Generalized linear mixture models for handling nonignorable dropouts in longitudinal studies

This paper presents a method for analysing longitudinal data when there are dropouts. In particular, we develop a simple method based on generalized linear mixture models for handling nonignorable dropouts for a variety of discrete and continuous outcomes. Statistical inference for the model parameters is based on a generalized estimating equations (GEE) approach (Liang and Zeger, 1986). The proposed method yields estimates of the model parameters that are valid when nonresponse is nonignorable under a variety of assumptions concerning the dropout process. Furthermore, the proposed method can be implemented using widely available statistical software. Finally, an example using data from a clinical trial of contracepting women is used to illustrate the methodology.     

Residuals for proportional hazards models with interval-censored survival data

We develop diagnostic tools for use with proportional hazards models for interval-censored survival data. We propose counterparts to the Cox-Snell, Lagakos (or martingale), deviance, and Schoenfeld residuals. Many of the properties of these residuals carry over to the interval-censored case. In particular, the interval-censored versions of the Lagakos and Schoenfeld residuals may be derived as components of suitable score statistics. The Lagakos residuals may be used to check regression relationships, while the Schoenfeld residuals can help to detect nonproportional hazards in semiparametric models. The methods apply to parametric models and to the semiparametric model with discrete observation times.     

The D' measure of overall gametic disequilibrium between pairs of multiallelic loci

Abstract The D ' coefficient is one of the most commonly used measures of the extent of gametic disequilibrium between multiallelic loci. It has been suggested that the range of the D ' measure of overall disequilibrium between pairs of multiallelic loci depends on allele frequencies, except under some very restricted conditions. Nevertheless, the problem of dependence of the range of D ' has not been characterized under a wide set of possible polymorphisms. Evaluation of the utility of D ' as a measure of the strength of overall disequilibrium between all possible pairs of alleles at two multiallelic loci requires better knowledge of its range than is currently available. In this work, the conditions of polymorphism under which the range of D ' is frequency independent are given. It is found that the range of D ' is more often independent of allelic frequencies than is commonly thought. Furthermore, the range of D ' undergoes only small fluctuations as a function of the polymorphisms at the loci. Numerical cases and microsatellite data from humans are used for illustration. These observations indicate that the D ' coefficient is a useful tool for the estimation and comparison of the extent of overall disequilibrium across pairs of multiallelic loci.     

Evidence for a nutritional disorder of Oxalis acetosella L. on acid forest soils; II. Diagnostic field experiments and nutrient solution studies

Different field experiments were performed to discover the main factor(s) responsible for the poor leaf growth, moderate propagation and leaf chlorosis of Oxalis acetosella in an old Norway spruce stand with acid soil (Höglwald Forest). A previous study had suggested, that Ca (or Mg) deficiency or Mn toxicity could be involved.In a Main Diagnostic Field Experiment with an intact population, Ca and Mg were either applied as sulphate or carbonate to distinguish between nutritional and pH effects. Mn and Si were also applied to aggravate or overcome possible effects of Mn toxicity. Enhancement experiments with different amounts of CaSO₄ were conducted to investigate the Ca dose-effect relationship under field conditions. Additional trials with SrCO₃, BaCO₃ and NaHCO₃ had the goal to raise the soil pH without supply of nutrients.Greenhouse experiments with Oxalis acetosella supplemented the field studies by investigating the Ca and Mn dose-effect relationships under controlled conditions. Growth, vitality and nutrition of Oxalis were studied in a nutrient solution culture at pH 4.0 over a range of concentrations of Ca (20 to 5000 mol L_-1) and Mn (5 to 1000 mol L_-1) respectively. Furthermore, the effects of two contrasting ammonium/nitrate ratios were tested. The nutritional composition of the basal nutrient solution and the microclimate in the greenhouse were as far as possible adjusted to the environmental conditions of the plant in the Höglwald Forest.All these studies led to the conclusion, that the moderate growth and vitality of Oxalis in the Höglwald Forest was mainly due to an insufficient Ca supply, rather than an effect of Mg deficiency, low soil pH or Mn toxicity. The application of CaSO₄ caused a similar stimulation of the growth as CaCO₃. A clearly positive, close CaSO₄ dose-effect relationship was detected in field experiments as well as in the nutrient solution study. The same type of leaf chloroses as in the field was reproduced through low Ca nutrient solutions. Predominant ammonium nutrition may significantly impair Ca uptake.Oxalis acetosella displayed a relatively high leaf tissue tolerance of excessive Mn. There was no indication for a Mn-induced Ca deficiency in the Höglwald Forest. Enhanced Si uptake led to a partial vitalization of Oxalis; the reason for that remained unclear.     

Regulatory Perspectives on the Significance of Ecological Changes as Reported in Ecological Risk Assessments

A broad range of perspectives exists regarding the interpretation of potentially adverse ecological changes in ecological risk assessments conducted under Superfund and RCRA. While USEPA's Proposed Guidelines for Ecological Risk Assessment recommend determining whether predicted changes are adverse based on the nature of effects, intensity of effects, spatial scale, temporal scale, and potential for recovery, the guidelines do not provide specific stan dards for judging adversity. Hence, implementation of the proposed guide lines varies with each risk manager's subjective judgments regarding the relative importance of each of these five criteria. In an effort to increase consistency in the scientific interpretation of ecological risk assessments, the following practices are recommended. First, measures of effects should focus on levels of ecological organization that are more complex than the individual organ ism. Second, multiple lines of evidence should be evaluated for each assessment endpoint. Third, bioequivalence testing should be used in place of traditional statistical testing (e.g., Student t-test) because the goal of bioequivalence testing is to answer the biologically relevant question of whether measurements differ by, at most, a biologically small amount. Fourth, in defining biologically small differences, site-specific and species specific conditions should be considered to the greatest extent possible. Fifth, where the outcomes of multiple lines of evidence contradict one another, the risk assessor should employ a quantitative approach to weighing the evidence based on the scientific defensibility of each measure of effect.     

Quantitative multiplexed strategies for human Lyme disease serological testing

Lyme disease, which is primarily caused by infection with the bacterium Borrelia burgdorferi in the United States or other Borrelia species internationally, presents an ongoing challenge for diagnostics. Serological testing is the primary means of diagnosis but testing approaches differ widely, with varying degrees of sensitivity and specificity. Moreover, there is currently no reliable test to determine disease resolution following treatment. A distinct challenge in Lyme disease diagnostics is the variable patterns of human immune response to a plurality of antigens presented by Borrelia spp. during the infection. Thus, multiplexed testing approaches that capture these patterns and detect serological response against multiple antigens may be the key to prompt, accurate Lyme disease diagnosis. In this review, current state-of-the-art multiplexed diagnostic approaches are presented and compared with respect to their diagnostic accuracy and their potential for monitoring response to treatment.     

Molecular taxonomic identification in the absence of a ‘barcoding gap’: a test with the endemic flora of the Canarian oceanic hotspot

We use a comprehensive subset of Canarian angiosperms corresponding to 23 families, 35 genera and 60 Canarian endemic taxa to test whether this flora is suitable to taxonomic identification with the two proposed plant DNA barcode sequences and whether these sequences may reveal the existence of cryptic species overlooked by morphology. The rate of discrimination success between the insular congeneric samples using the rbcL+matK combination and a ‘character‐based’ approach (where we use only the combination of nucleotide positions in an alignment that allows unambiguous species identification) is higher (82.29%) than that obtained with the ‘distance‐based’ approach (80.20%) used by the CBOL Plant Working Group in 2009 and also when compared with tests conducted in other floras. This suggests that the molecular identification of the Canarian endemic flora can be achieved as successfully as in other floras where the incidence of radiation is not as relevant. The facts that (i) a distance‐based criterion was unable to discriminate between congeneric and conspecific comparisons and (ii) only the character‐based discrimination criterion resolved cases that the distance‐based criterion did not, further support the use of a character discrimination approach for a more efficient DNA barcoding of floras from oceanic islands like the Canaries. Thus, a barcoding gap seems not to be necessary for the correct molecular characterization of the Canarian flora. DNA barcodes also suggest the possible existence of cryptic taxa to be further investigated by morphology and that the current taxonomic status of some of the taxa analysed may need revision.     

Survey of electromagnetic field exposure in bedrooms of residences in lower Austria

Previous investigations of exposure to electric, magnetic, or electromagnetic fields (EMF) in households were either about electricity supply EMFs or radio frequency EMFs (RF‐EMFs). We report results from spot measurements at the bedside that comprise electrostatic fields, extremely low‐frequency electric fields (ELF‐EFs), extremely low‐frequency magnetic fields (ELF‐MFs), and RF‐EMFs. Measurements were taken in 226 households throughout Lower Austria. In addition, effects of simple reduction measures (e.g., removal of clock radios or increasing their distance from the bed, turning off Digital Enhanced Cordless Telecommunication (DECT) telephone base stations) were assessed. All measurements were well below International Commission on Non‐Ionizing Radiation Protection (ICNIRP) guideline levels. Average night‐time ELF‐MFs (long‐term measurement from 10 pm to 6 am, geometric mean over households) above 100 nT were obtained in 2.3%, and RF‐EMFs above 1000 µW/m² in 7.1% of households. Highest ELF‐EFs were primarily due to lamps beside the bed (max = 166 V/m), and highest ELF‐MFs because of transformers of devices (max = 1030 nT) or high current of power lines (max = 380 nT). The highest values of RF‐EMFs were caused by DECT telephone base stations (max = 28979 µW/m²) and mobile phone base stations (max = 4872 µW/m²). Simple reduction measures resulted in an average decrease of 23 nT for ELF‐MFs, 23 V/m for ELF‐EFs, and 246 µW/m² for RF‐EMFs. A small but statistically significant correlation between ELF‐MF exposure and overall RF‐EMF levels of R = 0.16 (P = 0.008) was computed that was independent of type (flat, single family) and location (urban, rural) of houses. Bioelectromagnetics 31:200–208, 2010. © 2009 Wiley‐Liss, Inc.     

A Bayesian Chi‐Squared Goodness‐of‐Fit Test for Censored Data Models

Summary We propose a Bayesian chi‐squared model diagnostic for analysis of data subject to censoring. The test statistic has the form of Pearson's chi‐squared test statistic and is easy to calculate from standard output of Markov chain Monte Carlo algorithms. The key innovation of this diagnostic is that it is based only on observed failure times. Because it does not rely on the imputation of failure times for observations that have been censored, we show that under heavy censoring it can have higher power for detecting model departures than a comparable test based on the complete data. In a simulation study, we show that tests based on this diagnostic exhibit comparable power and better nominal Type I error rates than a commonly used alternative test proposed by Akritas (1988, Journal of the American Statistical Association 83, 222–230). An important advantage of the proposed diagnostic is that it can be applied to a broad class of censored data models, including generalized linear models and other models with nonidentically distributed and nonadditive error structures. We illustrate the proposed model diagnostic for testing the adequacy of two parametric survival models for Space Shuttle main engine failures.