Gerta Rücker

Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials.

Funnel plots, and tests for funnel plot asymmetry, have been widely used to examine bias in the results of meta-analyses. Funnel plot asymmetry should not be equated with publication bias, because it has a number of other possible causes. This article describes how to interpret funnel plot asymmetry, recommends appropriate tests, and explains the implications for choice of meta-analysis model

Low‐dose aspirin for secondary cardiovascular prevention – cardiovascular risks after its perioperative withdrawal versus bleeding risks with its continuation – review and meta‐analysis

Objectives.  Low‐dose aspirin given for secondary prevention of cardiovascular disease is frequently withdrawn prior to surgical or diagnostic procedures to reduce bleeding complications. This may expose patients to increased cardiovascular morbidity and mortality. Aim of the study was to review and quantify cardiovascular risks because of periprocedural aspirin withdrawal and bleeding risks with the continuation of aspirin.

y-Randomization and its variants in QSPR/QSAR

y-Randomization is a tool used in validation of QSPR/QSAR models, whereby the performance of the original model in data description (r2) is compared to that of models built for permuted (randomly shuffled) response, based on the original descriptor pool and the original model building procedure. We compared y-randomization and several variants thereof, using original response, permuted response, or random number pseudoresponse and original descriptors or random number pseudodescriptors, in the typical setting of multilinear regression (MLR) with descriptor selection. For each combination of number of observations (compounds), number of descriptors in the final model, and number of descriptors in the pool to select from, computer experiments using the same descriptor selection method result in two different mean highest random r2 values. A lower one is produced by y-randomization or a variant likewise based on the original descriptors, while a higher one is obtained from variants that use random number pseudodescriptors. The difference is due to the intercorrelation of real descriptors in the pool. We propose to compare an original models r2 to both of these whenever possible. The meaning of the three possible outcomes of such a double test is discussed. Often y-randomization is not available to a potential user of a model, due to the values of all descriptors in the pool for all compounds not being published. In such cases random number experiments as proposed here are still possible. The test was applied to several recently published MLR QSAR equations, and cases of failure were identified. Some progress also is reported toward the aim of obtaining the mean highest r2 of random pseudomodels by calculation rather than by tedious multiple simulations on random number variables.

Pharmacotherapy for borderline personality disorder: Cochrane systematic review of randomised trials

BACKGROUND Many patients with borderline personality disorder receive pharmacological treatment, but there is uncertainty about the usefulness of such therapies. AIMS To evaluate the evidence of effectiveness of pharmacotherapy in treating different facets of the psychopathology of borderline personality disorder. METHOD A Cochrane Collaboration systematic review and meta-analysis of randomised comparisons of drug v. placebo, drug v. drug, or single drug v. combined drug treatment in adult patients with borderline personality disorder was conducted. Primary outcomes were overall disorder severity as well as specific core symptoms. Secondary outcomes comprised associated psychiatric pathology and drug tolerability. RESULTS Twenty-seven trials were included in which first- and second-generation antipsychotics, mood stabilisers, antidepressants and omega-3 fatty acids were tested. Most beneficial effects were found for the mood stabilisers topiramate, lamotrigine and valproate semisodium, and the second-generation antipsychotics aripiprazole and olanzapine. However, the robustness of findings is low, since they are based mostly on single, small studies. Selective serotonin reuptake inhibitors so far lack high-level evidence of effectiveness. CONCLUSIONS The current evidence from randomised controlled trials suggests that drug treatment, especially with mood stabilisers and second-generation antipsychotics, may be effective for treating a number of core symptoms and associated psychopathology, but the evidence does not currently support effectiveness for overall severity of borderline personality disorder. Pharmacotherapy should therefore be targeted at specific symptoms.

Why add anything to nothing? The arcsine difference as a measure of treatment effect in meta-analysis with zero cells.

For clinical trials with binary endpoints there are a variety of effect measures, for example risk difference, risk ratio and odds ratio (OR). The choice of metric is not always straightforward and should reflect the clinical question. Additional issues arise if the event of interest is rare. In systematic reviews, trials with zero events in both arms are encountered and often excluded from the meta-analysis.The arcsine difference (AS) is a measure which is rarely considered in the medical literature. It appears to have considerable promise, because it handles zeros naturally, and its asymptotic variance does not depend on the event probability.This paper investigates the pros and cons of using the AS as a measure of intervention effect. We give a pictorial representation of its meaning and explore its properties in relation to other measures. Based on analytical calculation of the variance of the arcsine transformation, a more conservative variance estimate for the rare event setting is proposed. Motivated by a published meta-analysis in cardiac surgery, we examine the statistical properties of the various metrics in the rare event setting.We find the variance estimate of the AS to be more stable than that of the log-OR, even if events are rare. However, parameter estimation is biased if the groups are markedly unbalanced. Though, from a theoretical viewpoint, the AS is a natural choice, its practical use is likely to continue to be limited by its less direct interpretation.

ON TOPOLOGICAL INDICES, BOILING POINTS, AND CYCLOALKANES

The experimental boiling points (bp) of saturated hydrocarbons (acyclic through polycyclic) up to decanes are systematically compiled. The bp values are classified into groups of lower or higher reliability according to the accuracy and frequency with which they were reproduced by independent researchers. For each hydrocarbon structure the values of several simple topological indices (TI) of widely varying origin are given, including the values of molecular walk counts of various lengths and their sum. The sensitivity of the TIs for structural changes within comprehensive groups of cyclic saturated hydrocarbons is evaluated, and the total walk count is found to be most sensitive. By multilinear regression structure−bp correlations are obtained for various comprehensive compound samples. Both the detour index and the walk counts are found to play a major role in the best models. Comparison of the bp models obtained with those from the recent literature reveals significant improvements for both cyclic and a...

Network meta‐analysis, electrical networks and graph theory

Network meta-analysis is an active field of research in clinical biostatistics. It aims to combine information from all randomized comparisons among a set of treatments for a given medical condition. We show how graph-theoretical methods can be applied to network meta-analysis. A meta-analytic graph consists of vertices (treatments) and edges (randomized comparisons). We illustrate the correspondence between meta-analytic networks and electrical networks, where variance corresponds to resistance, treatment effects to voltage, and weighted treatment effects to current flows. Based thereon, we then show that graph-theoretical methods that have been routinely applied to electrical networks also work well in network meta-analysis. In more detail, the resulting consistent treatment effects induced in the edges can be estimated via the Moore-Penrose pseudoinverse of the Laplacian matrix. Moreover, the variances of the treatment effects are estimated in analogy to electrical effective resistances. It is shown that this method, being computationally simple, leads to the usual fixed effect model estimate when applied to pairwise meta-analysis and is consistent with published results when applied to network meta-analysis examples from the literature. Moreover, problems of heterogeneity and inconsistency, random effects modeling and including multi-armed trials are addressed. Copyright

Differential Effectiveness of Placebo Treatments: A Systematic Review of Migraine Prophylaxis

IMPORTANCE When analyzing results of randomized clinical trials, the treatment with the greatest specific effect compared with its placebo control is considered to be the most effective one. Although systematic variations of improvements in placebo control groups would have important implications for the interpretation of placebo-controlled trials, the knowledge base on the subject is weak. OBJECTIVE To investigate whether different types of placebo treatments are associated with different responses using the studies of migraine prophylaxis for this analysis. DESIGN, SETTING, AND PARTICIPANTS We searched relevant sources through February 2012 and contacted the authors to identify randomized clinical trials on the prophylaxis of migraine with an observation period of at least 8 weeks after randomization that compared an experimental treatment with a placebo control group. We calculated pooled random-effects estimates according to the type of placebo for the proportions of treatment response. We performed meta-regression analyses to identify sources of heterogeneity. In a network meta-analysis, direct and indirect comparisons within and across trials were combined. Additional analyses were performed for continuous outcomes. EXPOSURE Active migraine treatment and the placebo control conditions. MAIN OUTCOMES AND MEASURES Proportion of treatment responders, defined as having an attack frequency reduction of at least 50%. Other available outcomes in order of preference included a reduction of 50% or greater in migraine days, the number of headache days, or headache score or a significant improvement as assessed by the patients or their physicians. RESULTS Of the 102 eligible trials, 23 could not be included in the meta-analyses owing to insufficient data. Sham acupuncture (proportion of responders, 0.38 [95% CI, 0.30-0.47]) and sham surgery (0.58 [0.37-0.77]) were associated with a more pronounced reduction of migraine frequency than oral pharmacological placebos (0.22 [0.17-0.28]) and were the only significant predictors of response in placebo groups in multivariable analyses (P = .005 and P = .001, respectively). Network meta-analysis confirmed that more patients reported response in sham acupuncture groups than in oral pharmacological placebo groups (odds ratio, 1.88 [95% CI, 1.30-2.72]). Corresponding analyses for continuous outcomes showed similar findings. CONCLUSIONS AND RELEVANCE Sham acupuncture and sham surgery are associated with higher responder ratios than oral pharmacological placebos. Clinicians who treat patients with migraine should be aware that a relevant part of the overall effect they observe in practice might be due to nonspecific effects and that the size of such effects might differ between treatment modalities.

The relationship between quality of research and citation frequency

BackgroundCitation counts are often regarded as a measure of the utilization and contribution of published articles. The objective of this study is to assess whether statistical reporting and statistical errors in the analysis of the primary outcome are associated with the number of citations received.MethodsWe evaluated all original research articles published in 1996 in four psychiatric journals. The statistical and reporting quality of each paper was assessed and the number of citations received up to 2005 was obtained from the Web of Science database. We then examined whether the number of citations was associated with the quality of the statistical analysis and reporting.ResultsA total of 448 research papers were included in the citation analysis. Unclear or inadequate reporting of the research question and primary outcome were not statistically significantly associated with the citation counts. After adjusting for journal, extended description of statistical procedures had a positive effect on the number of citations received. Inappropriate statistical analysis did not affect the number of citations received. Adequate reporting of the primary research question, statistical methods and primary findings were all associated with the journal visibility and prestige.ConclusionIn this cohort of published research, measures of reporting quality and appropriate statistical analysis were not associated with the number of citations. The journal in which a study is published appears to be as important as the statistical reporting quality in ensuring dissemination of published medical science.

Ranking treatments in frequentist network meta-analysis works without resampling methods

BackgroundNetwork meta-analysis is used to compare three or more treatments for the same condition. Within a Bayesian framework, for each treatment the probability of being best, or, more general, the probability that it has a certain rank can be derived from the posterior distributions of all treatments. The treatments can then be ranked by the surface under the cumulative ranking curve (SUCRA). For comparing treatments in a network meta-analysis, we propose a frequentist analogue to SUCRA which we call P-score that works without resampling.MethodsP-scores are based solely on the point estimates and standard errors of the frequentist network meta-analysis estimates under normality assumption and can easily be calculated as means of one-sided p-values. They measure the mean extent of certainty that a treatment is better than the competing treatments.ResultsUsing case studies of network meta-analysis in diabetes and depression, we demonstrate that the numerical values of SUCRA and P-Score are nearly identical.ConclusionsRanking treatments in frequentist network meta-analysis works without resampling. Like the SUCRA values, P-scores induce a ranking of all treatments that mostly follows that of the point estimates, but takes precision into account. However, neither SUCRA nor P-score offer a major advantage compared to looking at credible or confidence intervals.