Explorative statistical analysis and the valuation of hypotheses

Clinical research has to include data from various sources. In many cases, the structure of the data allows explorative analysis only. This paper is concerned with situations in which no special hypotheses from a large set of possible ones have been predetermined for statistical testing or wherein tests are planned for hypotheses on only very few variables out of a high dimensional response as in controlled clinical trials. Explorative analysis of those experiments frequently uses p-values for the valuation of the hypotheses not designated for testing. It is shown by means of an example to what extent conclusions may be wrong if they result from ignoring (or misunderstanding) the explorative character of such p-values. Posterior probabilities with respect to suitable priors are proposed as an alternative explorative method for the valuation of hypotheses. The advantages of this procedure are discussed.     

Causal mediation analysis with multiple mediators in the presence of treatment noncompliance

Randomized experiments are often complicated because of treatment noncompliance. This challenge prevents researchers from identifying the mediated portion of the intention‐to‐treated (ITT) effect, which is the effect of the assigned treatment that is attributed to a mediator. One solution suggests identifying the mediated ITT effect on the basis of the average causal mediation effect among compliers when there is a single mediator. However, considering the complex nature of the mediating mechanisms, it is natural to assume that there are multiple variables that mediate through the causal path. Motivated by an empirical analysis of a data set collected in a randomized interventional study, we develop a method to estimate the mediated portion of the ITT effect when both multiple dependent mediators and treatment noncompliance exist. This enables researchers to make an informed decision on how to strengthen the intervention effect by identifying relevant mediators despite treatment noncompliance. We propose a nonparametric estimation procedure and provide a sensitivity analysis for key assumptions. We conduct a Monte Carlo simulation study to assess the finite sample performance of the proposed approach. The proposed method is illustrated by an empirical analysis of JOBS II data, in which a job training intervention was used to prevent mental health deterioration among unemployed individuals.     

A run‐based procedure to identify time‐lagged gene clusters in microarray experiments

Motivation: The analysis of gene‐expression data obtained from microarray experiments can be useful to identify regulatory relationship between genes. Genes with a common functional role have similar expression patterns across different microarray experiments. These similar expression patterns are perhaps due to co‐regulation of genes in the same functional group. Most of the existing methods available for the identification of the regulatory relationships are either made for comparing two genes at a time or methods are not computationally efficient in the identification of the regulatory relationships. The procedures adopted by these methods do not use complete information contained in the data set. In this paper, we propose a statistical procedure, which will use the information contained in the data set to cluster genes that show similar patterns. The proposed procedure compares several genes at a time instead of pair‐wise comparisons as done in some of the other procedures. The proposed procedure provides gene clusters based on time‐lagged data sets with more details. The proposed method provides a numerical value that would facilitate in comparing different sets of data obtained from different expressions. It also provides the identification of the gene involved and the time point at which the observation is made so that proper medicine can be developed for the gene‐specific and time‐specific disease. Results: We applied the proposed procedure on the Spellman data set (Mol. Biol. Cell 1998; 9 (12): 3273–3297) and compared our procedure with some of the other existing procedures. We found that our procedure is more computationally efficient than Ji and Tan (Bioinformatics 2005; 21 :509–516), event method and edge detection procedures. The proposed procedure also provides more details about the clusters than Ji and Tan (Bioinformatics 2005; 21 :509–516), event method and edge detection procedures. The proposed procedure is really simple to apply as compared with other available procedures in the literature including Ji and Tan (Bioinformatics 2005; 21 :509–516), event method and edge detection procedures. Copyright © 2008 John Wiley & Sons, Ltd.     

Improved method for the determination of urinary delta-aminolevulinic acid in lead workers

An improved method for the determination of urinary delta-aminolevulinic acid (ALA) as an indicator of lead absorption is described. It is a modification of the analytical procedure described by Tomokuni and Ogata in 1972. In the present method, the spot urine from lead-exposed workers which was diluted 3-fold with distilled water beforehand was used as the sample to eliminate as much as possible the urinary matrix interference. In addition, methyl acetoacetate was used as the condensation reagent for ALA-pyrrole, because the color intensity of ALA-pyrrole was higher with the use of methyl acetoacetate than that with ethyl acetoacetate which was used in the original method. The urinary ALA value determined by the present method was similar to that measured using an internal-standard addition method. This method is suitable as a rapid and inexpensive screening procedure for the routine quantitative analysis of urinary ALA.     

Genome scan for association and linkage

In this paper, a genome search is performed on the GAW Problem 1 data, in an attempt to determine which, if any, of the marker loci are associated and/or linked with the disease. Since there was no clear indication from the data of the likely mode of inheritance, methods were used which did not require such assumptions to be made. A two-stage procedure was used to test for association. Firstly a standard unmatched case-control test was applied to all the loci. The family-based method of Self et al. [1991] was then applied to those loci which gave a positive result in the first stage. This procedure correctly detected loci 1 and 2, and that disease risk was increased for homozygote carriers of the disease allele at each locus, although a false positive result was also found. The affected sib pair method of Holmans [1993] was also applied to the data, although the sample contained far too few sib-pairs for such an analysis to be effective. This analysis failed to find any of the disease loci. ©1995 Wiley-Liss, Inc.     

A method for the analysis of repeated binary outcomes in randomized clinical trials with non-compliance

When analysing repeated binary data from randomized trials, the model-based approaches, such as generalized estimating equations, are frequently used. Such methods ignore compliance information and give the model-based intention-to-treat estimate of treatment effect. In this paper, the design-based (randomization-based) semi-parametric estimation procedure is given in the estimation of causal risk difference. The resulting risk difference estimator is interpreted as an extension of the instrumental variables estimator for a binary outcome which has the causal interpretation. Extension of the proposed method to stratified analysis is given for data from stratified randomization or meta-analysis. It yields a Mantel-Haenszel type risk difference estimator. As a special case of stratified analysis, the pattern mixture model which stratifies the data by pattern of missing data is performed. Application of the proposed method to a trial in which endpoints were the occurrences of fever over three courses is provided. The same ideas are applied to the causal risk ratio estimation.     

An application of statistical matching with the survey of income and education and the 1976 Health Interview Survey.

This article outlines an alternative procedure to household surveys for obtaining individual observation-level data. The procedure, called statistical matching, integrates data on an individual observation from one source with data on a different observation identified as the "best matching" or "most similar" record from a second source. The best match is determined by objective statistical criteria. Also reported is a significant application of the procedure between the Survey of Income and Education and the 1976 National Health Interview Survey. The success of merging these two large, nationally representative data files shows statistical matching as a viable method of creating databases for health services research.     

A geometric approach to the analysis of physiological flow data

Physiological flow data are common in various medical fields. Examples include urinary, blood and expiratory flows. They are widely used in assessing functions in the urinary, circulatory, or pulmonary systems, respectively. Current statistical methods for analysing these flow data in clinical trials are either univariate analyses, which do not utilize all the information together, or some conventional multivariate methods (such as regression analyses) which yield results that do not render clear medical interpretations. This paper presents a new approach to analysing the flow data, using urinary flow as the primary focus. The basic idea and technical steps are applicable to other flow data as well. The proposed method aims to transform the flow measurements back to the shape of the flow graphs. Since the whole geometric pattern of the flow graph provides more information about the patient's flow condition than any individual flow parameter alone, the method is a meaningful way of combining and analysing the flow data in both statistical and clinical senses. The method is a three-stage procedure. Patients are classified into three classes in the first stage and then ranked in sequence in the second stage, according to the geometry of the shape pattern and some clinical criteria. The classification procedure is shown to be very reliable when compared with the clinician's visual evaluation, and hence can be implemented by computer programming to aid clinical trials involving many patients. The whole ranking score is then readily analysed at the third stage for comparing treatment effects by the analysis of covariance method based on ranks, with the post-treatment score as the response variable and the baseline score as the covariate. An example of a urinary flow data set is provided to illustrate the use of the procedure.     

Accuracy of risk indicators of the Infant Life Defense Program in a region of the S?o Paulo State, Brazil]

The Infant Life Defense Program of Health Department of Bauru City SP, Brazil, has diagnostic criteria for the inclusion of newborns. The diagnostic criteria combine eleven social and clinical indices of infant mortality risk, defined by in the light of a review of the literature and previous case studies. The indices are easily collected at the hospitals during delivery. The objective of this study is to propose an alternative diagnostic criteria, using the same social and clinical indices, of greater sensitivity and same proportion of children included in the program. The data on the newborns were collected between May 11, 1986 and November 10, 1987. The mortality period was defined as between 7 days and 6 months, which was the follow-up period for the infants enrolled in the program. The method of analysis was the determination for each index of the crude relative risk in a univariate analysis and the adjusted relative risk using the logistic regression procedure. A score system was constructed on the basis of the sum of the excess risk of each index.     

Interval estimation of quantile ratios applied to anti‐cancer drug screening by xenograft experiments

The current practice in analyzing data from anti‐cancer drug screening by xenograft experiments lacks statistical consideration to account for experimental noise, and a sound inference procedure is necessary. A novel confidence bound and interval procedure for estimating quantile ratios developed in this paper fills the void. Justified by rigorous large‐sample theory and a simulation study of small‐sample performance, the proposed method performs well in a wide range of scenarios involving right‐skewed distributions. By providing rigorous inference and much more interpretable statistics that account for experimental noise, the proposed method improves the current practice of analyzing drug activity data in xenograft experiments. The proposed method is fully nonparametric, simple to compute, performs equally well or better than known nonparametric methods, and is applicable to any statistical inference of a ‘fold change’ that can be formulated as a quantile ratio. Copyright © 2010 John Wiley & Sons, Ltd.     

Further analysis of the theoretical effectiveness of the TwoDay method of family planning.

This article validates the theoretical effectiveness of a simple approach to identify the fertile window of the menstrual cycle. The TwoDay method identifies all days in the cycle in which the woman notices cervical secretions, and the days immediately following these days, as the period in which the woman should consider herself fertile. Women who use this method are counseled to avoid unprotected intercourse on these days. The theoretical effectiveness of the TwoDay method was tested previously by applying the method rules to the menstrual cycles of women from a large data set from the World Health Organization (WHO). For the current study, we administered the same analysis to a data set from an Italian Ovulation Method center. These data are better suited for the analysis than were the WHO data because they identify all days with secretions. Results suggest that the method can be highly effective in helping women to identify correctly the days on which they should avoid unprotected intercourse if they do not wish to become pregnant, although some users may identify a few days as fertile that actually are not.     

Analysis of the degree of precision of a computer-assisted dietary survey method

A computer-assisted method for the registration of food intakes in real time according to a concise and simple procedure was subjected to a series of controls in order to assess precision. The method employs the 'portion' of a composite dish as unit of measure. The constancy of the portion was tested. The results show that the recipes of the diet of a subject may be stored and utilized in deferred time without loss of precision: data to be recorded in real time are then very limited. The length of period of analysis was also studied. The time of investigation should not be shorter than two weeks to obtain precise information on the feeding habits of an individual subject, whereas for a group of subjects the diary of a single day provides information of sufficient precision.     

Data generation for the Cox proportional hazards model with time‐dependent covariates: a method for medical researchers

The proliferation of longitudinal studies has increased the importance of statistical methods for time‐to‐event data that can incorporate time‐dependent covariates. The Cox proportional hazards model is one such method that is widely used. As more extensions of the Cox model with time‐dependent covariates are developed, simulations studies will grow in importance as well. An essential starting point for simulation studies of time‐to‐event models is the ability to produce simulated survival times from a known data generating process. This paper develops a method for the generation of survival times that follow a Cox proportional hazards model with time‐dependent covariates. The method presented relies on a simple transformation of random variables generated according to a truncated piecewise exponential distribution and allows practitioners great flexibility and control over both the number of time‐dependent covariates and the number of time periods in the duration of follow‐up measurement. Within this framework, an additional argument is suggested that allows researchers to generate time‐to‐event data in which covariates change at integer‐valued steps of the time scale. The purpose of this approach is to produce data for simulation experiments that mimic the types of data structures applied that researchers encounter when using longitudinal biomedical data. Validity is assessed in a set of simulation experiments, and results indicate that the proposed procedure performs well in producing data that conform to the assumptions of the Cox proportional hazards model. Copyright © 2013 John Wiley & Sons, Ltd.     

Dose estimate of exposure to radioisotopes in molecular and cellular biology.

A method for prospectively evaluating the annual equivalent doses and effective dose to biomedical researchers working with unsealed radioisotopes, and their classification, is presented here. Simplified formulae relate occupational data to a reasonable overestimate of the annual effective dose, and the equivalent doses to the hands and to the skin. The procedure, up to the classification of personnel and laboratories, can be made fully automatic, using a common spreadsheet on a personal computer. The method is based on occupational data, accounting for the amounts of each radioisotope used by a researcher, the time of exposure and the overall amounts employed in the laboratories where experiments are performed. The former data serve to forecast a contribution to the dose arising from a researcher's own work, the latter to a forecast of an 'environmental' contribution deriving simply from the presence in a laboratory where other people are working with radioisotopes. The estimates of the doses due to one's own radioisotope handling and to 'environment' were corrected for accidental exposure, considered as a linear function of the manipulated activity or of the time spent in the laboratories respectively, and summed up to give the effective dose. The effective dose associated with some common experiments in molecular and cellular biology is pre-evaluated by this method.     

Robust estimation of the causal effect of time-varying neighborhood factors on health outcomes

The fundamental difficulty of establishing causal relationships between an exposure and an outcome in observational data involves disentangling causality from confounding factors. This problem underlies much of neighborhoods research, which abounds with studies that consider associations between neighborhood characteristics and health outcomes in longitudinal data. Such analyses are confounded by selection issues; individuals with above average health outcomes (or associated characteristics) may self-select into advantaged neighborhoods. Techniques commonly used to assess causal inferences in observational longitudinal data, such as inverse probability of treatment weighting (IPTW), may be inappropriate in neighborhoods data due to unique characteristics of such data. We advance the IPTW toolkit by introducing a procedure based on a multivariate kernel density function which is more appropriate for neighborhoods data. The proposed weighting method is applied in conjunction with a marginal structural model. Our empirical analyses use longitudinal data from the Health and Retirement Study; our exposure of interest is an index of neighborhood socioeconomic status (NSES), and we examine its influence on cognitive function. Our findings illustrate the importance of the choice of method for IPTW—the comparison weighting methods provide poor balance across the set of covariates (which is not the case for our preferred procedure) and yield misleading results when applied in the outcomes models. The utility of the multivariate kernel is also validated via simulation. In addition, our findings emphasize the importance of IPTW—controlling for covariates within a regression without IPTW indicates that NSES affects cognition, whereas IPTW-weighted models fail to show a statistically significant effect.     

Evaluation of the deuterium dilution technique against the test-weighing procedure for the determination of breast milk intake

A noninvasive isotopic method for the determination of milk intake in breast-fed infants was evaluated against the conventional method of test-weighing. In experiment 1, 48-h breast milk intake was estimated concurrently by the test-weighing procedure and by the deuterium dilution technique. In experiment 2, the isotopic method, modified to minimize fluctuations in water flux, was evaluated against a 24-h test-weighing. In experiment 1, the mean 48-h milk intake estimated by the isotopic method (1616 +/- 353 ml) was significantly higher than that measured by the test-weighing procedure (1449 +/- 234 ml) (p less than 0.01). In experiment 2, 24-h milk intake as determined by test-weighing and deuterium dilution averaged 878.0 +/- 188.1 and 690.7 +/- 141.4 ml/day, respectively, and differed significantly (p less than 0.001). The ability of the deuterium dilution method to predict values obtained by the test-weighing procedure was unsatisfactory for individual estimations. These experiments indicate that the deuterium dilution technique is unacceptable for the determination of breast-milk intake in individuals, but may be satisfactory for a population estimation.     

Evaluation of procedures for hygienic hand-disinfection: controlled parallel experiments on the Vienna test model

Controlled parallel experiments were performed on the Vienna test model for the evaluation of procedures for hygienic hand-disinfection in three laboratories (Vienna, Mainz, Birmingham). The degerming activity of four procedures, each taking 1 min, was assessed repeatedly and compared with that of a standard disinfection procedure (ST) using isopropanol 60% (v/v). The mean log reductions (mean log RF) for each procedure were as follows: n-propanol 50% (v/v) 4.85 and 5.14 in Vienna (V) and Mainz (M) respectively, ethanol 70% (v/v) + chlorhexidine-gluconate 0.5% (w/v), 4.01 (V), 3.76 (M) and 4.00 in Birmingham (B). Washing procedures were less effective, mean log RF's of 3.19 (V), 3.49 (M) and 3.04 (B) were obtained with povidone-iodine soap, and 2.91 (V), 3.37 (M) and 3.27 (B) with a liquid phenolic soap. Analysis of variance on the data from Vienna and Mainz revealed significant differences of means not only between procedures ('preparations') but also on repeat testing. To compensate for the influence of variables such as test subjects, laboratory and day, the Vienna test model provides a method of standardization by testing a ST in parallel with the test procedure (P). Standardization of the results was obtained by pair-wise substraction, log RFPi-log RFSTi. Analysis of variance on the resulting values demonstrated that comparability of the results between laboratories and on repeat testing was achieved. The relative variation of the measurements within the laboratories ranged from 0.9 to 4.2%. As assessed by power-analysis, a disinfection procedure will be detected as significantly (P = 0.1) inferior to the standard processes in 95 of 100 experiments if it produces a mean log RF that is at least 0.55-0.65 log units smaller than that of the standard.     

Statistical properties and inference of the antimicrobial MIC test

A common method for measuring the drug-specific minimum inhibitory concentration (MIC) of an antibacterial agent is via a two-fold broth dilution test known as the MIC test. Because this procedure implicitly rounds data upward, inference based on unadjusted measurements is biased and overestimates bacterial resistance to a drug. We detail this test procedure and its associated bias, which, in many cases, has an expected value of approximately 0.5 on the log(2) scale. In addition, new bias-corrected estimates of resistance are proposed. A numeric example is used to illustrate the extent to which the traditional resistance estimate can overestimate the true proportion of resistant strains, a phenomenon which is remedied by using the proposed estimates.     

A validation study on a method to estimate nutrient intake by family members through a household-based food-weighing survey

The aim of the present study is to investigate the validity of a new method to estimate the food intake of individual subjects by a household-based dietary survey. The new method is based on the combination of household-based food weighing and approximating the proportions by which family members shared each dish or food in the household, which has been one of the components of the National Nutrition Survey, Japan, since 1995. We analyzed two sets of data from 64 volunteers (female students taking a dietitian course and their mothers) in 32 households by the approximated proportion method (method A) and the individual-based food weighing method (B) as a reference measurement. Energy and macronutrient intake by individual subjects estimated by method A was highly correlated to the corresponding values by method B (Pearson's correlation coefficients: r=0.90-0.92). Average energy intake was likely to be underestimated by method A compared with method B, being lower by 94 kcal (6.2% of the reference value in method B), at least in the young and middle-aged female adults that were the subjects of this study. When intake of boiled rice was separately analyzed, underestimation of energy intake by method A was 44 kcal, which contributed to approximately 50% of the total magnitude of the underestimation. The procedure manual for the National Nutrition Survey requests the participants to individually weigh the amount of boiled rice taken by family members in the household. However, this procedure is not observed in many actual settings. Therefore, following this procedure would be an effective measure to improve the accuracy of the dietary data.     

Handwipe sampling and analysis procedure for the measurement of dermal contact with pesticides

A handwipe sampling and analysis procedure was developed for the measurement of dermal contact to pesticides. This procedure utilizes cellulose dressing sponges wetted with 2-propanol. A two-step wiping procedure is described that ensures that the entire hand is sampled. Removal efficiency experiments show that dry residues of the pesticides chlorpyrifos and pyrethrin I are quantitatively removed from hands immediately following contact. Results suggest that the procedure may remove pesticide residues that are deeply embedded in the skin and not removed by soap-and-water washing. Extraction efficiency studies for 29 other pesticides show that the proposed extraction method may be applicable for a wide range of pesticides including phenoxy-acid herbicides. Field testing of the procedure indicates that it is easily implemented by sampling personnel and readily accepted by children.