Using a large, commercial dataset, we benchmark our model's performance against its deterministic counterpart. We also present inferential theory to guide the selection of model parameters controlled by the analyst. Results reveal that the new approach produces robust solutions in the face of measurement error. Out-of-sample coverage for individuals drawn from the target population is significantly higher than corresponding solutions from published methods.

Using Free and Proprietary Data to Compare Shortest-Path Lengths for Effective Pedestrian Routing in Street Networks

Ubiquitous mobile devices, such as smartphones, led to an increased popularity of pedestrian-related routing applications over the past few years. Because pedestrians typically aim to minimize their walking distance, especially in nonrecreational and multimodal trips, pedestrian routing systems will be fully used only if they can find the correct shortest path and thus help to avoid unnecessary detours. The standard equipment of car navigation systems based on the Global Positioning System several years ago led to the availability of accurate street network data for car-based routing applications. However, pedestrian routing applications should consider pedestrian-related network segments besides those used by motorized traffic, including footpaths and pedestrian bridges. The authors of this paper performed a shortest-path analysis of pedestrian routes for cities in Germany and the United States. For a set of 1,000 randomly generated origin–destination pairs, the authors compared the lengths of pedestrian routes that were computed by different freely available network sources, such as OpenStreetMap and TIGER/Line data, and proprietary data sets, such as TomTom, NAVTEQ, and ATKIS. The results showed that freely available data sources such as OpenStreetMap provided a relatively comprehensive option for cities in which commercial pedestrian data sets were not yet available.

Classification bias in commercial business lists for retail food stores in the U.S.

BackgroundAspects of the food environment such as the availability of different types of food stores have recently emerged as key modifiable factors that may contribute to the increased prevalence of obesity. Given that many of these studies have derived their results based on secondary datasets and the relationship of food stores with individual weight outcomes has been reported to vary by store type, it is important to understand the extent to which often-used secondary data correctly classify food stores. We evaluated the classification bias of food stores in Dun & Bradstreet (D&B) and InfoUSA commercial business lists.MethodsWe performed a full census in 274 randomly selected census tracts in the Chicago metropolitan area and collected detailed store attributes inside stores for classification. Store attributes were compared by classification match status and store type. Systematic classification bias by census tract characteristics was assessed in multivariate regression.ResultsD&B had a higher classification match rate than InfoUSA for supermarkets and grocery stores, while InfoUSA was higher for convenience stores. Both lists were more likely to correctly classify large supermarkets, grocery stores, and convenience stores with more cash registers and different types of service counters (supermarkets and grocery stores only). The likelihood of a correct classification match for supermarkets and grocery stores did not vary systemically by tract characteristics whereas convenience stores were more likely to be misclassified in predominately Black tracts.ConclusionResearches can rely on classification of food stores in commercial datasets for supermarkets and grocery stores whereas classifications for convenience and specialty food stores are subject to some systematic bias by neighborhood racial/ethnic composition.

The Street Network Evolution of Crowdsourced Maps: OpenStreetMap in Germany 2007–2011

The OpenStreetMap (OSM) project is a prime example in the field of Volunteered Geographic Information (VGI). Worldwide, several hundred thousand people are currently contributing information to the “free” geodatabase. However, the data contributions show a geographically heterogeneous pattern around the globe. Germany counts as one of the most active countries in OSM; thus, the German street network has undergone an extensive development in recent years. The question that remains is this: How does the street network perform in a relative comparison with a commercial dataset? By means of a variety of studies, we show that the difference between the OSM street network for car navigation in Germany and a comparable proprietary dataset was only 9% in June 2011. The results of our analysis regarding the entire street network showed that OSM even exceeds the information provided by the proprietary dataset by 27%. Further analyses show on what scale errors can be reckoned with in the topology of the street network, and the completeness of turn restrictions and street name information. In addition to the analyses conducted over the past few years, projections have additionally been made about the point in time by which the OSM dataset for Germany can be considered “complete” in relative comparison to a commercial dataset.

P2-17-06: Patterns of Bone Density Evaluation in a Community Population Treated with Aromatase Inhibitors.

Abstract Aromatase inhibitors (AIs) lead to an increased risk of bone loss and fracture. Fracture rates in adjuvant AI studies have ranged from 2.3% to 11%, with higher rates seen in studies with longer duration of AI therapy and longer follow up. The American Society of Clinical Oncology recommends baseline bone density testing upon initiation of AI therapy and repeat testing every 1–2 years while on therapy. There are few data regarding the incorporation of these guidelines into clinical practice. We sought to evaluate patterns of bone density testing in a community-based cohort of breast cancer patients treated with AIs. Methods: We obtained encounter and pharmacy data from the HealthCore Integrated Research Database, a fully integrated commercial payer dataset including patients enrolled in WellPoint insurance plans. We identified 9138 women aged ≥50 years who had at least 2 diagnosis codes for breast cancer between 2001 and 2007 (followed through 2008), and who filled at least 1 prescription for an AI. We identified bone density testing using encounter data and used logistic regression to assess patient demographic and clinical characteristics associated with baseline bone density testing (any test from 6 months before through 6 months after the first prescription for AI). Among 2038 women who continued AI therapy for ≥2 years, we used logistic regression to assess factors associated with receipt of any bone density during the 2-year period. Results: Overall, 41.6% of women underwent baseline bone density testing. Rates of bone density testing increased over time, from 26.6% in 2002 to 44.7% in 2008 (adjusted odds ratio [AOR] for 2008 vs. 2002=2.02, 95% confidence interval [CI]=1.53−2.68). Older women were less likely to undergo baseline bone density testing (AOR for women age ≥69 vs. 51–59=0.64, 95% CI=0.56−0.72). Women taking proton pump inhibitors, which have been linked to bone density loss, were also less likely to undergo baseline bone density testing (AOR=0.86, 95% CI=0.73−1.00). Women living in areas of lower education (p&amp;lt;.001), and areas with a higher proportion of blacks (p=.02) or Hispanics (p&amp;lt;.001), as well as women previously treated with tamoxifen (p=.02), were all also less likely to undergo baseline bone density. Women receiving bisphosphonates in the year before initiation of an AI were more likely to undergo baseline bone density (AOR=1.33, 95% CI=1.16−1.51). Among women on AIs for at least 2 years, 59.9% of women underwent a bone density within 2 years of starting treatment. In adjusted analyses, earlier year of AI initiation (p&amp;lt;.001), living in areas with lower education levels (p&amp;lt;.001), and having been previously treated with tamoxifen (p&amp;lt;.001) were all associated with a decreased likelihood of undergoing bone density testing. Conclusions: Despite the increased risk of fracture in women treated with AIs and guidelines recommending regular bone density evaluation, 58.4% of women starting an AI and 40.1% of women on long-term therapy did not undergo bone density evaluation in this community-based population. Older women, who might be especially vulnerable to bone loss and fracture, were less likely to undergo baseline bone density evaluation. More attention is needed to ensure that preventable fractures are avoided in patients taking AIs. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr P2-17-06.

Options for monitoring and estimating historical carbon emissions from forest degradation in the context of REDD+.

Measuring forest degradation and related forest carbon stock changes is more challenging than measuring deforestation since degradation implies changes in the structure of the forest and does not entail a change in land use, making it less easily detectable through remote sensing. Although we anticipate the use of the IPCC guidance under the United Framework Convention on Climate Change (UNFCCC), there is no one single method for monitoring forest degradation for the case of REDD+ policy. In this review paper we highlight that the choice depends upon a number of factors including the type of degradation, available historical data, capacities and resources, and the potentials and limitations of various measurement and monitoring approaches. Current degradation rates can be measured through field data (i.e. multi-date national forest inventories and permanent sample plot data, commercial forestry data sets, proxy data from domestic markets) and/or remote sensing data (i.e. direct mapping of canopy and forest structural changes or indirect mapping through modelling approaches), with the combination of techniques providing the best options. Developing countries frequently lack consistent historical field data for assessing past forest degradation, and so must rely more on remote sensing approaches mixed with current field assessments of carbon stock changes. Historical degradation estimates will have larger uncertainties as it will be difficult to determine their accuracy. However improving monitoring capacities for systematic forest degradation estimates today will help reduce uncertainties even for historical estimates.