PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Bayesian analysis”

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.

At least 829 records · Page 46Linked to original sources
PubMed2005

Accuracy of MSI testing in predicting germline mutations of MSH2 and MLH1: a case study in Bayesian meta-analysis of diagnostic tests without a gold standard.

Microsatellite instability (MSI) testing is a common screening procedure used to identify families that may harbor mutations of a mismatch repair (MMR) gene and therefore may be at high risk for hereditary colorectal cancer. A reliable estimate of sensitivity and specificity of MSI for detecting germline mutations of MMR genes is critical in genetic counseling and colorectal cancer prevention. Several studies published results of both MSI and mutation analysis on the same subjects. In this article we perform a meta-analysis of these studies and obtain estimates that can be directly used in counseling and screening. In particular, we estimate the sensitivity of MSI for detecting mutations of MSH2 and MLH1 to be 0.81 (0.73-0.89). Statistically, challenges arise from the following: (a) traditional mutation analysis methods used in these studies cannot be considered a gold standard for the identification of mutations; (b) studies are heterogeneous in both the design and the populations considered; and (c) studies may include different patterns of missing data resulting from partial testing of the populations sampled. We address these challenges in the context of a Bayesian meta-analytic implementation of the Hui-Walter design, tailored to account for various forms of incomplete data. Posterior inference is handled via a Gibbs sampler.

Adaptor Proteins, Signal Transducing↗
PubMed2006

Bayesian fMRI data analysis with sparse spatial basis function priors.

In previous work we have described a spatially regularised General Linear Model (GLM) for the analysis of brain functional Magnetic Resonance Imaging (fMRI) data where Posterior Probability Maps (PPMs) are used to characterise regionally specific effects. The spatial regularisation is defined over regression coefficients via a Laplacian kernel matrix and embodies prior knowledge that evoked responses are spatially contiguous and locally homogeneous. In this paper we propose to finesse this Bayesian framework by specifying spatial priors using Sparse Spatial Basis Functions (SSBFs). These are defined via a hierarchical probabilistic model which, when inverted, automatically selects an appropriate subset of basis functions. The method includes non-linear wavelet shrinkage as a special case. As compared to Laplacian spatial priors, SSBFs allow for spatial variations in signal smoothness, are more computationally efficient and are robust to heteroscedastic noise. Results are shown on synthetic data and on data from an event-related fMRI experiment.

Algorithms↗
PubMed2005

Introduction to Bayesian methods II: fundamental concepts.

The use of Bayesian design and analysis is burgeoning. In this introduction to Bayesian methods, we provide basic examples of Bayesian thinking and formalism on which more complicated and comprehensive approaches are built. These include adjusting estimates using related information, the use of Bayes theorem in diagnostic testing, the relationship of the prior and posterior distributions for situations where both the data and prior distribution are Gaussian, and the key steps in a Bayesian analysis. If Bayesian methods are carefully developed and applied, they have excellent objective (i.e., frequentist) properties, providing marvelous tools to help improve the FDA regulatory process.

Bayes Theorem↗
PubMed2006

Self-monitoring of glucose in type 2 diabetes mellitus: a Bayesian meta-analysis of direct and indirect comparisons.

OBJECTIVE: To evaluate the relative effectiveness of interventions with self-monitoring blood glucose and self-monitoring of urine glucose, versus interventions without self-monitoring, in terms of HbA(1c) reductions in type 2 diabetes mellitus. METHODS: Thirteen published full reports on randomised controlled trials investigating the effects of self-monitoring glucose were identified by a systematic search of Medline, Embase, the Cochrane Library (1966-Nov 2005) and previous reviews. Three types of studies were included: self-monitoring of blood glucose versus no self-monitoring, self-monitoring of blood glucose versus self-monitoring of urine glucose and self-monitoring of blood glucose with regular feedback versus monitoring without feedback. The internal validity of studies was assessed systematically by two reviewers, using 13 criteria of a validated list. Results from the three types of studies were analysed simultaneously with a Bayesian metaanalysis of direct and indirect comparisons. RESULTS: Adjusted for baseline HbA(1c) level and internal validity, interventions with self-monitoring of blood glucose showed a reduction in HbA(1c) of 0.40 percentage-points (%) (95% credible interval [CrI] 0.07 to 0.70%) in comparison to interventions without self-monitoring. Regular feedback more than doubled the HbA(1c) reduction. Self-monitoring of urine glucose showed comparable results to interventions without self-monitoring (0.02% decrease in HbA(1c); 95% CrI -0.62 to 0.70%). There is a 88% probability that interventions with self-monitoring blood glucose are more effective than interventions with urine glucose monitoring (relative reduction in HbA(1c) is 0.38%, 95% CrI -0.30 to 1.00%). CONCLUSION: The randomized clinical trials performed to date provided positive results on the effectiveness of interventions with self-monitoring of blood glucose in type 2 diabetes mellitus. Regular medical feedback of the monitored HbA(1c) levels is important. Furthermore, self-monitoring of blood glucose is likely to be more effective than self-monitoring of urine glucose.

Bayes Theorem↗
PubMed2003

Evidence for use of coronary stents. A hierarchical bayesian meta-analysis.

BACKGROUND: Coronary stents are widely used in interventional cardiology, but a current quantitative systematic overview comparing routine coronary stenting with standard percutaneous transluminal coronary angioplasty (PTCA) and restricted stenting (provisional stenting) has not been published. PURPOSE: To summarize results from all randomized clinical trials comparing routine coronary stenting with standard PTCA. DATA SOURCES: Electronic databases were searched by using the key words angioplasty and stent. References from identified articles were also reviewed. In addition, several prominent general medical and cardiology journals were searched and agencies known to perform systematic reviews were consulted. STUDY SELECTION: All comparative randomized clinical trials were included, except those involving primary angioplasty for the treatment of acute myocardial infarction. DATA EXTRACTION: A specified protocol was followed, and two of the authors independently extracted the data. Outcomes assessed were total mortality, myocardial infarction, angiographic restenosis, coronary artery bypass surgery, repeated PTCA, and freedom from angina. DATA SYNTHESIS: The results were synthesized by using a Bayesian hierarchical random-effects model. A total of 29 trials involving 9918 patients were identified. There was no evidence for a difference between routine coronary stenting and standard PTCA in terms of deaths or myocardial infarctions (odds ratio, 0.90 [95% credible interval [CrI], 0.72 to 1.11]) or the need for coronary artery bypass surgery (odds ratio, 1.01 [CrI, 0.79 to 1.31]). Coronary stenting reduced the rate of restenosis (odds ratio, 0.52 [CrI, 0.37 to 0.69]) and the need for repeated PTCA (odds ratio, 0.59 [CrI, 0.50 to 0.68]). The trials showed a wide range of crossover rates from PTCA to stenting. By use of a multiplicative model, each 10% increase in crossover rate decreased the need for repeated angioplasty by approximately 8% (odds ratio multiplying factor, 1.08 [CrI, 0.98 to 1.18]). Routine stenting probably reduces the need for repeated angioplasty by fewer than 4 to 5 per 100 treated persons compared with PTCA with provisional stenting. Studies were not blinded and suggest a bias with a possible overestimation of this benefit. CONCLUSIONS: In the controlled environment of randomized clinical trials, routine coronary stenting is safe but probably not associated with important reductions in rates of mortality, acute myocardial infarction, or coronary artery bypass surgery compared with standard PTCA with provisional stenting. Coronary stenting is associated with substantial reductions in angiographic restenosis rates and the subsequent need for repeated PTCA, although this benefit may be overestimated because of trial designs. The incremental benefit of routine stenting for reducing repeated angioplasty diminishes as the crossover rate of stenting with conventional PTCA increases.

Aged↗
PubMed2001

Probabilistic cost-effectiveness analysis of HIV prevention. Comparing a Bayesian approach with traditional deterministic sensitivity analysis.

In cost-effectiveness analysis, the incremental cost-effectiveness ratio is used to measure economic efficiency of a new intervention, relative to an existing one. However, costs and effects are seldom known with certainty. Uncertainty arises from two main sources: uncertainty regarding correct values of intervention-related parameters and uncertainty associated with sampling variation. Recently, attention has focused on Bayesian techniques for quantifying uncertainty. We computed the Bayesian-based 95% credible interval estimates of the incremental cost-effectiveness ratio of several related HIV prevention interventions and compared these results with univariate sensitivity analyses. The conclusions were comparable, even though the probabilistic technique provided additional information.

Bayes Theorem↗
PubMed2007

Bayesian meta-analysis and meta-regression for gene-disease associations and deviations from Hardy-Weinberg equilibrium.

Violation of Hardy-Weinberg equilibrium (HWE) can raise doubts about the validity of the conclusions from genetic association studies. However, for most currently performed gene-disease association studies, the available tests have low power to detect deviations from HWE. We consider this issue from a meta-analysis perspective, and suggest an approach to estimate the deviation and investigate its relationship with the observed genetic effects. Different degrees of deviation from HWE have previously been proposed as a potential source of heterogeneity across studies. We present a hierarchical meta-regression model that can be applied to test this assumption, using the concept of the fixation coefficient. We re-analyse seven meta-analyses to illustrate these methods. The uncertainty in the genetic effect estimate tended to increase once the fixation coefficient was taken into account. Dependence of the genetic effect size on the deviation from HWE was found in one meta-analysis, while in the other six examples, deviations from HWE did not clearly explain between-study heterogeneity in the genetic effects. The proposed hierarchical models allow the synthesis of data across gene-disease association studies with appropriate consideration of HWE issues.

Alleles↗
PubMed2003

Bayesian meta-analysis for longitudinal data models using multivariate mixture priors.

We propose a class of longitudinal data models with random effects that generalizes currently used models in two important ways. First, the random-effects model is a flexible mixture of multivariate normals, accommodating population heterogeneity, outliers, and nonlinearity in the regression on subject-specific covariates. Second, the model includes a hierarchical extension to allow for meta-analysis over related studies. The random-effects distributions are decomposed into one part that is common across all related studies (common measure), and one part that is specific to each study and that captures the variability intrinsic between patients within the same study. Both the common measure and the study-specific measures are parameterized as mixture-of-normals models. We carry out inference using reversible jump posterior simulation to allow a random number of terms in the mixtures. The sampler takes advantage of the small number of entertained models. The motivating application is the analysis of two studies carried out by the Cancer and Leukemia Group B (CALGB). In both studies, we record for each patient white blood cell counts (WBC) over time to characterize the toxic effects of treatment. The WBCs are modeled through a nonlinear hierarchical model that gathers the information from both studies.

Bayes Theorem↗
PubMedDate not supplied

Sequential analysis in a Bayesian model of diastolic blood pressure measurement.

A sequential method for diagnosing or excluding hypertension based on the Bayesian model of diastolic blood pressure presented in a companion article is presented. The likelihood ratio method of Wald is modified to include the effects of a prior probability distribution and to constrain the strategy to achieve specified positive and negative predictive values. The resulting formulas for upper and lower limits to diagnose and exclude diastolic hypertension can be evaluated using a hand calculator and a table of areas of the standard normal distribution. The strategy is illustrated for a population having a blood pressure distribution similar to that of the cohort screened for participation in the Hypertension Detection and Follow-up Program, with 90 mm Hg as the cutoff defining hypertension and required positive and negative predictive values of 95%. The performance of the strategy was simulated using Monte Carlo methods. The median number of readings required for diagnosis is three, and 80% of subjects are diagnosed in 11 or fewer readings. In contrast to the strategy's 95% predictive values, a fixed-number-of-measurements strategy requiring the same mean number of measurements has a positive predictive value of only 83% and a negative predictive value of 96%. When the parameters of the model have been properly measured or estimated, this method is practical, efficient, and accurate for diagnosing hypertension in a known population.

Bayes Theorem↗
PubMed2005

Synthesising qualitative and quantitative evidence: a review of possible methods.

BACKGROUND: The limitations of traditional forms of systematic review in making optimal use of all forms of evidence are increasingly evident, especially for policy-makers and practitioners. There is an urgent need for robust ways of incorporating qualitative evidence into systematic reviews. OBJECTIVES: In this paper we provide a brief overview and critique of a selection of strategies for synthesising qualitative and quantitative evidence, ranging from techniques that are largely qualitative and interpretive through to techniques that are largely quantitative and integrative. RESULTS: A range of methods is available for synthesising diverse forms of evidence. These include narrative summary, thematic analysis, grounded theory, meta-ethnography, meta-study, realist synthesis, Miles and Huberman's data analysis techniques, content analysis, case survey, qualitative comparative analysis and Bayesian meta-analysis. Methods vary in their strengths and weaknesses, ability to deal with qualitative and quantitative forms of evidence, and type of question for which they are most suitable. CONCLUSIONS: We identify a number of procedural, conceptual and theoretical issues that need to be addressed in moving forward with this area, and emphasise the need for existing techniques to be evaluated and modified, rather than inventing new approaches.

Bayes Theorem↗
PubMed2003

Analysis of a Bayesian repeated measures model for detecting differences in GP prescribing habits.

A linear mixed model is used to detect a change, if any, in the prescribing habits in the UK at the general practice (family medicine) level due to an educational intervention given repeated measures data before and after the intervention and a control group. Inferences are corrected for general practice size and fundholding status. The estimates of the model parameters are obtained using Bayesian inference by applying Gibbs sampling. We develop three different priors for the parameters of the model. These three priors correspond to 'sceptical,' 'reference' and 'enthusiastic' priors in terms of the opinion about the treatment effects that they represent. We compare the results obtained by using these three priors for the parameters in the random effects model.

Anti-Inflammatory Agents, Non-Steroidal↗
PubMed2005

Systematically reviewing qualitative and quantitative evidence to inform management and policy-making in the health field.

Policy-makers and managers have always used a wide range of sources of evidence in making decisions about policy and the organization of services. However, they are under increasing pressure to adopt a more systematic approach to the utilization of the complex evidence base. Decision-makers must address complicated questions about the nature and significance of the problem to be addressed; the nature of proposed interventions; their differential impact; cost-effectiveness; acceptability and so on. This means that Cochrane-style reviews alone are not sufficient. Rather, they require access to syntheses of high-quality evidence that include research and non-research sources, and both qualitative and quantitative research findings. There is no single, agreed framework for synthesizing such diverse forms of evidence and many of the approaches potentially applicable to such an endeavour were devised for either qualitative or quantitative synthesis and/or for analysing primary data. This paper describes the key stages in reviewing and synthesizing qualitative and quantitative evidence for decision-making and looks at various strategies that could offer a way forward. We identify four basic approaches: narrative (including traditional 'literature reviews' and more methodologically explicit approaches such as 'thematic analysis', 'narrative synthesis', 'realist synthesis' and 'meta-narrative mapping'), qualitative (which convert all available evidence into qualitative form using techniques such as 'meta-ethnography' and 'qualitative cross-case analysis'), quantitative (which convert all evidence into quantitative form using techniques such as 'quantitative case survey' or 'content analysis') and Bayesian meta-analysis and decision analysis (which can convert qualitative evidence such as preferences about different outcomes into quantitative form or 'weights' to use in quantitative synthesis). The choice of approach will be contingent on the aim of the review and nature of the available evidence, and often more than one approach will be required.

Delivery of Health Care↗
PubMed2001

Beta-blockers in congestive heart failure. A Bayesian meta-analysis.

PURPOSE: Congestive heart failure is an important cause of patient morbidity and mortality. Although several randomized clinical trials have compared beta-blockers with placebo for treatment of congestive heart failure, a meta-analysis quantifying the effect on mortality and morbidity has not been performed recently. DATA SOURCES: The MEDLINE, Cochrane, and Web of Science electronic databases were searched from 1966 to July 2000. References were also identified from bibliographies of pertinent articles. STUDY SELECTION: All randomized clinical trials of beta-blockers versus placebo in chronic stable congestive heart failure were included. DATA EXTRACTION: A specified protocol was followed to extract data on patient characteristics, beta-blocker used, overall mortality, hospitalizations for congestive heart failure, and study quality. DATA SYNTHESIS: A hierarchical random-effects model was used to synthesize the results. A total of 22 trials involving 10 135 patients were identified. There were 624 deaths among 4862 patients randomly assigned to placebo and 444 deaths among 5273 patients assigned to beta-blocker therapy. In these groups, 754 and 540 patients, respectively, required hospitalization for congestive heart failure. The probability that beta-blocker therapy reduced total mortality and hospitalizations for congestive heart failure was almost 100%. The best estimates of these advantages are 3.8 lives saved and 4 fewer hospitalizations per 100 patients treated in the first year after therapy. The probability that these benefits are clinically significant (>2 lives saved or >2 fewer hospitalizations per 100 patients treated) is 99%. Both selective and nonselective agents produced these salutary effects. The results are robust to any reasonable publication bias. CONCLUSIONS: beta-Blocker therapy is associated with clinically meaningful reductions in mortality and morbidity in patients with stable congestive heart failure and should be routinely offered to all patients similar to those included in trials.

Adrenergic beta-Antagonists↗
PubMed2002

Molecular systematics of the Eastern Fence Lizard (Sceloporus undulatus): a comparison of Parsimony, Likelihood, and Bayesian approaches.

Phylogenetic analysis of large datasets using complex nucleotide substitution models under a maximum likelihood framework can be computationally infeasible, especially when attempting to infer confidence values by way of nonparametric bootstrapping. Recent developments in phylogenetics suggest the computational burden can be reduced by using Bayesian methods of phylogenetic inference. However, few empirical phylogenetic studies exist that explore the efficiency of Bayesian analysis of large datasets. To this end, we conducted an extensive phylogenetic analysis of the wide-ranging and geographically variable Eastern Fence Lizard (Sceloporus undulatus). Maximum parsimony, maximum likelihood, and Bayesian phylogenetic analyses were performed on a combined mitochondrial DNA dataset (12S and 16S rRNA, ND1 protein-coding gene, and associated tRNA; 3,688 bp total) for 56 populations of S. undulatus (78 total terminals including other S. undulatus group species and outgroups). Maximum parsimony analysis resulted in numerous equally parsimonious trees (82,646 from equally weighted parsimony and 335 from weighted parsimony). The majority rule consensus tree derived from the Bayesian analysis was topologically identical to the single best phylogeny inferred from the maximum likelihood analysis, but required approximately 80% less computational time. The mtDNA data provide strong support for the monophyly of the S. undulatus group and the paraphyly of "S. undulatus" with respect to S. belli, S. cautus, and S. woodi. Parallel evolution of ecomorphs within "S. undulatus" has masked the actual number of species within this group. This evidence, along with convincing patterns of phylogeographic differentiation suggests "S. undulatus" represents at least four lineages that should be recognized as evolutionary species.

Animals↗