Borrowing strength from external trials in a meta-analysis

There exists a variety of situations in which a random effects meta-analysis might be undertaken using a small number of clinical trials. A problem associated with small meta-analyses is estimating the heterogeneity between trials. To overcome this problem, information from other related studies may be incorporated into the meta-analysis. A Bayesian approach to this problem is presented using data from previous meta-analyses in the same therapeutic area to formulate a prior distribution for the heterogeneity. The treatment difference parameters are given non-informative priors. Further, related trials which compare one or other of the treatments of interest with a common third treatment are included in the model to improve inference on both the heterogeneity and the treatment difference. Two approaches to estimating relative efficacy are considered, namely a general parametric approach and a method explicit to binary data. The methodology is illustrated using data from 26 clinical trials which investigate the prevention of cirrhosis using beta-blockers and sclerotherapy. Both sources of external information lead to more precise posterior distributions for all parameters, in particular that representing heterogeneity.     

The relative benefits of meta-analysis conducted with individual participant data versus aggregated data.

The authors describe the relative benefits of conducting meta-analyses with (a) individual participant data (IPD) gathered from the constituent studies and (b) aggregated data (AD), or the group-level statistics (in particular, effect sizes) that appear in reports of a study’s results. Given that both IPD and AD are equally available, meta-analysis of IPD is superior to meta-analysis of AD. IPD meta-analysis permits synthesists to perform subgroup analyses not conducted by the original collectors of the data, to check the data and analyses in the original studies, to add new information to the data sets, and to use different statistical methods. However, the cost of IPD meta-analysis and the lack of available IPD data sets suggest that the best strategy currently available is to use both approaches in a complementary fashion such that the first step in conducting an IPD meta-analysis would be to conduct an AD meta-analysis. Regardless of whether a meta-analysis is conducted with IPD or AD, synthesists must remain vigilant in how they interpret their results. They must avoid ecological fallacies, Simpson’s paradox, and interpretation of synthesis-generated evidence as supporting causal inferences. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

History of depression and smoking cessation outcome: A meta-analysis.

The authors conducted a meta-analysis of published studies to (a) evaluate the premise that a history of major depression is associated with failure to quit smoking and (b) identify factors that moderate the relationship between history of depression and cessation outcome. Fifteen studies met the selection requirements and were coded for various study methodology and treatment characteristics. DSTAT was used to calculate individual study effect sizes, determine the mean effect size across studies. and test for moderator effects. No differences in either short-term (≤ 3 months) or long-term abstinence rates (≥ 6 months) were observed between smokers positive versus negative for history of depression. Lifetime history of major depression does not appear to be an independent risk factor for cessation failure in smoking cessation treatment. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A comparison of various estimators of a treatment difference for a multi-centre clinical trial

When a clinical trial is conducted at more than one centre it is likely that the true treatment effect will not be identical at each centre. In other words there will be some degree of treatment-by-centre interaction. A number of alternative approaches for dealing with this have been suggested in the literature. These include frequentist approaches with a fixed or random effects model for the observed data and Bayesian approaches. In the fixed effects model, there are two common competing estimators of the treatment difference, based on weighted or unweighted estimates from individual centres. Which one of these should be used is the subject of some controversy and we do not intend to take a particular methodological position in this paper. Our intention is to provide some insight into the relative merits of the indicated range of possible estimators of the treatment effect. For the fixed effects model, we also look at the merits of using a preliminary test for interaction assuming a 10 per cent significance level for the test. In order to make comparisons we have simulated a 'typical' trial which compares an active drug with a placebo in the treatment of hypertension, using systolic blood pressure as the primary variable. As well as allowing the treatment effect to vary between centres, we have concentrated on the particular case where one centre is out of line with the others in terms of its true treatment difference. The various estimators that result from the different approaches are compared in terms of mean squared error and power to reject the null hypothesis of no treatment difference. Overall, the approach that uses the fixed effects weighted estimator of overall treatment difference is recommended as one that has much to offer.     

A Meta-Analysis of Massage Therapy Research.

Massage therapy (MT) is an ancient form of treatment that is now gaining popularity as part of the complementary and alternative medical therapy movement. A meta-analysis was conducted of studies that used random assignment to test the effectiveness of MT. Mean effect sizes were calculated from 37 studies for 9 dependent variables. Single applications of MT reduced state anxiety, blood pressure, and heart rate but not negative mood, immediate assessment of pain, and cortisol level. Multiple applications reduced delayed assessment of pain. Reductions of trait anxiety and depression were MT's largest effects, with a course of treatment providing benefits similar in magnitude to those of psychotherapy. No moderators were statistically significant, though continued testing is needed. The limitations of a medical model of MT are discussed, and it is proposed that new MT theories and research use a psychotherapy perspective. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Cumulative research knowledge and social policy formulation: The critical role of meta-analysis.

For many years, policymakers expressed increasing frustration with social science research. On every issue there were studies arguing for diametrically opposed conclusions. Methods of meta-analysis that correct for the effects of sampling error have shown that almost all such conflicting results were caused by sampling error. Furthermore, the effects of sampling error are greatly exaggerated by using significance test methodology. In many areas, meta-analysis has now provided dependable answers to the original research questions. Meta-analysis is now increasingly being used by policymakers, by textbook writers, and by theorists to provide the basic facts needed to draw both practical and explanatory conclusions. Sophisticated meta-analysis procedures are now used to correct for the effects of other study imperfections, such as measurement error, range restriction, and artificial dichotomization. In domains where the data on artifacts are available, the effect sizes in necessarily imperfect studies have been found to be considerably understated. Path analysis can be applied to the findings from meta-analysis to yield improved causal analyses that result in both explanation of results and improved generalization of results to new settings. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Links between physical fitness and cardiovascular reactivity and recovery to psychological stressors: A meta-analysis.

A meta-analysis of published studies with adult human participants was conducted to evaluate whether physical fitness attenuates cardiovascular reactivity and improves recovery from acute psychological stressors. Thirty-three studies met selection criteria; 18 were included in recovery analyses. Effect sizes and moderator influences were calculated by using meta-analysis software. A fixed effects model was fit initially; however, between-studies heterogeneity could not be explained even after inclusion of moderators. Therefore, to account for residual heterogeneity, a random effects model was estimated. Under this model, fit individuals showed significantly attenuated heart rate and systolic blood pressure reactivity and a trend toward attenuated diastolic blood pressure reactivity. Fit individuals also showed faster heart rate recovery, but there were no significant differences in systolic blood pressure or diastolic blood pressure recovery. No significant moderators emerged. Results have important implications for elucidating mechanisms underlying effects of fitness on cardiovascular disease and suggest that fitness may be an important confound in studies of stress reactivity. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Fish oil therapy for IgA nephropathy: efficacy and interstudy variability

Published reports examining the efficacy of fish oil for preserving renal function in immunoglobulin A (IgA) nephropathy have yielded conflicting results. This investigation was a meta-analysis conducted to determine whether the medical literature supports this therapy. In addition, the sources of variability among published findings were examined. Studies were combined using a random effects model. Five controlled studies were identified, two with positive results and three with negative results. Forty-four percent of the between-study variance could be attributed to differences in follow-up times and, less significantly, the number of renal function measurements; a weighting procedure was developed, eliminating this variance from the combined result. When all studies were combined, the mean effect, +0.25 +/- 0.23 SD (positive effects indicate that treatment was superior to control), was not statistically significant; however, the probability of at least a minor beneficial effect was 75%. Mixed-effects regression suggested that this therapy may be more effective among individuals with more proteinuria. The medical literature, therefore, does not prove the efficacy of fish oil therapy in IgA nephropathy, but suggests that an additional placebo-controlled trial is warranted. A sample-size calculation indicated that such a trial should be larger than those to date or should attempt to increase the treatment effect, perhaps by treating for more than 2 yr or enrolling more severely proteinuric individuals.     

Combining studies using effect sizes and quality scores: application to bone loss in postmenopausal women

This article presents a random effects model that uses effect sizes (ES) and quality scores to integrate results from investigations. An empirical example is given with data obtained from a meta-analysis on the effectiveness of physical activity in the prevention of bone loss in healthy postmenopausal women. A Medline search was performed to locate relevant studies published in French or English between January 1966 and May 1996. Three independent reviewers extracted data from studies. Effect sizes were calculated according to the method of Hedges and Olkin. A modified version of Chalmers' scale was utilized to calculate quality scores. DerSimonian and Laird's method with incorporation of the quality scores was used to estimate the overall effect size. Quality scores and the inverse of the variances were included as weights when combining studies. The overall estimate and standard error (SE) of the effect of physical activity on spinal bone mineral density loss in healthy postmenopausal women was ESoverall = 0.4263 (1.1361). When compared to other meta-analysis methods such as the fixed effects model and the model of DerSimonian and Laird without the quality score (DL), the new model generated comparable estimators (fixed effects model's ESoverall (SE) = 1.2724 (0.0139), DLs ESoverall (SE) = 0.3958 (1.2370)). Due to the heterogeneity that existed between studies, a random effects model was more appropriate then a fixed effects model. However, it resulted in wider confidence intervals, as expected. It was shown empirically that the model using quality scores generated narrower confidence intervals than the model of DL alone. The inclusion of covariates such as quality scores in meta-analyses permits the quantification of the variation between studies.     

Meta-analysis methodology for combining non-parametric sibpair linkage results: genetic homogeneity and identical markers

Meta-analysis methodology is developed for combining sibpair linkage results across multiple studies employing different study designs, some employing quantitative traits (e.g., blood pressure) and some employing qualitative traits (e.g., clinical hypertension), under the assumption that the underlying (disease) trait loci are the same. Pooling results based on three commonly used sibpair methods is considered: the affected sibpair method for dichotomous traits and, for quantitative traits, the Haseman-Elston regression method and the Risch-Zhang extremely discordant sibpair method. The proportion of genes shared identical by descent (IBD) by a sibpair of certain trait outcomes is chosen as a common effect to be pooled across studies. Variation in the observed IBD proportions among individual studies is modeled using a random effects model. A heterogeneity test is provided to assess the variability among individual studies. When results from all three types of studies are available, we derive pooled estimates of IBD proportions both for sibpairs with extremely concordant trait values and for sibpairs with extremely discordant trait values, and construct a combined test of linkage based on the difference of the two estimates. Simulation studies demonstrate the need for and the advantage of meta-analysis of linkage results. We also present some guidelines for reporting linkage studies bearing potential future meta-analysis in mind.     

A likelihood approach to meta-analysis with random effects

In a meta-analysis of a set of clinical trials, a crucial but problematic component is providing an estimate and confidence interval for the overall treatment effect theta. Since in the presence of heterogeneity a fixed effect approach yields an artificially narrow confidence interval for theta, the random effects method of DerSimonian and Laird, which incorporates a moment estimator of the between-trial components of variance sigma B2, has been advocated. With the additional distributional assumptions of normality, a confidence interval for theta may be obtained. However, this method does not provide a confidence interval for sigma B2, nor a confidence interval for theta which takes account of the fact that sigma B2 has to be estimated from the data. We show how a likelihood based method can be used to overcome these problems, and use profile likelihoods to construct likelihood based confidence intervals. This approach yields an appropriately widened confidence interval compared with the standard random effects method. Examples of application to a published meta-analysis and a multicentre clinical trial are discussed. It is concluded that likelihood based methods are preferred to the standard method in undertaking random effects meta-analysis when the value of sigma B2 has an important effect on the overall estimated treatment effect.     

Effects of psychotherapy with children and adolescents revisited: A meta-analysis of treatment outcome studies.

A meta-analysis of child and adolescent psychotherapy outcome research tested previous findings using a new sample of 150 outcome studies and weighted least squares methods. The overall mean effect of therapy was positive and highly significant. Effects were more positive for behavioral than for nonbehavioral treatments, and samples of adolescent girls showed better outcomes than other Age?×?Gender groups. Paraprofessionals produced larger overall treatment effects than professional therapists or students, but professionals produced larger effects than paraprofessionals in treating overcontrolled problems (e.g., anxiety and depression). Results supported the specificity of treatment effects: Outcomes were stronger for the particular problems targeted in treatment than for problems not targeted. The findings shed new light on previous results and raise significant issues for future study. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Meta-analysis of clinical trials for chemotherapy in cancer

Meta-analysis has attracted great interest among clinical practitioners in recent years, leading to a steady output of related publications. Meta-analytic articles are easily found in the MEDLINE database using the publication-type option. This paper reviews how to use and understand meta-analysis with a special reference to chemotherapy applied to cancer patients. It is described in relationship to evidence-based medicine (EBM) and clinical practice guidelines. Cochrane collaboration is also referred to as an active voluntary organization conducting meta-analysis. In the technical sections, statistical issues and graphic representations are clearly illustrated using the example of hepatic arterial infusion for colorectal cancer patients. The difference between fixed and random effects models is briefly explained. Finally, an example from Cochrane Library, namely progestagen therapy for endometrial cancer, is illustrated to show the implications of meta-analysis for clinical practice.     

Computing Travel Time Reliability in Transportation Networks with Multistates and Dependent Link Failures

The measurement of transportation network travel time reliability is imperative to provide drivers with accurate route guidance information and to generate the shortest path (or alternative paths) connecting origins and destinations, especially under conditions of varying demands and limited capacities. Previous studies assumed that link failures in a road network are statistically independent and that reliability probabilities are perfectly determinable. In real life, these assumptions are dubious, because the failure of a link in one particular area does not necessarily result in the complete failure of the neighboring link, but may lead to deterioration of its performance. This paper presents a new methodology to study the multistate system reliability analysis of transportation networks for which one cannot formulate an “all or nothing” type of failure criterion and in which dependent link failures are considered. The methodology is presented, using a numerical example, for computing the probability that travel time between an origin and a destination may exceed a threshold.     

The effects of psychological therapies under clinically representative conditions: A meta-analysis.

Recently, concern has arisen that meta-analyses overestimate the effects of psychological therapies and that those therapies may not work under clinically representative conditions. This meta-analysis of 90 studies found that therapies are effective over a range of clinical representativeness. The projected effects of an ideal study of clinically representative therapy are similar to effect sizes in past meta-analyses. Effects increase with larger dose and when outcome measures are specific to treatment. Some clinically representative studies used self-selected treatment clients who were more distressed than available controls, and these quasi-experiments underestimated therapy effects. This study illustrates the joint use of fixed and random effects models, use of pretest effect sizes to study selection bias in quasi-experiments, and use of regression analysis to project results to an ideal study in the spirit of response surface modeling. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Simulation of Homogeneous and Partially Isotropic Random Fields

A rigorous methodology for the simulation of homogeneous and partially isotropic multidimensional random fields is introduced. The property of partial isotropy of the random field is explicitly incorporated in the derivation of the algorithm. This consideration reduces significantly the computational effort associated with the generation of sample functions, as compared with the case when only the homogeneity in the field is taken into account. The approach is based on the spectral representation method, utilizes the fast Fourier transform, and generates simulations with random variability in both their amplitudes and phases, or in their phases only. Spatially variable seismic ground motions experiencing loss of coherence are generated as an example application of the developed approach.     

Probabilistic Material Degradation under High Temperature, Fatigue, and Creep

This paper describes the development of methodology that provides for quantification of uncertainty in the lifetime material strength degradation of structural components of aerospace propulsion systems subjected to a number of diverse random effects. The methodology is embodied in the two computer programs PROMISS and PROMISC. These programs form a material‐resistance model used in an aerospace structural‐reliability program NESSUS. A probabilistic material‐degradation model, in the form of a postulated randomized multifactor interaction equation, is used to quantify lifetime material strength. Each multiplicative term in the model has the property that if the current value of an effect equals the ultimate value, then the lifetime strength will be zero. Also, if the current value of an effect equals the reference value, the term equals one and lifetime strength is not affected by that particular effect. Presently, the model includes three effects that typically reduce lifetime strength: high temperature, mechanical fatigue, and creep. The paper also includes the statistical analysis of experimental data for INCONEL 718 obtained from the open literature. This statistical analysis of data provided regression parameters for use as the empirical material constants of the model, thus calibrating the model specifically for INCONEL 718. Model calibration was carried out for three variables, namely, high temperature, mechanical fatigue, and creep. Finally, using the PROMISS computer program, a sensitivity study was performed with the calibrated random model illustrating the effect of each variable upon random lifetime strength.     

Comparative efficacy of individual and group psychotherapy: A meta-analytic perspective.

Recent reviews of the group psychotherapy literature indicate that group is a beneficial and cost-effective treatment format. However, collective findings on the differential efficacy of group when compared with individual therapy remain problematic, incomplete, or controversial. To remedy this problem, the authors conducted a meta-analysis of 23 outcome studies that directly compared the effectiveness of the individual and group therapy formats when they were used within the same study. Results were consistent with previous reports that indicated no difference in outcome between the group and individual formats. This finding generally held true when client, therapist, methodology, treatment, and group variables were examined for possible relationship with effect sizes comparing group and individual therapy. Results bolster past findings that group therapy can be used as an efficacious cost-effective alternative to individual therapy under many different conditions. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

A nonparametric approach to a survival study with surrogate endpoints

A nonparametric estimator for the joint distribution of a survival time and surrogate response time, which may occur earlier during follow-up, is presented. In the absence of the surrogate response variable, the estimator reduces to the Kaplan Meier nonparametric estimator for the survival time alone. The estimator derived in this paper is done so in a particular novel way using an exchangeable process (reinforced random walks) to model individual observations. The methodology introduced in the paper is readily extended to modelling multiple state processes.