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 793 records · Page 44Linked to original sources
PubMed1993

"The Göteborg Quality of Life Instrument"--a psychometric evaluation of assessments of symptoms and well-being among women in a general population.

OBJECTIVE: To examine with psychometric analysis techniques the potential for constructing valid composite variables of "The Göteborg Quality of Life Instruments". DESIGN: Prospective population study of women in 1974-75 and in 1980-81. SETTING: City of Göteborg, Sweden. PARTICIPANTS: Representative samples of the general population of women in five age strata between 44 and 66 years of age in 1974-75, followed 1980-81 including new cohorts of women aged 26 and 38. 1302 women were examined in 1974-75 and 1408 in 1980-81. MAIN OUTCOME MEASURE: "The Göteborg Quality of Life instrument", constructed in the early 1970s to measure symptoms and well-being in a population study of men. RESULTS: Four multi-item scales were identified with satisfactory reliability and validity, i.e., they met defined criteria of multi-item scales and benefit from higher reliability than single questions. The psychological symptom scale comprised all ten questions intended to reflect depression and tension, while the physical symptom scale included all six pain questions, the general symptoms dizziness, breathlessness, and nausea, and general fatigue that belonged to both symptom scales. The social well-being scale comprises all well-being questions believed to reflect the broad concept of life satisfaction. The physical-mental wellbeing scale reflected satisfaction with both general and mental health. Several symptom and well-being questions were clearly outside the main clusters and should be treated as single items. CONCLUSIONS: Use of a few, distinct scales instead of a number of single questions should increase statistical power because the number of comparisons is greatly reduced and problems of chance findings are thus minimized.

Adult↗
PubMed2004

A stochastic model of vaccine trials for endemic infections using group randomization.

To clarify the determinants of vaccine trial power for non-typable Haemophilus influenzae, we constructed stochastic SIS models of infection transmission in small units (e.g. day-care centres) to calculate the equilibrium distribution of the number infected. We investigated how unit size, contact rate (modelled as a function of the unit size), external force of infection and infection duration affected the statistical power for detection of vaccine effects on susceptibility or infectiousness. Given a frequency-dependent contact rate, the prevalence, proportion of infections generated internally and the power to detect vaccine effects each increased slightly with unit size. Under a density-dependent model, unit size had much stronger effects. To maximize information allowing inference from vaccine trials, contact functions should be empirically evaluated by studying units of differing size and molecular methods should be used to help distinguish internal vs. external transmission.

Disease Transmission, Infectious↗
PubMed2006

Homocysteine-lowering trials for prevention of cardiovascular events: a review of the design and power of the large randomized trials.

BACKGROUND: Dietary supplementation with folic acid and vitamin B12 lowers blood homocysteine concentrations by about 25% to 30% in populations without routine folic acid fortification of food and by about 10% to 15% in populations with such fortification. In observational studies, 25% lower homocysteine has been associated with about 10% less coronary heart disease (CHD) and about 20% less stroke. METHODS: We reviewed the design and statistical power of 12 randomized trials assessing the effects of lowering homocysteine with B-vitamin supplements on risk of cardiovascular disease. RESULTS: Seven of these trials are being conducted in populations without fortification (5 involving participants with prior CHD and 2 with prior stroke) and 5 in populations with fortification (2 with prior CHD, 2 with renal disease, and 1 with prior stroke). These trials may not involve sufficient number of vascular events or last long enough to have a good chance on their own to detect reliably plausible effects of homocysteine lowering on cardiovascular risk. But, taken together, these 12 trials involve about 52,000 participants: 32,000 with prior vascular disease in unfortified populations and 14,000 with vascular disease and 6000 with renal disease in fortified populations. Hence, a combined analysis of these trials should have adequate power to determine whether lowering homocysteine reduces the risk of cardiovascular events within just a few years. CONCLUSION: The strength of association of homocysteine with risk of cardiovascular disease may be weaker than had previously been believed. Extending the duration of treatment in these trials would allow any effects associated with prolonged differences in homocysteine concentrations to emerge. Establishing a prospective meta-analysis of the ongoing trials of homocysteine lowering should ensure that reliable information emerges about the effects of such interventions on cardiovascular disease outcomes.

Cardiovascular Diseases↗
PubMed2003

Methodological and statistical considerations in measuring alcohol treatment effects.

This article addresses the design and selection of outcome methods for treatment research from a statistical perspective. Two of the major statistical and methodological issues relevant to the selection of dependent variable(s) are statistical power and social utility. Power is the paramount issue in research design. Power can be improved by measures and methods tailored to the predicted effects of treatment. These effects should vary from treatment to treatment, suggesting that no single outcome measure is suitable for all purposes. In estimating power, researchers should also consider the time-varying nature of most types of treatment effects. Given these considerations, timeline procedures have advantages because they allow the derivation of measures linked to specific treatment components. They also permit the study of time-varying effects of variables such as life stress. An analysis on the relationship between drinking and self-reported stress is presented as an illustration of how advanced statistical methods, in conjunction with carefully gathered data, can provide process data on how treatment can affect individual responses to stressors or other events. From a social utility perspective, however, having research focus on narrower and narrower slices of behavior in the search for power raises questions about the value of this research to clients and society. For dealing with this dilemma, it will be necessary to build a series of studies linking improvement in specific aspects of short-term outcome to longer term outcome and ultimately benefit to society.

Alcoholism↗
PubMed2005

Powering our way to the elusive side effect: a composite outcome 'basket' of predefined designated endpoints in each organ system should be included in all controlled trials.

OBJECTIVE: To review the methodological insights gained from the CLASS and VIGOR coxib trials and make suggestions for primary safety outcomes in future trials. STUDY DESIGN AND SETTING: We reviewed the methodology and findings of recent coxib trials and similar studies that have produced considerable controversy. RESULTS: There were statistical power problems with the primary and unexpected endpoints in both coxib trials. One approach for decreasing sample size requirements is using composite outcomes, wherein different clinically relevant endpoints are combined. By virtue of increasing the event rate fewer patients are required to detect a relative treatment effect of 50%. CONCLUSIONS: The recent withdrawal from the market of rofecoxib took years; the approach proposed here could shorten this time. It is important to revisit the comprehensiveness of the ability to detect important outcomes. Composite outcomes can be more varied in content, including mortality and morbidity. The proposed basket would consist of predefined designated endpoints in each organ system that could be associated with the population one is treating, and not just expected to be related to treatment. The goal is to anticipate and have a reasonable chance of detecting the unexpected side effect.

Cyclooxygenase Inhibitors↗
PubMed2005

Adjusting for treatment effects in studies of quantitative traits: antihypertensive therapy and systolic blood pressure.

A population-based study of a quantitative trait may be seriously compromised when the trait is subject to the effects of a treatment. For example, in a typical study of quantitative blood pressure (BP) 15 per cent or more of middle-aged subjects may take antihypertensive treatment. Without appropriate correction, this can lead to substantial shrinkage in the estimated effect of aetiological determinants of scientific interest and a marked reduction in statistical power. Correction relies upon imputation, in treated subjects, of the underlying BP from the observed BP having invoked one or more assumptions about the bioclinical setting. There is a range of different assumptions that may be made, and a number of different analytical models that may be used. In this paper, we motivate an approach based on a censored normal regression model and compare it with a range of other methods that are currently used or advocated. We compare these methods in simulated data sets and assess the estimation bias and the loss of power that ensue when treatment effects are not appropriately addressed. We also apply the same methods to real data and demonstrate a pattern of behaviour that is consistent with that in the simulation studies. Although all approaches to analysis are necessarily approximations, we conclude that two of the adjustment methods appear to perform well across a range of realistic settings. These are: (1) the addition of a sensible constant to the observed BP in treated subjects; and (2) the censored normal regression model. A third, non-parametric, method based on averaging ordered residuals may also be advocated in some settings. On the other hand, three approaches that are used relatively commonly are fundamentally flawed and should not be used at all. These are: (i) ignoring the problem altogether and analysing observed BP in treated subjects as if it was underlying BP; (ii) fitting a conventional regression model with treatment as a binary covariate; and (iii) excluding treated subjects from the analysis. Given that the more effective methods are straightforward to implement, there is no argument for undertaking a flawed analysis that wastes power and results in excessive bias.

Aged↗
PubMed2001

Predicting sulcus size using ocular measurements.

PURPOSE: To predict sulcus size using ocular measurements. SETTING: Michel Pop Clinics, Montreal, Quebec, Canada. METHODS: Forty-three eyes were evaluated using several techniques. Ultrasound biomicroscopy (UBM) echograms were taken to measure the anterior chamber depth (ACD), sulcus size, and central corneal thickness. The limbus size was measured with a caliper. Axial length, ACD, and pachymetry were measured by contact ultrasonography. Refraction and corneal power were also evaluated. RESULTS: The coefficient of linear regression was 0.05 between the limbus and the sulcus size (P =.78), 0.76 between ultrasonography and UBM ACD measurements (P <.001), and 0.69 between ultrasonography and UBM pachymetry (P <.001). Paired t tests showed that ultrasound and UBM ACD measurements were not statistically different (P =.70) but that ultrasound and UBM pachymetry measurements were (P <.001). The sulcus versus limbus difference was 0.6 mm for myopia and 0.3 mm for hyperopia. A backward elimination multiple regression performed with all measures to predict sulcus size resulted in the following formula: Sulcus size = 18.9 - 0.023 x sphere + 0.15 x mean keratometry (R = 0.49; P =.005; statistical power = 0.89; standard error of estimate = 0.5 mm). CONCLUSION: Traditional estimation of sulcus size through limbal measurement is inadequate because limbus size alone cannot predict sulcus size. A general formula using the sphere and the mean corneal power can help predict sulcus size. Corneal power was significantly and negatively correlated with sulcus and limbus size as well as sphere. The standard error of sulcus measurement by UBM was 0.4 mm.

Adult↗
PubMed2003

Mapping quantitative trait loci underlying triploid endosperm traits.

Endosperm, which is derived from two polar nuclei fusing with one sperm, is a triploid tissue in cereals. Endosperm tissue determines the grain quality of cereals. Improving grain quality is one of the important breeding objectives in cereals. However, current statistical methods for mapping quantitative trait loci (QTL) under diploid genetic control have not been effective for dealing with endosperm traits because of the complexity of their triploid inheritance. In this paper, we derive for the first time the conditional probabilities of F(3) endosperm QTL genotypes given different flanking marker genotypes in F(2) plants. Using these probabilities, we develop a multiple linear regression method implemented via the iteratively reweighted least-squares (IRWLS) algorithm and a maximum likelihood method (ML) implemented via the expectation-maximization (EM) algorithm to map QTL underlying endosperm traits. We use the mean value of endosperm traits of F(3) seeds as the dependent variable and the expectations of genotypic indicators for additive and dominance effect of a putative QTL flanked by a pair of markers as independent variables for IRWLS mapping. However, if an endosperm trait is measured quantitatively using a single endosperm sample, the ML mapping method can be used to separate the two dominance effects. Efficiency of the methods is verified through extensive Monte Carlo simulation studies. Results of simulation show that the proposed methods provide accurate estimates of both the QTL effects and locations with very high statistical power. With these methods, we are now ready to map endosperm traits, as we can for regular quantitative trait under diploid control.

Chromosome Mapping↗
PubMed2005

[Proper handling of correlated data in rehabilitation research].

Many study designs in rehabilitation science give rise to correlated data. For example, patients are followed over time, different responses are measured for each patient, or patients are observed in logical units. Standard statistical methods, however, are only valid for independent responses, and careless application of these methods for actually correlated observations might give erroneous results. By means of a simple example, we show how using methods for correlated data can indeed give a gain in statistical power. In the following, different approaches (Summary measures, Repeated Measurement ANOVA, MANOVA, and Mixed Models) to deal with correlated data are presented. We conclude that among these, the Mixed Models approach is the method of choice because it allows flexible modelling of correlation structure and is, meanwhile, also available in standard statistical software packages.

Analysis of Variance↗
PubMed2002

Cancer incidence among Finnish nuclear reactor workers.

Because of their well-documented exposures to repeated low doses of ionizing radiation, nuclear reactor workers offer an opportunity to assess cancer risk from low-dose radiation. A cohort of all 15,619 Finnish nuclear reactor workers was established through dose-monitoring records. A questionnaire survey revealed no substantial differences in consumption of tobacco or alcohol between different exposure groups nor between nuclear power company employees and contract workers. In the follow-up for cancer incidence, no clear excess in cancer incidence was observed overall, nor was any observed in any of the specific cancer types studied. There was little evidence for an association between cancer incidence and cumulative radiation dose, but the statistical power was limited. More precise estimates will be available from an international collaborative study of nuclear industry workers, including our cohort.

Adult↗
PubMed2002

Signal-driven computations in speech processing.

Learning a language requires both statistical computations to identify words in speech and algebraic-like computations to discover higher level (grammatical) structure. Here we show that these computations can be influenced by subtle cues in the speech signal. After a short familiarization to a continuous speech stream, adult listeners are able to segment it using powerful statistics, but they fail to extract the structural regularities included in the stream even when the familiarization is greatly extended. With the introduction of subliminal segmentation cues, however, these regularities can be rapidly captured.

Adult↗
PubMed1999

[Methodological aspects of biostatistics and epidemiology in a clinical trial].

Various methodological issues in biostatistics and epidemiology which are relevant for the development of a clinical trial are discussed. A brief set of definitions regarding clinical trials and methods for the assignment of treatments is presented. In addition, methods to reduce the possibility of bias that may be introduced into a clinical trial are summarized. The different elements that must be included in the informed consent form are defined. General considerations for data analysis and for sample size required to obtain an adequate statistical power are illustrated. Finally, critical questions that can be used for the review of the literature on clinical trials are suggested.

Aged↗
PubMed2005

[Markov Chain Monte Carlo Method of multiple imputation for longitudinal data with missing values in the survey of maternal and children health].

OBJECTIVE: To deal with arbitrary missing pattern in longitudinal data of the Survey of Maternal and Child Health and make the most appropriate inferences with multiple imputation (MI) for further analysis. METHODS: SAS 9.0 was used for Markov Chain Monte Carlo (MCMC) method of MI procedure to impute missing values and combine inferences. RESULTS: The result is acceptable as the data set was imputed 5 times. CONCLUSION: MI is able to solve a variety of problems in missing data sets and to improve the statistical power, especially with the use of MCMC method, for complicated missing data sets.

Bias↗
PubMed2000

Vagus nerve stimulation is associated with mood improvements in epilepsy patients.

Vagus nerve stimulation (VNS) has gained increasing acceptance for treatment of drug-resistant seizures. The aim of this study was to evaluate effects of VNS on depressed mood in epilepsy patients during the first 6 months after implantation of the stimulation device. This study was conducted as an addition to the international multisite randomized and double-blind controlled trial on anti-seizure effects of VNS (EO3). Only adult patients with >4/month medication-resistant complex-partial seizures were included (N=11). During the acute phase of the study (3 months after implantation), patients were randomly assigned to low (stimulation detection) versus high stimulation (maximal tolerability, maximum 1.75 mA). Mood and mood changes were recorded based on standardized psychiatric rating scales and self-report questionnaires. Patients were assessed 4 weeks before (baseline) as well as 3 and 6 months after implantation. Significant positive mood effects were observed in most scales and subscales at the 3-month follow-up (P<0.05). Mood improvements were sustained at the 6-month follow-up and were independent of effects on seizure activity (9/11 mood responders versus 2/11 seizure responders). Mood effects appeared more pronounced in the high stimulation group after the acute study phase, but findings were not significant (P<0.10). VNS is associated with mood improvements in patients with epilepsy, but to confirm VNS dose effects, studies with more statistical power are needed.

Adult↗
PubMedDate not supplied

A critical view of the utility of positive controls in a test battery.

Many behavioral teratologists underestimate the cumulative experimental effect from several studies, and regard a seeming inconsistency in results as a substantial inconsistency, as revealed by a survey of the attitudes toward significance test results. Such erroneous attitudes toward significance test results may lead to a confusion in the treatment of positive controls. Although the use of positive controls has been recommended, in practice, too small sample sizes lead to low statistical power in the administration of a test battery. The effectiveness of the positive control to verify the sensitivity of a test battery is discussed critically.

Animals↗
PubMed2007

Supplementary analyses regarding Langevin, Langevin, and Curnoe's (2007) findings on fraternal birth order in homosexual men.

A recent article by Langevin, Langevin, and Curnoe (2007) reported mixed results regarding the fraternal birth order effect, that is, the repeatedly observed finding that older brothers correlate with homosexuality in later-born males. Using a fraternal birth order index computed as older brothers minus younger brothers, Langevin et al. found that the "homoerotic" probands were born later among their brothers than were the "heteroerotic" probands in their full sample (N = 1194) and in their subsample over age 19 (N = 1122), but not in their subsample over age 31 (N = 698) or in their subsample with mothers over age 46 at the proband's birth (N = 727). The present writer concluded that the results obtained with the larger samples are more reliable, based on analyses demonstrating that (1) the larger samples are unlikely to be seriously affected by incomplete sibships, and (2) the smaller samples have poor statistical power. A separate analysis, based on an approximate reconstruction of Langevin et al.'s raw data, indicated that their heteroerotic probands reported a ratio of 104 older brothers per 100 older sisters, which is close to the normative population value of 106, whereas their homoerotic probands reported a ratio of 137, indicating a statistically significant excess of older brothers. These results suggest that Langevin et al.'s data showed significant evidence of a fraternal birth order effect and that their data were consistent with previous studies of this phenomenon.

Age Factors↗
PubMed1997

The Concerted Action 'Heart' European registry on clinical application of mechanical circulatory support systems: bridge to transplant. The Registry Scientific Committee.

OBJECTIVE: The goal of this paper is to identify in the field of mechanical support as bridge to transplant, by statistical analysis, variables influencing survival during support (transplanted patients) and the overall survival (discharged after transplant). METHODS: Clinical factors are analysed in 258 patients in the period 1986-1993. All variables were analyzed by a univariate and multivariate analysis. RESULTS: The indications for mechanically circulatory support were hemodynamic deterioration before transplantation in 177 (69%), post acute myocardial infarction in 40 (15%), postcardiotomy cardiogenic shock in 20 (8%), graft failure in 12 (5%) and cardiac rejection 9 (3%). The devices implanted have been: pneumatic VAD in 145 cases (56%), electromechanical LVAS in 15 cases (6%), TAH in 78 cases (30%) and centrifugal pumps in 20 cases (8%). The patients were supported for period ranging from 2 h to 623 days (mean 18.3 days +/- 43.2). The type of support was: LVAD 50 cases (20%), RVAD 3 cases (1%), BVAD 127 cases (49%), and TAH 78 cases (30%). Bleeding occurred in 84 patients (32.5%), infections in 83 patients (32.1%); 21 embolic complications were reported in 16 patients (6%). Renal failure occurred in 64 cases (25%) requiring dialysis in 33 (13%); respiratory failure in 47 cases (18%); neurological impairment was noted in 22 patients (9%). One hundred-sixty patients were transplanted (62%) and 104 ultimately discharged (40% out of total 258 patients and 65% out of 160 transplanted patients). Among postoperative parameters, renal failure, TAH, neurological impairment and infection shown statistical power. Some pre- and post-operative variables were identified as independent risk factors for overall mortality: age, indication for graft failure, all indications different from cardiomyopathy, neurological impairment, renal insufficiency, infection, bleeding and any type of support different from LVAD. The improvement in the success rate in the last 2 years is statistically significant (P = 0.0282) considering both the percentage of transplanted patients and of discharged patients. CONCLUSIONS: The results are encouraging if mechanical support is performed in patients with deterioration while awaiting transplant, when LVAD is feasible and effective, when an ideal timing of transplant during support period is identified.

Adolescent↗
PubMed2005

Power and sample size for clinical trials when efficacy is required in multiple endpoints: application to an Alzheimer's treatment trial.

BACKGROUND: When the efficacy of a treatment in a randomized controlled trial is required for multiple primary endpoints, trial design and analysis differ from trial requiring efficacy in only one of the multiple endpoints. METHODS: We consider a two-arm clinical trial requiring efficacy analysis for multiple primary endpoints, formulating the appropriate null and alternative hypotheses for the test of treatment efficacy. We study the significance level/statistical power of an intersection-union test (IUT) in this situation. We compare IUT with the intuitive approach (selecting the maximum sample size over those obtained from testing individual primary endpoints one by one) for determination of sample size. RESULTS: The proposed IUT reserves the same Type I error rate as shared by all endpoint-specific tests. The statistical power of the proposed IUT is no more than the minimum from the individual tests. The maximum sample size from multiple endpoint-specific tests is often inadequate for the test of treatment efficacy, especially when the standardized effect sizes are similar. Finally, the IUT can be applied to Alzheimer's disease treatment trials in which two primary endpoints are typically used. CONCLUSIONS: The IUT is a valid method for use in the design and analysis of clinical trials requiring efficacy at multiple primary endpoints.

Alzheimer Disease↗