SEARCH · PubMed Health
Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Explore the source record for details and available documents.
The possibility of reduction of numbers of fresh coumarin and normal plasmas has been studied in a multicentre manual prothrombin (PT) calibration of high international sensitivity index (ISI) rabbit and low ISI human reference thromboplastins at 14 laboratories. The number of calibrant plasmas was reduced progressively by a computer program which generated random numbers to provide 1000 different selections for each reduced sample at each participant laboratory. Results were compared with those of the full set of 20 normal and 60 coumarin plasma calibrations. With the human reagent, 20 coumarins and seven normals still achieved the W.H.O. precision limit (3% CV of the slope), but with the rabbit reagent reduction < 50 coumarins with 17 normal plasmas led to unacceptable CV. Little reduction of numbers from the full set of 80 fresh plasmas appears advisable. For maximum confidence, when calibrating the ISI of a new reagent, it would always seem worthwhile to use the full number of patients' and normal plasmas at each centre for the initial calibration, i.e. 60 fresh coumarin and 20 fresh normal plasmas. Smaller numbers should achieve the required precision with the lower ISI reagent.
Explore the source record for details and available documents.
We studied the design effects for population-based surveys that estimated the prevalence of wasting and stunting malnutrition in Malawi, Zambia, Indonesia, and Nepal, and studied the magnitude of different types of malnutrition clustering within villages. Weight, height, and midupper-arm circumference were measured on all children or on systematic samples of children in randomly selected villages. Design effects ranged from 0.53 for low height-for-age in Zambia to 6.12 for low weight-for-age in Nepal. If all sampled clusters were of size 30, as is often the case for nutrition surveys, design effects would have ranged from 0.44 for low height-for-age in Zambia to 2.59 for low midupper-arm circumference in Zambia. Malnutrition did cluster within villages. Stunting malnutrition clustered less than did wasting malnutrition. Nutrition surveys using clusters of 30 can sample fewer clusters than currently recommended if basic prevalence and cluster information are available prior to sample selection.
Explore the source record for details and available documents.
Proper randomisation means little if investigators cannot include all randomised participants in the primary analysis. Participants might ignore follow-up, leave town, or take aspartame when instructed to take aspirin. Exclusions before randomisation do not bias the treatment comparison, but they can hurt generalisability. Eligibility criteria for a trial should be clear, specific, and applied before randomisation. Readers should assess whether any of the criteria make the trial sample atypical or unrepresentative of the people in which they are interested. In principle, assessment of exclusions after randomisation is simple: none are allowed. For the primary analysis, all participants enrolled should be included and analysed as part of the original group assigned (an intent-to-treat analysis). In reality, however, losses frequently occur. Investigators should, therefore, commit adequate resources to develop and implement procedures to maximise retention of participants. Moreover, researchers should provide clear, explicit information on the progress of all randomised participants through the trial by use of, for instance, a trial profile. Investigators can also do secondary analyses on, for instance, per-protocol or as-treated participants. Such analyses should be described as secondary and non-randomised comparisons. Mishandling of exclusions causes serious methodological difficulties. Unfortunately, some explanations for mishandling exclusions intuitively appeal to readers, disguising the seriousness of the issues. Creative mismanagement of exclusions can undermine trial validity.
We develop a new statistic for testing the equality of two multivariate mean vectors. A scaled chi-squared distribution is proposed as an approximating null distribution. Because the test statistic is based on componentwise statistics, it has the advantage over Hotelling's T2 test of being applicable to the case where the dimension of an observation exceeds the number of observations. An appealing feature of the new test is its ability to handle missing data by relying on only componentwise sample moments. Monte Carlo studies indicate good power compared to Hotelling's T2 and a recently proposed test by Srivastava (2004, Technical Report, University of Toronto). The test is applied to drug discovery data.
Explore the source record for details and available documents.
A logistic regression model may be used to provide predictions of outcome for individual patients at another centre than where the model was developed. When empirical data are available from this centre, the validity of predictions can be assessed by comparing observed outcomes and predicted probabilities. Subsequently, the model may be updated to improve predictions for future patients. As an example, we analysed 30-day mortality after acute myocardial infarction in a large data set (GUSTO-I, n = 40 830). We validated and updated a previously published model from another study (TIMI-II, n = 3339) in validation samples ranging from small (200 patients, 14 deaths) to large (10,000 patients, 700 deaths). Updated models were tested on independent patients. Updating methods included re-calibration (re-estimation of the intercept or slope of the linear predictor) and more structural model revisions (re-estimation of some or all regression coefficients, model extension with more predictors). We applied heuristic shrinkage approaches in the model revision methods, such that regression coefficients were shrunken towards their re-calibrated values. Parsimonious updating methods were found preferable to more extensive model revisions, which should only be attempted with relatively large validation samples in combination with shrinkage.
We have compared microarray data generated on Affymetrix chips from standard (8 micrograms) or low (100 nanograms) amounts of total RNA. We evaluated the gene signals and gene fold-change estimates obtained from the two methods and validated a subset of the results by real time, polymerase chain reaction assays. The correlation of low RNA derived gene signals to gene signals obtained from standard RNA was poor for less to moderately abundant genes. Genes with high abundance showed better correlation in signals between the two methods. The signal correlation between the low RNA and standard RNA methods was improved by including a reference sample in the microarray analysis. In contrast, the fold-change estimates for genes were better correlated between the two methods regardless of the magnitude of gene signals. A reference sample based method is suggested for studies that would end up comparing gene signal data from a combination of low and standard RNA templates; no such referencing appears to be necessary when comparing fold-changes of gene expression between standard and low template reactions.
In many practical classification problems it is important to distinguish false positive from false negative results when evaluating the performance of the classifier. This is of particular importance for medical diagnostic tests. In this context, receiver operating characteristic (ROC) curves have become a standard tool. Here we apply this concept to characterize the performance of a simple neural network. Investigating the binary classification of a perceptron we calculate analytically the shape of the corresponding ROC curves. The influence of the size of the training set and the prevalence of the quality considered are studied by means of a statistical-mechanics analysis.
We develop the idea of using data from the first 'few' patients entered in a clinical trial to estimate the final trial size needed to have specified power for rejecting H0 in favour of H1 if a real difference exists. When comparing means derived from Normally distributed data, there is no important effect on test size, power or expected trial size, provided that a minimum of about 20 degrees of freedom are used to estimate residual variance. Relative advantages and disadvantages of using larger internal pilot studies are presented. These revolve around crude expectations of the final study size, recruitment rate, duration of follow-up and practical constraints on the ability to prevent the circulation of unblinded randomization codes to investigators and those involved in editing and checking data.
Authentication systems based on biometric features (e.g., fingerprint impressions, iris scans, human face images, etc.) are increasingly gaining widespread use and popularity. Often, vendors and owners of these commercial biometric systems claim impressive performance that is estimated based on some proprietary data. In such situations, there is a need to independently validate the claimed performance levels. System performance is typically evaluated by collecting biometric templates from n different subjects, and for convenience, acquiring multiple instances of the biometric for each of the n subjects. Very little work has been done in 1) constructing confidence regions based on the ROC curve for validating the claimed performance levels and 2) determining the required number of biometric samples needed to establish confidence regions of prespecified width for the ROC curve. To simplify the analysis that address these two problems, several previous studies have assumed that multiple acquisitions of the biometric entity are statistically independent. This assumption is too restrictive and is generally not valid. We have developed a validation technique based on multivariate copula models for correlated biometric acquisitions. Based on the same model, we also determine the minimum number of samples required to achieve confidence bands of desired width for the ROC curve. We illustrate the estimation of the confidence bands as well as the required number of biometric samples using a fingerprint matching system that is applied on samples collected from a small population.
Diagnostic uncertainty, particularly prior to the receipt of test results, is common in medicine. When randomized trials are being conducted to evaluate therapies in these situations, patients must be randomized in the absence of a definitive diagnosis. This article explores some of the implications of diagnostic uncertainty for the design and analysis of randomized clinical trials, using specific therapeutic examples to illustrate the important points.
Molecular epidemiological association studies use valuable biosamples and incur costs. Statistical methods for early genotyping termination may conserve biosamples and costs. Group sequential methods (GSM) allow early termination of studies on the basis of interim comparisons. Simulation studies evaluated the application of GSM using data from a case-control study of GST genotypes and prostate cancer. Group sequential boundaries (GSB) were defined in the EAST-2000 software and were evaluated for study termination when early evidence suggested that the null hypothesis of no association between genotype and disease was unlikely to be rejected. Early termination of GSTM1 genotyping, which demonstrated no association with prostate cancer, occurred in >90% of the simulated studies. On average, 36.4% of biosamples were saved from unnecessary genotyping. In contrast, for GSTT1, which demonstrated a positive association, inappropriate termination occurred in only 6.6%. GSM may provide significant cost and sample savings in molecular epidemiology studies.
OBJECTIVE: To assess whether there are significant differences in ease of use and quality of samples among several bone biopsy needles currently available. DESIGN: Eight commonly used, commercially available bone biopsy needles of different gauges were evaluated. Each needle was used to obtain five consecutive samples from a lamb lumbar pedicle. Subjective assessment of ease of needle use, ease of sample removal from the needle and sample quality, before and after fixation, was graded on a 5-point scale. The number of attempts necessary to reach a 1 cm depth was recorded. Each biopsy specimen was measured in the gross state and after fixation. RESULTS: The RADI Bonopty 15 g and Kendall Monoject J-type 11 g needles were rated the easiest to use, while the Parallax Core-Assure 11 g and the Bard Ostycut 16 g were rated the most difficult. Parallax Core-Assure and Kendall Monoject needles had the highest quality specimen in the gross state; Cook Elson/Ackerman 14 g and Bard Ostycut 16 g needles yielded the lowest. The MD Tech without Trap-Lok 11 g needle had the highest quality core after fixation, while the Bard Ostycut 16 g had the lowest. There was a significant difference in pre-fixation sample length between needles (P<0.0001), despite acquiring all cores to a standard 1 cm depth. Core length and width decrease in size by an average of 28% and 42% after fixation. CONCLUSION: Bone biopsy needles vary significantly in performance. Detailed knowledge of the strengths and weaknesses of different needles is important to make an appropriate selection for each individual's practice.
A horizontal, fluorophore-enhanced, repetitive extragenic palindromic-PCR (rep-PCR) DNA fingerprinting technique (HFERP) was developed and evaluated as a means to differentiate human from animal sources of Escherichia coli. Box A1R primers and PCR were used to generate 2,466 rep-PCR and 1,531 HFERP DNA fingerprints from E. coli strains isolated from fecal material from known human and 12 animal sources: dogs, cats, horses, deer, geese, ducks, chickens, turkeys, cows, pigs, goats, and sheep. HFERP DNA fingerprinting reduced within-gel grouping of DNA fingerprints and improved alignment of DNA fingerprints between gels, relative to that achieved using rep-PCR DNA fingerprinting. Jackknife analysis of the complete rep-PCR DNA fingerprint library, done using Pearson's product-moment correlation coefficient, indicated that animal and human isolates were assigned to the correct source groups with an 82.2% average rate of correct classification. However, when only unique isolates were examined, isolates from a single animal having a unique DNA fingerprint, Jackknife analysis showed that isolates were assigned to the correct source groups with a 60.5% average rate of correct classification. The percentages of correctly classified isolates were about 15 and 17% greater for rep-PCR and HFERP, respectively, when analyses were done using the curve-based Pearson's product-moment correlation coefficient, rather than the band-based Jaccard algorithm. Rarefaction analysis indicated that, despite the relatively large size of the known-source database, genetic diversity in E. coli was very great and is most likely accounting for our inability to correctly classify many environmental E. coli isolates. Our data indicate that removal of duplicate genotypes within DNA fingerprint libraries, increased database size, proper methods of statistical analysis, and correct alignment of band data within and between gels improve the accuracy of microbial source tracking methods.
Formulae are derived for the required number of failures needed in a randomized clinical trial with a binary or survival endpoint in which treatments are compared using multiple units within each patient. Calculation of the required total numbers of units and patients are illustrated by examples. These show that naive use of conventional calculations could underestimate the number of units and patients required, particularly if the number of units per patient is small.