Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Statistical Power”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 865 records · Page 48Linked to original sources
PubMed1986

Determination of bioequivalence of psychotropic drugs and concerns involving product interchange.

There is a growing debate among researchers and practitioners concerning the validity of the Food and Drug Administration (FDA) standards for establishing generic interchange through assignment of "therapeutically equivalent" designations for noninnovator or generic drugs. This debate has particular significance for psychotropic drugs which are used in the management of patients with severely debilitating mental diseases. Thus, the controversy primarily focuses on the appropriateness of using the FDA's therapeutic equivalence designation as a criterion for interchange. This is particularly important in light of the lack of well-defined standards for bioequivalence studies and in view of the effect of mandatory substitution laws and financial incentives that encourage generic dispensing that can lead to frequent and indiscriminate interchange among multiple generic brands. Specific concerns include the validity of assay techniques for drug in biologic fluids, statistical power analysis, the appropriateness of the "70/70 rule," and the relevance of studies carried out in healthy normal volunteers in the determination of bioequivalence.

Antipsychotic Agents↗
PubMed1998

Goodness-of-fit tests for GEE modeling with binary responses.

Analysis of data with repeated measures is often accomplished through the use of generalized estimating equations (GEE) methodology. Although methods exist for assessing the adequacy of the fitted models for uncorrelated data with likelihood methods, it is not appropriate to use these methods for models fitted with GEE methodology. We propose model-based and robust (empirically corrected) goodness-of-fit tests for GEE modeling with binary responses based on partitioning the space of covariates into distinct regions and forming score statistics that are asymptotically distributed as chi-square random variables with the appropriate degrees of freedom. The null distribution and the statistical power of the proposed goodness-of-fit tests were assessed using simulated data. The proposed goodness-of-fit tests are illustrated by two examples using data from clinical studies.

Antidepressive Agents↗
PubMedDate not supplied

Genetic dissection of age-related changes of immune function in mice.

Understanding of the genetic basis of normal and abnormal development of the immune response is an enormous undertaking. The immune response, at the most minimal level, involves interactions of antigen presenting cells (APCs), T and B cells. Each of these cells produce cell surface and soluble factors (cytokines) that affect both autocrine and paracrine functions. A second level of complexity needs to consider the development of the macrophage/monocyte lineage as well as the production of the common lymphoid precursor which undergoes distinct maturation steps in the thymus and periphery to form mature T cells as well as in BM (BM) and lymphoid organs to form mature B cells. A third level of complexity involves the immune response to infectious agents including viruses and also the response to tumour antigens. In addition, there are imbalances that predispose to decreased responses (immunodeficiencies) or increased responses (autoimmunity). A fourth level of complexity involves attempts to understand the differences in the immune response that occurs at a very young age, in adults, and at a very old age. This review will focus on the use of C57BL/6 J X DBA/2 J (BXD) recombinant inbred (RI) strains of mice to map genetic loci associated with the production of lymphoid precursors in the BM, development of T cells in the thymus, and T-cell responses to stimulation in the peripheral lymphoid organs in adult and in aged mice. Strategies to improve the power and precision in which complex traits such as the age-related immune response can be mapped is limited with the current set of 35 strains of BXD mice. Strategies to increase these strains by generating recombinant intercross (RIX) strains of mice are being developed to enable this large set of lines to detect quantitative trait loci (QTLs) with a much higher consistency and statistical power. More importantly, the resolution with which these QTLs can be mapped would be greatly improved and, in many cases, adequate to carry out direct identification of candidate genes. It is likely that, given the complexity of the immune system development, the number of cells involved in an immune response, and especially the changes in the immune system with ageing, mapping hundreds of genes will be required to fully understand age-related changes in the immune response. This review outlines ongoing and future strategies that will enable the mapping and identification of these genes.

Aging↗
PubMed2004

Assessing the probability that a positive report is false: an approach for molecular epidemiology studies.

Too many reports of associations between genetic variants and common cancer sites and other complex diseases are false positives. A major reason for this unfortunate situation is the strategy of declaring statistical significance based on a P value alone, particularly, any P value below.05. The false positive report probability (FPRP), the probability of no true association between a genetic variant and disease given a statistically significant finding, depends not only on the observed P value but also on both the prior probability that the association between the genetic variant and the disease is real and the statistical power of the test. In this commentary, we show how to assess the FPRP and how to use it to decide whether a finding is deserving of attention or "noteworthy." We show how this approach can lead to improvements in the design, analysis, and interpretation of molecular epidemiology studies. Our proposal can help investigators, editors, and readers of research articles to protect themselves from overinterpreting statistically significant findings that are not likely to signify a true association. An FPRP-based criterion for deciding whether to call a finding noteworthy formalizes the process already used informally by investigators--that is, tempering enthusiasm for remarkable study findings with considerations of plausibility.

Analysis of Variance↗
PubMed1995

Treating menstruating women with thrombolytic therapy: insights from the global utilization of streptokinase and tissue plasminogen activator for occluded coronary arteries (GUSTO-I) trial.

OBJECTIVES: The purpose of this study was to examine the clinical implications of administering thrombolytic therapy to menstruating women with acute myocardial infarction. BACKGROUND: Although anecdotal case reports have suggested that thrombolytic therapy is safe during menstruation, the risk of increased bleeding in menstruating women receiving such therapy is poorly defined. METHODS: We identified menstruating women who received thrombolytic therapy by soliciting information on all North American women enrolled in the GUSTO-I trial and then collected additional information about them with use of a one-page data form. We compared the characteristics and outcomes of these women with other GUSTO-I patient populations, including all North American women below the median age of menopause, all women and all patients. RESULTS: The median age of the 12 menstruating women was 46 years; 75% were cigarette smokers. The median hospital stay was 7 days, 2 fewer than the overall stay in GUSTO-I. None of these women died or had a stroke or severe bleeding. Three patients (25%) had moderate bleeding (vaginal in two patients [66%]) that required transfusion compared with 11% of all GUSTO-I patients and all North American premenopausal women (p = 0.13) and 17% of all female GUSTO-I patients (p = 0.47). Because of the small sample size of 12 women, the power was low (0.37) to detect the observed difference in moderate bleeding. The median nadir hematocrit was 33% in the menstruating women compared with 34% in the premenopausal women and all women. The median time from symptom onset to treatment for the 12 women was 3.7 h, which was 0.9 h longer than the overall median in the trial (p = 0.09). CONCLUSIONS: Although there was no statistically significant increase in bleeding risk during menstruation, this fact may be a result of low statistical power rather than a lack of effect. Thus, the results suggest that there may be a clinically significant increase in the risk of moderate bleeding. Nevertheless, the GUSTO-I experience is consistent with the concept that the lifesaving benefit of thrombolytic therapy for acute myocardial infarction should generally not be withheld because of active menstruation.

Aged↗
PubMed2001

1,1-Dichloro-2,2-bis(p-chlorophenyl)ethylene and polychlorinated biphenyls and breast cancer: combined analysis of five U.S. studies.

BACKGROUND: Environmental exposure to organochlorines has been examined as a potential risk factor for breast cancer. In 1993, five large U.S. studies of women located mainly in the northeastern United States were funded to evaluate the association of levels of 1,1-dichloro-2,2-bis(p-chlorophenyl) ethylene (DDE) and polychlorinated biphenyls (PCBs) in blood plasma or serum with breast cancer risk. We present a combined analysis of these results to increase precision and to maximize statistical power to detect effect modification by other breast cancer risk factors. METHODS: We reanalyzed the data from these five studies, consisting of 1400 case patients with breast cancer and 1642 control subjects, by use of a standardized approach to control for confounding and assess effect modification. We calculated pooled odds ratios (ORs) and 95% confidence intervals (CIs) by use of the random-effects model. All statistical tests were two-sided. RESULTS: When we compared women in the fifth quintile of lipid-adjusted values with those in the first quintile, the multivariate pooled OR for breast cancer associated with PCBs was 0.94 (95% CI = 0.73 to 1.21), and that associated with DDE was 0.99 (95% CI = 0.77 to 1.27). Although in the original studies there were suggestions of elevated breast cancer risk associated with PCBs in certain groups of women stratified by parity and lactation, these observations were not evident in the pooled analysis. No statistically significant associations were observed in any other stratified analyses, except for an increased risk with higher levels of PCBs among women in the middle tertile of body mass index (25-29.9 kg/m(2)); however, the risk was statistically nonsignificantly decreased among heavier women. CONCLUSIONS: Combined evidence does not support an association of breast cancer risk with plasma/serum concentrations of PCBs or DDE. Exposure to these compounds, as measured in adult women, is unlikely to explain the high rates of breast cancer experienced in the northeastern United States.

Body Weight↗
PubMed2006

Review of microarray experimental design strategies for genetical genomics studies.

Genetical genomics approaches provide a powerful tool for studying the genetic mechanisms governing variation in complex traits. By combining information on phenotypic traits, pedigree structure, molecular markers, and gene expression, such studies can be used for estimating heritability of mRNA transcript abundances, for mapping expression quantitative trait loci (eQTL), and for inferring regulatory gene networks. Microarray experiments, however, can be extremely costly and time consuming, which may limit sample sizes and statistical power. Thus it is crucial to optimize experimental designs by carefully choosing the subjects to be assayed, within a selective profiling approach, and by cautiously controlling systematic factors affecting the system. Also, a rigorous strategy should be used for allocating mRNA samples across assay batches, slides, and dye labeling, so that effects of interest are not confounded with nuisance factors. In this presentation, we review some selective profiling strategies for genetical genomics studies, including the selection of individuals for increased genetic dissimilarity and for a higher number of recombination events. Efficient designs for studying epistasis are also discussed, as well as experiments for inferring heritability of transcriptional levels. It is shown that solving an optimal design problem generally requires a numerical implementation and that the optimality criteria should be intimately related to the goals of the experiment, such as the estimation of additive, dominance, and interacting effects, localizing putative eQTL, or inferring genetic and environmental variance components associated with transcriptional abundances.

Animals↗
PubMed2003

Concurrent strength and endurance training: the influence of dependent variable selection.

Twenty-six active university students were randomly allocated to resistance (R, n = 9), endurance (E, n = 8), and concurrent resistance and endurance (C, n = 9) training conditions. Training was completed 3 times per week in all conditions, with endurance training preceding resistance training in the C group. Resistance training involved 4 sets of upper- and lower-body exercises with loads of 4-8 repetition maximum (RM). Each endurance training session consisted of five 5-minute bouts of incremental cycle exercise at between 40 and 100% of peak oxygen uptake (.VO2peak). Parameters measured prior to and following training included strength (1RM and isometric and isokinetic [1.04, 3.12, 5.20, and 8.67 rad.s(-1)] strength), .VO2peak and Wingate test performance (peak power output [PPO], average power, and relative power decline). Significant improvements in 1RM strength were observed in the R and C groups following training. .VO2peak significantly increased in E and C but was significantly reduced in R after training. Effect size (ES) transformations on the other dependent variables suggested that performance changes in the C group were not always similar to changes in the R or E groups. These ES data suggest that statistical power and dependent variable selection are significant issues in enhancing our insights into concurrent training. It may be necessary to assess a range of performance parameters to monitor the relative effectiveness of a particular concurrent training regimen.

Adolescent↗
PubMed2004

Linkage disequilibrium as a signature of selective sweeps.

The hitchhiking effect of a beneficial mutation, or a selective sweep, generates a unique distribution of allele frequencies and spatial distribution of polymorphic sites. A composite-likelihood test was previously designed to detect these signatures of a selective sweep, solely on the basis of the spatial distribution and marginal allele frequencies of polymorphisms. As an excess of linkage disequilibrium (LD) is also known to be a strong signature of a selective sweep, we investigate how much statistical power is increased by the inclusion of information regarding LD. The expected pattern of LD is predicted by a genealogical approach. Both theory and simulation suggest that strong LD is generated in narrow regions at both sides of the location of beneficial mutation. However, a lack of LD is expected across the two sides. We explore various ways to detect this signature of selective sweeps by statistical tests. A new composite-likelihood method is proposed to incorporate information regarding LD. This method enables us to detect selective sweeps and estimate the parameters of the selection model better than the previous composite-likelihood method that does not take LD into account. However, the improvement made by including LD is rather small, suggesting that most of the relevant information regarding selective sweeps is captured by the spatial distribution and marginal allele frequencies of polymorphisms.

Computer Simulation↗
PubMed1989

Potential and pitfalls of randomized clinical trials in cancer research.

Randomized clinical trials are a powerful tool to establish the superiority of investigational therapies over standard ones. Yet in cancer, particularly in colorectal cancer, the results of clinical trials have been disappointing and have raised some suspicion about the real worth of randomization. This paper attempts to show that the results of randomized trials have often been misinterpreted, either because they were subject to various sources of biases or because their statistical power was too low to allow a reasonably reliable assessment of true treatment benefits.

Bias↗
PubMed1997

Fully parametric and semi-parametric regression models for common events with covariate measurement error in main study/validation study designs.

The derivation of the likelihood function for binary data from two types of main study/validation study designs where model covariates are measured with error is elaborated. Rather than limiting consideration to a restricted family of models with convenient mathematical properties, we suggest that empirical considerations, customized to the data at hand, should drive model choices. The joint likelihood function for the main study, in which the covariates are measured with error, and the validation study, in which they are not, is maximized, and estimation and inference proceeds using standard theory. Although the choice of the measurement error model is driven by empirical considerations, the relatively small validation study sizes typically seen may lead to misspecification, resulting in bias in estimation and inference about exposure-disease relationships. By using a nonparametric form for the measurement error model, the resulting semi-parametric methods suggested by Robins, Rotnitzky, and Zhao (1994, Journal of the American Statistical Association 89, 864-866) and Robins, Hsieh, and Newey (1995, Journal of the Royal Statistical Society, Series B 57, 409-424) are free from bias due to misspecification of the measurement error model, trading efficiency for robustness as usual. These fully and semi-parametric methods are illustrated with a detailed example from a main study/validation study of the health effects of occupational exposure to chemotherapeutics among pharmacists (Valanis et al., 1993, American Journal of Hospital Pharmacy 50, 455-462). A constant, prevalence ratio model for common binary events, with gamma covariate measurement error, is derived and empirically verified by the available data. A careful reanalysis of the data, taking measurement error fully into account, leads to a threefold increase in the log relative risk and no loss of statistical power. The semi-parametric estimates are consistent with the parametric results, providing reassurance that important bias due to misspecification of the measurement error model is unlikely.

Analysis of Variance↗
PubMed1996

The effect of arthritis in the carpal joint on performance in Norwegian cold-blooded trotters.

The purpose of this study was to evaluate the impact of arthritis of the carpal joint on performance of Norwegian cold-blooded trotters. Two performance variables were used in the analyses. The first was the start status, for which horses that had started in one or more races within a certain age received the value 1, and horses that had not raced were correspondingly assigned the value 0. The second variable was the accumulated, transformed and standardized earnings (ATSE), which is the power transformation of earnings (earnings .2, with unraced horses assigned a value of zero) expressed as a standardized normal deviate by birth year. With the exception of the first year of racing, the number of horses that had raced was larger in the group of unaffected horses than in the groups with arthritis (carpitis or bilateral carpitis), although the difference between the groups was not significant for any of the age classes. A similar picture was observed for ATSE and, in general, the diseased horses earned less money. None of these differences was significant at the 5% level. However, the statistical power was less than 0.3, which means that the probability of detecting a true difference was less than 30%. The data were deemed inadequate to show a significant effect of arthritis on racing performance. This may only be achieved through investigations in which more of the error variance can be statistically modelled, and in which arthritis can be observed as an incidence rather than as a prevalence.

Age Factors↗
PubMed1999

An efficient design for verifying disease outcome status in large cohorts with rare exposures and low disease rates.

Cohort studies require the use of large samples when the risk of the event is very low. Databases that are large and population-based, such as Medicaid files, are frequently used for cohort studies, since they provide access to the large samples required for adequate statistical power at a relatively affordable cost. Epidemiologic studies using these databases typically require verification of reported diagnoses, however, because of the potential for errors in disease reporting. When exposure prevalence is also low, as in many pharmacoepidemiologic investigations of drug toxicity, there are few exposed cases compared to the number of unexposed cases. Verification of all unexposed presumptive cases through medical records is costly. We investigate the statistical efficiency of a design in which all exposed cases but only a subsample of the unexposed cases are verified. We show that good efficiency can usually be achieved with a small subsample of unexposed cases. Published in 1999 by John Wiley & Sons, Ltd.

Anti-Inflammatory Agents, Non-Steroidal↗
PubMed1994

Alcoholic beverages and breast cancer: some observations on published case-control studies.

We identified 38 case-control studies investigating possible associations between alcoholic beverage consumption and cancer of the female breast. Each study was characterized according to design features such as: control type (hospital or community based), risk factors controlled for, matching strategy, and statistical power. We examined the effect of these design variables on several outcome variables including identification of any significant elevation in odds ratio and characterization of any dose-response effect. The major finding of this study is that of a striking difference between hospital and community based controlled studies with respect to (1) the level of any estimated dose-response effect, and (2) the finding of statistically significant elevations in odds ratios at levels of consumption below 4 drinks per week. In summary, the generally weak associations reported in these case-control studies along with the measurement and/or selection biases implied by our findings would lead one to the conclusion that present evidence does not support a causal association. This conclusion seems to be in accord with results from cohort studies and with similar conclusions from several other reviews.

Alcohol Drinking↗
PubMed2005

Design and analysis of phase III trials with ordered outcome scales: the concept of the sliding dichotomy.

The conventional approach to the analysis of a Phase III trial in head injury or stroke takes an ordered scale measuring functional outcome and collapses the scale to a binary outcome of favorable versus unfavorable. This discards potentially relevant information which limits statistical power and moreover is not in accord with clinical practice. We propose an alternative approach where a favorable outcome is defined as better than would be expected, taking account of each individual patient's baseline prognosis. This is illustrated through a worked example based on data from a Phase III trial in head injury. The approach is also compared with the proportional odds model, which is another statistical approach that can exploit an ordered outcome scale. The approach raises issues of clinical, statistical, and regulatory importance, and we initiate what we believe needs to become a widespread debate amongst the community involved in clinical research in head injury and stroke.

Clinical Trials, Phase III as Topic↗
PubMed2006

Novel knowledge-based mean force potential at the profile level.

BACKGROUND: The development and testing of functions for the modeling of protein energetics is an important part of current research aimed at understanding protein structure and function. Knowledge-based mean force potentials are derived from statistical analyses of interacting groups in experimentally determined protein structures. Current knowledge-based mean force potentials are developed at the atom or amino acid level. The evolutionary information contained in the profiles is not investigated. Based on these observations, a class of novel knowledge-based mean force potentials at the profile level has been presented, which uses the evolutionary information of profiles for developing more powerful statistical potentials. RESULTS: The frequency profiles are directly calculated from the multiple sequence alignments outputted by PSI-BLAST and converted into binary profiles with a probability threshold. As a result, the protein sequences are represented as sequences of binary profiles rather than sequences of amino acids. Similar to the knowledge-based potentials at the residue level, a class of novel potentials at the profile level is introduced. We develop four types of profile-level statistical potentials including distance-dependent, contact, Phi/Psi dihedral angle and accessible surface statistical potentials. These potentials are first evaluated by the fold assessment between the correct and incorrect models generated by comparative modeling from our own and other groups. They are then used to recognize the native structures from well-constructed decoy sets. Experimental results show that all the knowledge-base mean force potentials at the profile level outperform those at the residue level. Significant improvements are obtained for the distance-dependent and accessible surface potentials (5-6%). The contact and Phi/Psi dihedral angle potential only get a slight improvement (1-2%). Decoy set evaluation results show that the distance-dependent profile-level potentials even outperform other atom-level potentials. We also demonstrate that profile-level statistical potentials can improve the performance of threading. CONCLUSION: The knowledge-base mean force potentials at the profile level can provide better discriminatory ability than those at the residue level, so they will be useful for protein structure prediction and model refinement.

Algorithms↗
PubMed2006

Dexamethasone in adults with community-acquired bacterial meningitis.

Bacterial meningitis in adults is a severe disease with high fatality and morbidity rates. Experimental studies have shown that the inflammatory response in the subarachnoid space is associated with an unfavourable outcome. In these experiments, corticosteroids, and in particular dexamethasone, were able to reduce the inflammatory cascades in the subarachnoid space. The use of corticosteroids as adjunctive therapy in adults with bacterial meningitis has been evaluated in six studies, performed over a time period of 40 years. Most studies on adjunctive dexamethasone therapy in adults with bacterial meningitis were limited by methodological flaws. In 2002, a study with sufficient statistical power to show significant differences was published. This European Dexamethasone Study showed that adjunctive dexamethasone therapy reduced the rate of unfavourable outcomes in adults with bacterial meningitis from 25% to 15%. In this study, adjunctive treatment with dexamethasone was given before or with the first dose of antibacterials, without serious adverse effects. A quantitative review showed a consistent beneficial effect of dexamethasone on mortality and a borderline statistical beneficial effect on neurological sequelae. On the basis of the available evidence, adjunctive dexamethasone therapy should be initiated before or with the first dose of antibacterials and continued for 4 days in all adults with suspected or proven bacterial meningitis, regardless of bacterial aetiology. In patients with both meningitis and septic shock, dexamethasone therapy cannot be unequivocally recommended, but the use of lower doses seems reasonable at present. Since prompt use of dexamethasone and appropriate antibacterials improves the prognosis of adults with bacterial meningitis, hospitals will require protocols to include dexamethasone with the initial antibacterial therapy.

Adult↗
PubMed2003

Comparison of rheumatoid arthritis clinical trial outcome measures: a simulation study.

OBJECTIVE: Isolated studies have suggested that continuous measures of response may be better than predefined, dichotomous definitions (e.g., the American College of Rheumatology 20% improvement criteria [ACR20]) for discriminating between rheumatoid arthritis (RA) treatments. Our goal was to determine the statistical power of predefined dichotomous outcome measures (termed "a priori"), compared with that of continuous measures derived from trial data in which there was no predefined response threshold (termed "data driven"), and to evaluate the sensitivity to change of these measures in the context of different treatments and early versus later-stage disease. In order to generalize beyond results from a single trial, we performed simulation studies. METHODS: We obtained summary data from trials comparing disease-modifying antirheumatic drugs (DMARDs) and from comparative coxib-placebo trials to test the power of 2 a priori outcomes, the ACR20 and improvement of the Disease Activity Score (DDAS), as well as 2 data-driven outcomes. We studied patients with early RA and those with later-stage RA (duration of <4 years and 4-9 years, respectively). We performed simulation studies, using the interrelationship of ACR core set measures in the trials to generate multiple trial data sets consistent with the original data. RESULTS: The data-driven outcomes had greater power than did the a priori measures. The DMARD comparison was more powerful in early disease than in later-stage disease (the sample sizes needed to achieve 80% power for the most powerful test were 64 for early disease versus 100 for later disease), but the coxib-versus-placebo comparison was less powerful in early disease than in later disease (the sample sizes needed to achieve 80% power were 200 and 100, respectively). When the effects of treatment on core set items were small and/or inconsistent, power was reduced, particularly for a less broadly based outcome (e.g., DDAS) compared with the ACR20. CONCLUSION: The simulation studies demonstrate that data-driven outcome definitions can provide better sensitivity to change than does the ACR20 or DDAS. Using such methods would improve power, but at the expense of trial standardization. The studies also show how patient population and treatment characteristics affect the power of specific outcome measures in RA clinical trials, and provide quantification of those effects.

Antirheumatic Agents↗