Comparing several score tests for interval censored data

I create a general model to perform score tests on interval censored data. Special cases of this model are the score tests of Finkelstein, Sun and Fay. Although Sun's was derived as a test for discrete data and Finkelstein's and Fay's tests were derived under a grouped continuous model, by writing all tests under one general model we see that as long as the regularity conditions hold, any of these three classes of tests may be applied to either grouped continuous or discrete data. I show the equivalence between the weighted logrank form of the general test and the form with a term for each individual, the form often used with permutation tests. From the weighted logrank form of the tests, we see that Sun's and Finkelstein's test are similar, giving constant (or approximately constant) weights to differences in survival distributions over time. In contrast, the proportional odds model (Fay's model with logistic error) gives more weight to early differences.     

ICEstimator在蚊虫毒理学半数致死浓度计算上的应用

目的探索ICEstimator用来计算半数致死浓度（LC50）值及其95％可信区间（95％CI）的可行性和优越性。方法利用ICEstimator分别计算敌百虫对白纹伊蚊、致倦库蚊的LC50及其95％CI和氟氯氰菊酯对白纹伊蚊、致倦库蚊的LC50及其95％CI,并与目前常用的计算方法（SPSS、SAS、DPS）对上述数值进行分析比较。结果经DPS／SAS/SPSS软件计算氟氯氰菊酯对白纹伊蚊及致倦库蚊的LC50（mg／ml）及其95％CI分别为5．03（4．68～5．38）、5．20（4．90～5．60）、5．14（4．83～5．46）、5．14（4．83～5．46）和5．31（4．58～6．03）、5．40（4．75～6．15）、5．28（2．37～7．11）、5．28（2．37～7．11）;敌百虫对白纹伊蚊及致倦库蚊的LC50（mg/ml）及其95％CI分别为92．92（83．27～102．58）、100．60（90．60～110．70）、96．00（87.88～105．33）、96．00（87．88～105．33）和1123．02（998-89～1247．14）、1123．70（800．60～1652．40）、1111．91（725．47～1745．88）、1111．90（725．46～1745．87）。经非参数K-W-Test检验,二者计算的数值及其95％CI差异均无统计学意义（χ^2=5．595,P=0．113）。结论用ICEstimator可快速得到某杀虫剂对某种蚊虫的LC50及其95％CI,与SPSS和SAS方法比较,该计算方法计算过程简单,速度快,更加有效地利用了初始数据。     

Modeling explicitly and mechanistically median lethal concentration as a function of time for risk assessment

A mechanistic model that explains how toxic effects depend on the duration of exposure has been developed. Derived from the dynamic energy budget (DEB)tox model, it expresses the hazard rate as a function of the toxic concentration in the organism. Using linear approximations in accordance with the general simplifications made in DEBtox, the concentration that induces x% of lethality (LCx) and in particular the lethal concentration 50% (LC50) are expressed explicitly as functions of time. Only three parameters are required: an asymptotic effect concentration, a time constant, and an effect velocity. More sophisticated (but still analytic) models are possible, describing more complex toxicity patterns such as an increase of sensitivity with time or, conversely, an adaptation. These models can be fitted to the common and widespread LC50 endpoints available from the literature for various aquatic species and chemicals. The interpretation of the values assigned to the parameters will help explain the toxicity processes and standardize toxicity values from different sources for comparisons.     

A simple method to calculate the confidence interval of a standardized mortality ratio (SMR)

In analyzing standardized mortality ratios (SMRs), it is of interest to calculate a confidence interval for the true SMR. The exact limits of a specific interval can be obtained by means of the Poisson distribution either within an iterative procedure or by one of the tables. The limits can be approximated in using one of various shortcut methods. In this paper, a method is described for calculating the exact limits in a simple and easy way. The method is based on the link between the chi 2 distribution and the Poisson distribution. Only a table of the chi 2 distribution is necessary.     

Comparing diagnostic tests: a simple graphic using likelihood ratios

The diagnostic abilities of two or more diagnostic tests are traditionally compared by their respective sensitivities and specificities, either separately or using a summary of them such as Youden's index. Several authors have argued that the likelihood ratios provide a more appropriate, if in practice a less intuitive, comparison. We present a simple graphic which incorporates all these measures and admits easily interpreted comparison of two or more diagnostic tests. We show, using likelihood ratios and this graphic, that a test can be superior to a competitor in terms of predictive values while having either sensitivity or specificity smaller. A decision theoretic basis for the interpretation of the graph is given by relating it to the tent graph of Hilden and Glasziou (Statistics in Medicine, 1996). Finally, a brief example comparing two serodiagnostic tests for Lyme disease is presented. Published in 2000 by John Wiley & Sons, Ltd.     

Non-parametric estimation for the difference or ratio of median failure times for paired observations

We propose a non-parametric method to calculate a confidence interval for the difference or ratio of two median failure times for paired observations with censoring. The new method is simple to calculate, does not involve non-parametric density estimates, and is valid asymptotically even when the two underlying distribution functions differ in shape. The method also allows missing observations. We report numerical studies to examine the performance of the new method for practical sample sizes.     

Shortcomings of the laboratory-derived median lethal concentration for predicting mortality in field populations: exposure duration and latent mortality

Exposure duration and intensity (concentration or dose) determine lethal effects of toxicants. However, environmental regulators have focused on exposure intensity and have considered duration only peripherally. Conventional testing for toxicology tends to fix exposure time and to use the median lethal concentration (LC50) at that time to quantify mortality. Fixing the exposure duration and selecting the 50% mortality level for reasons of statistical and logistical convenience result in the loss of ecologically relevant information generated at all other times and ignore latent mortality that manifests after the exposure ends. In the present study, we used survival analysis, which is widely employed in other fields, to include both time and concentration as covariates and to quantify latent mortality. This was done with two contrasting toxicants, copper sulfate (CuSO4) and sodium pentachlorophenol (NaPCP). Amphipods (Hyalella azteca) were exposed to different toxicant concentrations, and the percentage mortalities were noted both during and after the exposure ended. For CuSO4 at the conventional 48-h LC50 concentrations, the predicted proportions dead after including latent mortality were 65 to 85%, not 50%. In contrast, only 5% or fewer additional animals died if the latent mortality was included for NaPCP. The data (including exposure time, concentration, and proportion dead at each time) for each toxicant were then successfully fit with survival models. The proportion of organisms dying at any combination of exposure concentration and time can be predicted from such models. Survival models including latent mortality produced predictions of lethal effects that were more meaningful in an ecological or field context than those from conventional LC50 methods.     

Avoidance tests with small fish: determination of the median avoidance concentration and of the lowest-observed-effect gradient

The purpose of the present study was to develop sensitive, rapid, and easily quantified avoidance tests for small fish (Danio rerio) in order to provide important ecological information during toxicity assessments. Fish were exposed in three replicate linear flow-through chambers consisting of five compartments. The test system was found to provide a linear contamination gradient, with mean dilutions in each compartment of 90, 70, 50, 30, and 10%. Also, in the absence of a toxic gradient, the fish were uniformly distributed along the five-compartment chambers. Then the apparatus was evaluated by exposing fish to a concentration gradient of copper and a dilution gradient of a field sample contaminated with acid mine drainage (AMD). Avoidance was monitored at 24-h intervals up to 96 h of exposure. The avoidance of copper and AMD by D. rerio was confirmed. The apparatus enabled quantification of median avoidance effect concentrations or dilutions (EC50 or EDil50) and also lowest-observed-effect gradients, which express the minimum toxicant gradient eliciting avoidance, a parameter increasing the ecological relevance of the laboratory avoidance responses. For quantifying avoidance, a 24-h exposure was sufficient, as the 24- to 96-h EC50 and EDil50 values were similar. The avoidance response was easy and rapid to quantify, leading this test to routine use in environmental risk assessment.     

Simultaneous interval estimates of the odds ratio in studies with two or more comparisons

More than one odds ratio estimate will often arise from a single epidemiologic study. Examples of designs where this may occur include those where there is more than one case or control group, and investigations of several risk factors as part of the same study. Various methods for presenting multiple interval estimates are discussed, including: the naive method, the Bonferroni method, the Dunn method, the Scheffé method, and the Dunnett method. For rectangular regions the Dunnett method gives a region with the most appropriate confidence level, but this region contains a different set of odds ratio estimates than are implied by the usual significance tests. A confidence ellipse circumscribed by the Scheffé limits gives the best agreement with the significance tests. Each of these methods is illustrated with a numerical example.     

A measurement of the efficacy of nosocomial infection control using the 95 per cent confidence interval for infection rates

From 1981 through 1985, the authors studied the changes in monthly nosocomial infection rates at the University of Virginia Hospital in Charlottesville, Virginia using the 95% confidence interval for infection rates as a marker of the efficacy of infection control activities. For a 99-month baseline period, monthly infection rates were calculated and the 95% confidence interval was established. In the 60 study months, each monthly rate was compared with the 95% confidence interval for that particular month. At the end of each study year, the monthly infection rates were incorporated into the existing confidence interval. Of 60 monthly rates during the study period, 30 were below the confidence interval (p less than 0.00001), two were above the confidence interval (p = 0.23), and 28 were within the confidence interval. Since there was no reduction in surveillance activity, patient case-mix index, or laboratory sensitivity for organism recovery, these results suggest that monthly nosocomial infection rates at this hospital have decreased when compared with the baseline period. The use of the 95% confidence interval may provide a measure of the efficacy of infection control activities, suggest temporal intervals requiring more intensive infection surveillance, and provide a method for examining the variability in monthly infection rates.     

PET/CT性能检测中18F用量分析

目的探讨正电子发射计算机断层显像/计算机体层扫描仪(PET/CT)性能检测过程中18F的放射性活度及活度浓度准确使用量。方法按照NEMANU2-2001标准,采用其相关模体,结合飞利浦Gemini系列,GE Discovery系列和西门子Biograph系列PET/CT性能检测项目所使用18F的放射性活度或放射性活度浓度进行分析。结果①空间分辨率:飞利浦1.48～2.22GBq/ml;GE>185MBq/ml;西门子1.11GBq/ml);②散射因子、计数丢失和随机符合测量及精确性(计数丢失和随机符合校正):飞利浦481～555MBq;GE900MBq;西门子1.07GBq;③灵敏度:飞利浦7.4MBq;GE10MBq;西门子4.6MBq。结论各厂商提供的放射性活度及活度浓度可用于NEMA标准检测。     

Comparison of sublethal and lethal criteria for nine different chemicals in standardized toxicity tests using the earthworm Eisenia andrei.

In this study, the effects of nine different chemicals on the survival, growth, and reproduction of the earthworm species Eisenia andrei were determined using a recently developed method. Earthworms were exposed for 3 weeks to the test chemicals in an artificial soil substrate. Additional data on the acute toxicity of these chemicals were derived from the literature. For some chemicals, cocoon production was the most sensitive parameter (cadmium, chromium, paraquat, fentin, benomyl, phenmedipham), while for others cocoon hatchability was most sensitive (pentachlorophenol, parathion, carbendazim). In the case of parathion, growth of the worms seemed to be even more sensitive than reproduction. As an overall parameter for the effect on earthworm reproduction, the total number of juveniles produced per worm appeared to be a useful parameter. Differences between (acute) LC50 values and the lowest NOEC value for effects on growth and reproduction were different for each chemical. Difference was greatest for cadmium (a factor of greater than 100) and smallest for fentin, benomyl, and pentachlorophenol (a factor of 5-6).     

Reliability of 95% confidence interval revealed by expected quality-of-life scores: an example of nasopharyngeal carcinoma patients after radiotherapy using EORTC QLQ-C 30

Background  

Many researchers use observed questionnaire scores to evaluate score reliability and to make conclusions and inferences regarding quality-of-life outcomes. The amount of false alarms from medical diagnoses that would be avoided if observed scores were substituted with expected scores is interesting, and understanding these differences is important for the care of cancer patients. Using expected scores to estimate the reliability of 95% confidence intervals (CIs) is rarely reported in published papers. We investigated the reliability of patient responses to a quality-of-life questionnaire and made recommendations for future studies of the quality of life of patients.     

Statistical tests, <Emphasis Type="Italic">P</Emphasis> values,confidence intervals,and power: a guide to misinterpretations

Misinterpretation and abuse of statistical tests, confidence intervals, and statistical power have been decried for decades, yet remain rampant. A key problem is that there are no interpretations of these concepts that are at once simple, intuitive, correct, and foolproof. Instead, correct use and interpretation of these statistics requires an attention to detail which seems to tax the patience of working scientists. This high cognitive demand has led to an epidemic of shortcut definitions and interpretations that are simply wrong, sometimes disastrously so—and yet these misinterpretations dominate much of the scientific literature. In light of this problem, we provide definitions and a discussion of basic statistics that are more general and critical than typically found in traditional introductory expositions. Our goal is to provide a resource for instructors, researchers, and consumers of statistics whose knowledge of statistical theory and technique may be limited but who wish to avoid and spot misinterpretations. We emphasize how violation of often unstated analysis protocols (such as selecting analyses for presentation based on the P values they produce) can lead to small P values even if the declared test hypothesis is correct, and can lead to large P values even if that hypothesis is incorrect. We then provide an explanatory list of 25 misinterpretations of P values, confidence intervals, and power. We conclude with guidelines for improving statistical interpretation and reporting.     

Characterization of exposure to pollutants in spray painting industries using room air concentration values and an exposure test

Occupational-hygienic measurements of the interior air concentration had been performed in a large number of spray painting plants. The analyses stated that with most of the cases the exposition risk on varnishers is characterized by aerosol concentration and the concentration of lead and chromate in it. The concentrations of vapours of solvent mixtures are of subordinate importance.     

Unpredictable bias when using the missing indicator method or complete case analysis for missing confounder values: an empirical example

ObjectiveMissing indicator method (MIM) and complete case analysis (CC) are frequently used to handle missing confounder data. Using empirical data, we demonstrated the degree and direction of bias in the effect estimate when using these methods compared with multiple imputation (MI).Study Design and SettingFrom a cohort study, we selected an exposure (marital status), outcome (depression), and confounders (age, sex, and income). Missing values in “income” were created according to different patterns of missingness: missing values were created completely at random and depending on exposure and outcome values. Percentages of missing values ranged from 2.5% to 30%.ResultsWhen missing values were completely random, MIM gave an overestimation of the odds ratio, whereas CC and MI gave unbiased results. MIM and CC gave under- or overestimations when missing values depended on observed values. Magnitude and direction of bias depended on how the missing values were related to exposure and outcome. Bias increased with increasing percentage of missing values.ConclusionMIM should not be used in handling missing confounder data because it gives unpredictable bias of the odds ratio even with small percentages of missing values. CC can be used when missing values are completely random, but it gives loss of statistical power.