Short communication: On recognizing the proper experimental unit in animal studies in the dairy sciences.

Sound design of experiments combined with proper implementation of appropriate statistical methods for data analysis are critical for producing meaningful scientific results that are both replicable and reproducible. This communication addresses specific aspects of design and analysis of experiments relevant to the dairy sciences and, in so doing, responds to recent concerns raised in a letter to the editor of the Journal of Dairy Science regarding journal policy for research publications on pen-based animal studies. We further elaborate on points raised, rectify interpretation of important concepts, and show how aspects of statistical inference and elicitation of research conclusions are affected.

Accessing totally implantable venous access systems on the day of placement does not significantly increase the risk of infection.

PURPOSE: Totally implantable venous access systems (ports) are commonly placed and have a low complication rate. The most common complication is infection, which can have very negative effects on patients resulting in hospitalization and/or treatment delay in the setting of neoplasm. While a number of variables have been studied in relation to diminishing infectious rates, one remaining question is the effect of accessing the port on day of placement, which is the aim of this retrospective study. MATERIALS AND METHODS: After internal review board approval the electronic medical records of 2,006 patients who underwent port placement between 10/1/2008 and 9/30/2013 were reviewed. Of these patients 628 were excluded as they did not have complete placement and removal data available, leaving 1378 patients in our cohort. RESULTS: There was a significantly longer number of infection-free catheter days in the out-patient cohort as compared to the in-patient cohort (p = 0.027). In-patients mean day after placement when the port was first accessed (DAP) (0.5) was statistically earlier (closer to placement) than the out-patients DAP (7.2) (p<0.0001). However, the increased likelihood of infection could not be explained by DAP (p = 0.2029) even when controlling for in-patient and out-patient status (p = 0.97). CONCLUSIONS: Accessing the port on the day of placement does not significantly contribute to an increased likelihood of infection. This study seems to indicate that placing a port on the first day of outpatient therapy likely optimally balances respect for patient time with infection control.

Statistical considerations for a chronic bioassay study: Exposure to Decamethylcyclopentasiloxane (D5) and incidence of uterine endometrial adenocarcinomas in a 2-year inhalation study with Fischer rats.

Decamethylcyclopentasiloxane (D5) is a cyclic siloxane used in the production of industrial and consumer products. Four groups of 60 Fischer-344 female rats were analyzed for uterine endometrial adenocarcinoma (inhalation study with exposure levels in ppm/number of observed cases: 0/0, 10/1, 40/0, and 160/5) by exact regression (logistic, Poisson), the max poly-3 trend test, and a random effects probit model. When comparing the 160 ppm group to controls after 24 months, the incidence of adenocarcinomas was elevated (borderline significant); it was significant when all exposure levels were included. Four sets of (historical) control groups were formed, with varying heterogeneity. The effect of D5 was either significant or borderline significant when comparing all control sets to the 160 ppm group. When considering all exposure groups using any of the analysis methods, a significant effect was observed when the high dose group was included in the analysis; the effect was not significant when the high dose group was not included. The evidence tends to support the conclusion that D5 at the highest dose level (160 ppm) results in an increased incidence of adenocarcinomas. However, it is important to verify any potential effect through a biological investigation.

Identifying Heat Waves in Florida: Considerations of Missing Weather Data.

BACKGROUND: Using current climate models, regional-scale changes for Florida over the next 100 years are predicted to include warming over terrestrial areas and very likely increases in the number of high temperature extremes. No uniform definition of a heat wave exists. Most past research on heat waves has focused on evaluating the aftermath of known heat waves, with minimal consideration of missing exposure information. OBJECTIVES: To identify and discuss methods of handling and imputing missing weather data and how those methods can affect identified periods of extreme heat in Florida. METHODS: In addition to ignoring missing data, temporal, spatial, and spatio-temporal models are described and utilized to impute missing historical weather data from 1973 to 2012 from 43 Florida weather monitors. Calculated thresholds are used to define periods of extreme heat across Florida. RESULTS: Modeling of missing data and imputing missing values can affect the identified periods of extreme heat, through the missing data itself or through the computed thresholds. The differences observed are related to the amount of missingness during June, July, and August, the warmest months of the warm season (April through September). CONCLUSIONS: Missing data considerations are important when defining periods of extreme heat. Spatio-temporal methods are recommended for data imputation. A heat wave definition that incorporates information from all monitors is advised.

Predictors, spatial distribution, and occurrence of woody invasive plants in subtropical urban ecosystems.

We examined the spatial distribution, occurrence, and socioecological predictors of woody invasive plants (WIP) in two subtropical, coastal urban ecosystems: San Juan, Puerto Rico and Miami-Dade, United States. These two cities have similar climates and ecosystems typical of subtropical regions but differ in socioeconomics, topography, and urbanization processes. Using permanent plot data, available forest inventory protocols and statistical analyses of geographic and socioeconomic spatial predictors, we found that landscape level distribution and occurrence of WIPs was not clustered. We also characterized WIP composition and occurrence using logistic models, and found they were strongly related to the proportional area of residential land uses. However, the magnitude and trend of increase depended on median household income and grass cover. In San Juan, WIP occurrence was higher in areas of high residential cover when incomes were low or grass cover was low, whereas the opposite was true in Miami-Dade. Although Miami-Dade had greater invasive shrub cover and numbers of WIP species, San Juan had far greater invasive tree density, basal area and crown cover. This study provides an approach for incorporating field and available census data in geospatial distribution models of WIPs in cities throughout the globe. Findings indicate that identifying spatial predictors of WIPs depends on site-specific factors and the ecological scale of the predictor. Thus, mapping protocols and policies to eradicate urban WIPs should target indicators of a relevant scale specific to the area of interest for their improved and proactive management.

A pseudo-penalized quasi-likelihood approach to the spatial misalignment problem with non-normal data.

Spatially referenced datasets arising from multiple sources are routinely combined to assess relationships among various outcomes and covariates. The geographical units associated with the data, such as the geographical coordinates or areal-level administrative units, are often spatially misaligned, that is, observed at different locations or aggregated over different geographical units. As a result, the covariate is often predicted at the locations where the response is observed. The method used to align disparate datasets must be accounted for when subsequently modeling the aligned data. Here we consider the case where kriging is used to align datasets in point-to-point and point-to-areal misalignment problems when the response variable is non-normally distributed. If the relationship is modeled using generalized linear models, the additional uncertainty induced from using the kriging mean as a covariate introduces a Berkson error structure. In this article, we develop a pseudo-penalized quasi-likelihood algorithm to account for the additional uncertainty when estimating regression parameters and associated measures of uncertainty. The method is applied to a point-to-point example assessing the relationship between low-birth weights and PM2.5 levels after the onset of the largest wildfire in Florida history, the Bugaboo scrub fire. A point-to-areal misalignment problem is presented where the relationship between asthma events in Florida's counties and PM2.5 levels after the onset of the fire is assessed. Finally, the method is evaluated using a simulation study. Our results indicate the method performs well in terms of coverage for 95% confidence intervals and naive methods that ignore the additional uncertainty tend to underestimate the variability associated with parameter estimates. The underestimation is most profound in Poisson regression models.

UCH-L1 and GFAP Serum Levels in Neonates with Hypoxic-Ischemic Encephalopathy: A Single Center Pilot Study.

OBJECTIVE: We examined two potential biomarkers of brain damage in hypoxic-ischemic encephalopathy (HIE) neonates: glial fibrillary acidic protein (GFAP; a marker of gliosis) and ubiquitin C-terminal hydrolase L1 (UCH-L1; a marker of neuronal injury). We hypothesized that the biomarkers would be measurable in cord blood of healthy neonates and could serve as a normative reference for brain injury in HIE infants. We further hypothesized that higher levels would be detected in serum samples of HIE neonates and would correlate with brain damage on magnetic resonance imaging (MRI) and later developmental outcomes.? STUDY DESIGN: Serum UCH-L1 and GFAP concentrations from HIE neonates (n = 16) were compared to controls (n = 11). The relationship between biomarker concentrations of HIE neonates and brain damage (MRI) and developmental outcomes (Bayley-III) was examined using Pearson correlation coefficients and a mixed model design. RESULT: Both biomarkers were detectable in cord blood from control subjects. UCH-L1 concentrations were higher in HIE neonates (p < 0.001), and associated with cortical injury (p < 0.055) and later motor and cognitive developmental outcomes (p < 0.05). The temporal change in GFAP concentrations during (from birth to 96 h of age) predicted motor developmental outcomes (p < 0.05) and injury to the basal ganglia and white matter. CONCLUSION: Ubiquitin C-terminal hydrolase L1 and GFAP should be explored further as promising serum biomarkers of brain damage and later neurodevelopmental outcomes in neonates with HIE.

Estimated generalized least squares in spatially misaligned regression models with Berkson error.

In environmental studies, relationships among variables that are misaligned in space are routinely assessed. Because the data are misaligned, kriging is often used to predict the covariate at the locations where the response is observed. Using kriging predictions to estimate regression parameters in linear regression models introduces a Berkson error, which induces a covariance structure that is challenging to estimate. In addition, if the parameters associated with kriging (e.g. trend surface parameters and spatial covariance parameters) are estimated, then an additional uncertainty is introduced. We characterize the total measurement error as part of a broader class of Berkson error models and develop an estimated generalized least squares estimator using estimated covariance parameters. In working with the induced model, we fully account for the error structure and estimate the covariance parameters using likelihood-based methods. We provide insight into when it is important to fully account for the covariance structure induced from the different error sources. We assess the performance of the estimators using simulation and illustrate the methodology using publicly available data from the US Environmental Protection Agency.

RNA-seq: technical variability and sampling.

BACKGROUND: RNA-seq is revolutionizing the way we study transcriptomes. mRNA can be surveyed without prior knowledge of gene transcripts. Alternative splicing of transcript isoforms and the identification of previously unknown exons are being reported. Initial reports of differences in exon usage, and splicing between samples as well as quantitative differences among samples are beginning to surface. Biological variation has been reported to be larger than technical variation. In addition, technical variation has been reported to be in line with expectations due to random sampling. However, strategies for dealing with technical variation will differ depending on the magnitude. The size of technical variance, and the role of sampling are examined in this manuscript. RESULTS: In this study three independent Solexa/Illumina experiments containing technical replicates are analyzed. When coverage is low, large disagreements between technical replicates are apparent. Exon detection between technical replicates is highly variable when the coverage is less than 5 reads per nucleotide and estimates of gene expression are more likely to disagree when coverage is low. Although large disagreements in the estimates of expression are observed at all levels of coverage. CONCLUSIONS: Technical variability is too high to ignore. Technical variability results in inconsistent detection of exons at low levels of coverage. Further, the estimate of the relative abundance of a transcript can substantially disagree, even when coverage levels are high. This may be due to the low sampling fraction and if so, it will persist as an issue needing to be addressed in experimental design even as the next wave of technology produces larger numbers of reads. We provide practical recommendations for dealing with the technical variability, without dramatic cost increases.

A spatial model approach for assessing windbreak growth and carbon stocks.

Agroforestry, the deliberate integration of trees into agricultural operations, sequesters carbon (C) while providing valuable services on agricultural lands. However, methods to quantify present and projected C stocks in these open-grown woody systems are limited. As an initial step to address C accounting in agroforestry systems, a spatial Markov random field model for predicting the natural logarithm (log) of the mean aboveground volume of green ash ( Marsh.) within a shelterbelt, referred to as the log of aboveground volume, was developed using data from an earlier study and web-available soil and climate information. Windbreak characteristics, site, and climate variables were used to model the large-scale trend of the log of aboveground volume. The residuals from this initial model were correlated among sites up to 24 km from a point of interest. Therefore, a spatial dependence parameter was used to incorporate information from sites within 24 km into the prediction of the log of the aboveground volume. Age is an important windbreak characteristic in the model. Thus, the log of aboveground volume can be predicted for a given windbreak age and for values of other explanatory variables associated with a site of interest. Such predictions can be exponentiated to obtain predictions of aboveground volume for windbreaks without repeated inventory. With the capability of quantifying uncertainty, the model has the potential for large regional planning efforts and C stock assessments for many deciduous tree species used in windbreaks and riparian buffers once it is calibrated.

A comparison of errors in variables methods for use in regression models with spatially misaligned data.

When a response variable Y is measured on one set of points and a spatially varying predictor variable X is measured on a different set of points, X and Y have different supports and thus are spatially misaligned. To draw inference about the association between X and Y , X is commonly predicted at the points for which Y is observed, and Y is regressed on the predicted X. If X is predicted using kriging or some other smoothing approach, use of the predicted instead of the true (unobserved) X values in the regression results in unbiased estimates of the regression parameters. However, the naive standard errors of these parameters tend to be too small. In this article, two simulation studies are used to compare methods for providing appropriate standard errors in this spatial setting. Three of the methods are extended to the change-of-support case where X is observed at points, but Y is observed for areal units, and these approaches are also compared via simulation.

Advancing the science of EPA guidelines for sponsor-financed topical insect repellent efficacy studies.

In 1999, the US Environmental Protection Agency (EPA) published guidelines for product performance testing of skin-applied insect repellents, which provide guidance for topical insect repellent efficacy studies. EPA subsequently uses these sponsor-financed studies in their evaluation of proposed label claims. This paper reviews some of the statistical flaws in the proposed revisions to these guidelines and suggests possible improvements. This review is important because EPA's revisions to the 1999 guidelines do not address these issues.

The prevalence of injury for stroke caregivers and associated risk factors.

BACKGROUND: A vast body of research has shown that the emotional and physical demands of caregiving may lead to increased stress and burden. However, it is unknown whether these factors are associated with increased risk for injuries among caregivers. PURPOSE: Stroke survivors and their caregivers (N = 275) were surveyed using existing measures to explore the prevalence of physical injuries among caregivers and the types of injuries and factors associated with caregiver injury. METHODS: Caregiver measures included items from the National Alliance for Caregiving (NAC) survey and the short form of the Center for Epidemiologic Studies Depression Scale (CES-D). Veterans completed items on health and functioning from the Behavioral Risk Factors Surveillance Survey (BRFSS). RESULTS: We found that stroke caregivers who experience high burden levels and/or depression are considerably more likely to experience an increased risk for injury. Further, the vast majority of injured caregivers indicated that their injury interfered with their ability to provide care for the veteran. CONCLUSION: This suggests that injury on the part of the family caregiver may lead to the veteran's placement in a skilled nursing facility and lead to increased costs for the Department of Veteran's Affairs.

Linking health and environmental data in geographical analysis: it's so much more than centroids.

Programs and studies increasingly use existing data from multiple sources (e.g., surveillance systems, health registries, or governmental agencies) for analysis and inference. These data usually have been collected on different geographical or spatial units, with each varying from the ones of interest. Combining such disparate data creates statistical challenges. Florida's efforts to move toward implementing the Centers for Disease Control and Prevention (CDC)'s Environmental Public Health Tracking (EPHT) program aptly illustrate these concerns, which are typical of studies designed to measure the association between environmental and health outcomes. In this paper, we develop models of spatial associations between myocardial infarctions (MIs) and ambient ozone levels in Florida during August 2005 and use these models to illustrate the problems that can occur when making inferences from aggregated data, the concept of spatial support, and the importance of correct uncertainty assessment. Existing data on hospital discharges and emergency department visits were obtained from Florida's Agency for Health Care Administration. Environmental data were obtained from Florida's Department of Environmental Protection; sociodemographic data were obtained from the US Census Bureau; and data from CDC's Behavioral Risk Factor Surveillance System were used to provide additional information on other risk factors. We highlight the opportunities and challenges associated with combining disparate spatial data for EPHT analyses. We compare the results from two different approaches to data linkage, focusing on the need to account for spatial scale and the support of spatial data in the analysis. We use geographically weighted regression, not as a visual mapping tool, but as an inferential tool designed to indicate the need for spatial coefficients, a test that cannot be made by using the majority of Bayesian models. Finally, we use geostatistical simulation methods for uncertainty analysis to demonstrate its importance in models with predicted covariates. Our focus is on relatively simple methods and concepts that can be implemented with ESRI's(®) ArcGIS(®) software.

Education for worksite monitors of impaired nurses.

Boards of nursing sponsor programs, including those for alternatives to discipline, for recovering nurses. These programs rely on worksite monitors who are oftentimes other nurses or supervisors of nurses, to work with recovering nurses when they return to practice. The skills of these monitors vary with respect to understanding the monitor role and recognizing traits in chemical dependency and relapse. To determine the degree of content value and the best teaching method for monitors to learn program content, 17 currently active worksite monitors participated in a study to evaluate content value to 2 groups, new and experienced monitors, and to select the best method to teach 4 content topics. Results showed that current content was valued without necessary additions and that group instruction in urban areas was preferred over one-to-one instruction. Implementation of study outcomes yielded that issues of confidentiality made group instruction unsatisfactory.

Can a clinical continuing education course change behavior in dental hygiene practice?

Can a clinical continuing education course impact practitioner provision of care in everyday practice? National recommendations call for continuing education that is closely linked to practitioner learning at the point of care and for outcome assessment that measures the effectiveness of learning activities on the learner's practice behavior. Educational research has shown that interactive learning with clinical participation and the opportunity to practice is one of the most effective educational interventions. This study used an interactive educational intervention during a 3-day clinical course designed for dental hygienists. A follow-up survey was sent to determine whether the knowledge and skills that were taught in the course were applied subsequently to patients in practice. Sixty-one of 97 surveys were returned for a 63% response rate. Descriptive data including frequencies, means, medians, and standard deviations were obtained for all survey items. The majority of respondents reported moderate to high gains in knowledge and skills as well as application to patients in practice. The majority also identified continuing education as the primary source of information used when making changes in practice.

Assessing the association between environmental impacts and health outcomes: a case study from Florida.

The Centers for Disease Control and Prevention (CDC) created the Environmental Public Health Tracking (EPHT) program to integrate hazard monitoring, exposure, and health effects surveillance into a cohesive tracking network. Part of Florida's effort to move toward implementation of EPHT is to develop models of the spatial and temporal association between myocardial infarctions (MIs) and ambient ozone levels in Florida. Existing data were obtained from Florida's Agency for Health Care Administration, Florida's Department of Environmental Protection, the U.S. Census Bureau, and CDC's Behavioral Risk Factor Surveillance System. These data were linked by both ignoring spatial support and using block kriging, a support-adjusted approach. The MI data were indirectly standardized by age, race/ethnicity, and sex. The state of Florida was used as the comparison standard to compute the MI standardized event ratio (SER) for each county and each month. After the data were linked, global models were used initially to relate MIs to ambient ozone levels, adjusting for covariates. The global models provide an estimated relative MI SER for the state. Realizing that the association in MIs and ozone might change across locations, local models were used to estimate the relative MI SER for each county, again adjusting for covariates. Results differed, depending on whether the spatial support was ignored or accounted for in the models. The opportunities and challenges associated with EPHT analyses are discussed and future directions highlighted.

Geographic variation in poststroke depression among veterans with acute stroke.

This study compared patterns of poststroke depression (PSD) detection among veterans with acute stroke in eight U.S. geographic regions. Department of Veterans Affairs (VA) medical and pharmacy data as well as Medicare data were used. International Classification of Diseases-9th Revision depression codes and antidepressant medication dispensing were applied to define patients' PSD status 12 months poststroke. Logistic regression models were fit to compare VA PSD diagnosis and overall PSD detection between the regions. The use of VA medical data alone may underestimate the rate of PSD. Geographic variation in PSD detection depended on the data used. If VA medical data alone were used, we found no significant variation. If VA medical data were used along with Medicare and VA pharmacy data, we observed a significant variation in overall PSD detection across the regions after adjusting for potential risk factors. VA clinicians and policy makers need to consider enrollees' use of services outside the system when conducting program evaluation. Future research on PSD among veteran patients should use VA medical data in combination with Medicare and VA pharmacy data to obtain a comprehensive understanding of patients' PSD.

Should we fly or drive? The use of a computer-based system (ANTSS) for resource utilization.

Objective methods are needed to accurately allocate limited resources for the transport of critically ill patients. This manuscript presents the implementation of a novel internet-based charting system that employs a background physiologic acuity scoring system. Based on this score, the system, named ANTSS (Assessment of Neonates for Transport Scoring Software), assigns a color code to the patient (red, yellow, green). The system, based on the color coding, recommends a particular transport vehicle. ANTSS was initially evaluated using a survey to assess the systems effect on workload and the usefulness of the physiologic scoring. The survey was administered to the physicians who input the initial information. An objective count of the fields entered by the physicians using our older handwritten method was compared with the number of fields completed using ANTSS. Finally, a second survey was administered to members of the transport team to evaluate whether ANTSS-generated charts provided improved patient information to the end users of the information. The results from the surveys demonstrate that ANTSS improves the patient information provided to the transport team, provides the physician with useful objective information concerning the mode of transport, and increases the amount of completed fields in the transport record.

The impact of poststroke depression on healthcare use by veterans with acute stroke.

BACKGROUND AND PURPOSE: Poststroke depression (PSD) is common among stroke survivors, and it is associated with worse functional outcomes and increased poststroke mortality. Limited information is available about its impact on healthcare use. This study assessed the impact of PSD on healthcare use by veterans with acute stroke. METHODS: In this retrospective, observational national study, 5825 veterans with acute stroke were identified from Veterans Affairs' (VA) inpatient databases. To determine the patients' comprehensive PSD and use status, VA and Medicare fee-for-service inpatient and outpatient as well as VA pharmacy data were used. PSD was established if a patient had an inpatient or outpatient depression diagnosis or if a patient received one of the antidepressants within the VA 12 months postindex stroke. Healthcare use referred to the number of hospital stays, outpatient visits, and cumulative length of inpatient stays under both VA and Medicare fee-for-service programs. Poisson regression was fitted to estimate the impact of PSD on use controlling for sociodemographic, clinical, and disease severity factors. RESULTS: Forty-one percent of the sample had PSD. After adjusting for patient demographic and clinical factors, we found that the patients with stroke with PSD had significantly (P<0.0001) more hospitalizations, outpatient visits, and longer length of stays 12 months poststroke compared with these patients with stroke without PSD. CONCLUSIONS: Patients with PSD had greater 12-month poststroke healthcare use even when controlling for other demographic and clinical variables. Early detection and appropriate management of PSD for veterans with acute stroke may help reduce their poststroke healthcare use.