Detecting outlying trials in network meta-analysis.
Ontology highlight
ABSTRACT:
SUBMITTER: Zhang J
PROVIDER: S-EPMC4496319 | biostudies-literature | 2015 Aug
REPOSITORIES: biostudies-literature
Publications
Detecting outlying trials in network meta-analysis.
Zhang Jing J Fu Haoda H Carlin Bradley P BP
Statistics in medicine 20150408 19
Network meta-analysis (NMA) expands the scope of a conventional pairwise meta-analysis to simultaneously handle multiple treatment comparisons. However, some trials may appear to deviate markedly from the others and thus be inappropriate to be synthesized in the NMA. In addition, the inclusion of these trials in evidence synthesis may lead to bias in estimation. We call such trials trial-level outliers. To the best of our knowledge, while heterogeneity and inconsistency in NMA have been extensiv ...[more]
Similar Datasets
Project description:In this paper, we propose a method for automatically detecting the times during which game players exhibit specific behavior, such as when players commonly show excitement, concentration, immersion, and surprise. The proposed method detects such outlying behavior based on the game players' characteristics. These characteristics are captured non-invasively in a general game environment. In this paper, cameras were used to analyze observed data such as facial expressions and player movements. Moreover, multimodal data from the game players (i.e., data regarding adjustments to the volume and the use of the keyboard and mouse) was used to analyze high-dimensional game-player data. A support vector machine was used to efficiently detect outlying behaviors. We verified the effectiveness of the proposed method using games from several genres. The recall rate of the outlying behavior pre-identified by industry experts was approximately 70%. The proposed method can also be used for feedback analysis of various interactive content provided in PC environments.
| S-EPMC4690374 | biostudies-literature
Project description:Meta-analysis has become a widely used tool to combine results from independent studies. The collected studies are homogeneous if they share a common underlying true effect size; otherwise, they are heterogeneous. A fixed-effect model is customarily used when the studies are deemed homogeneous, while a random-effects model is used for heterogeneous studies. Assessing heterogeneity in meta-analysis is critical for model selection and decision making. Ideally, if heterogeneity is present, it should permeate the entire collection of studies, instead of being limited to a small number of outlying studies. Outliers can have great impact on conventional measures of heterogeneity and the conclusions of a meta-analysis. However, no widely accepted guidelines exist for handling outliers. This article proposes several new heterogeneity measures. In the presence of outliers, the proposed measures are less affected than the conventional ones. The performance of the proposed and conventional heterogeneity measures are compared theoretically, by studying their asymptotic properties, and empirically, using simulations and case studies.
| S-EPMC5106349 | biostudies-literature
Project description:In the absence of sufficient data directly comparing multiple treatments, indirect comparisons using network meta-analyses (NMAs) can provide useful information. Under current contrast-based (CB) methods for binary outcomes, the patient-centered measures including the treatment-specific event rates and risk differences (RDs) are not provided, which may create some unnecessary obstacles for patients to comprehensively trade-off efficacy and safety measures.We aim to develop NMA to accurately estimate the treatment-specific event rates.A Bayesian hierarchical model is developed to illustrate how treatment-specific event rates, RDs, and risk ratios (RRs) can be estimated. We first compare our approach to alternative methods using two hypothetical NMAs assuming a fixed RR or RD, and then use two published NMAs to illustrate the improved reporting.In the hypothetical NMAs, our approach outperforms current CB NMA methods in terms of bias. In the two published NMAs, noticeable differences are observed in the magnitude of relative treatment effects and several pairwise statistical significance tests from previous report.First, to facilitate the estimation, each study is assumed to hypothetically compare all treatments, with unstudied arms being missing at random. It is plausible that investigators may have selected treatment arms on purpose based on the results of previous trials, which may lead to 'nonignorable missingness' and potentially bias our estimates. Second, we have not considered methods to identify and account for potential inconsistency between direct and indirect comparisons.The proposed NMA method can accurately estimate treatment-specific event rates, RDs, and RRs and is recommended.
| S-EPMC3972291 | biostudies-literature
Project description:BackgroundNetwork meta-analysis (NMA), a generalization of conventional MA, allows for assessing the relative effectiveness of multiple interventions. Reporting bias is a major threat to the validity of MA and NMA. Numerous methods are available to assess the robustness of MA results to reporting bias. We aimed to extend such methods to NMA.MethodsWe introduced 2 adjustment models for Bayesian NMA. First, we extended a meta-regression model that allows the effect size to depend on its standard error. Second, we used a selection model that estimates the propensity of trial results being published and in which trials with lower propensity are weighted up in the NMA model. Both models rely on the assumption that biases are exchangeable across the network. We applied the models to 2 networks of placebo-controlled trials of 12 antidepressants, with 74 trials in the US Food and Drug Administration (FDA) database but only 51 with published results. NMA and adjustment models were used to estimate the effects of the 12 drugs relative to placebo, the 66 effect sizes for all possible pair-wise comparisons between drugs, probabilities of being the best drug and ranking of drugs. We compared the results from the 2 adjustment models applied to published data and NMAs of published data and NMAs of FDA data, considered as representing the totality of the data.ResultsBoth adjustment models showed reduced estimated effects for the 12 drugs relative to the placebo as compared with NMA of published data. Pair-wise effect sizes between drugs, probabilities of being the best drug and ranking of drugs were modified. Estimated drug effects relative to the placebo from both adjustment models were corrected (i.e., similar to those from NMA of FDA data) for some drugs but not others, which resulted in differences in pair-wise effect sizes between drugs and ranking.ConclusionsIn this case study, adjustment models showed that NMA of published data was not robust to reporting bias and provided estimates closer to that of NMA of FDA data, although not optimal. The validity of such methods depends on the number of trials in the network and the assumption that conventional MAs in the network share a common mean bias mechanism.
| S-EPMC3537713 | biostudies-literature
Project description:We present use of a simple formula to calculate the number of pairwise comparisons of interventions within a single trial or network meta-analyses. We used the data from our previous network meta-analysis to build a study-based register and enumerated the direct pairwise comparisons from the trials therein. We then compared this with the number of comparisons predicted by use of the formula and finally with the reported number of comparisons (indirect or direct) within the network meta-analysis. A total of 133 trials of 8 interventions were selected which included 163 comparisons. The network of these showed 16 unique direct comparisons. The formula predicted an expected 28 indirect or direct comparisons and this is the number that were indeed reported. The formula produces an accurate enumeration of the potential comparisons within a single trial or network meta-analysis. Its use could help transparency of reporting should a shortfall occur between comparisons actually used and the potential total.
| S-EPMC6402080 | biostudies-literature
Project description:BackgroundProstate cancer (PCa) is common, with it being the 2nd most prevalent cancer in men worldwide and the 6th leading cause of death in men. Screening for any type of cancer aims to increase the chances of successful treatment through early detection of the disease. There were some systematic reviews (SRs) evaluated the diagnostic value of biomarkers for the diagnosis of PCa and no studies have been conducted to analyze the quality of these SRs. We are not clear which kind of marker is the best choice. Thus, this study aims to assess the methodologic quality of the SRs and reanalyze the published data based on SRs for the biomarkers to find the optimal biomarker for the early diagnosis of PCa.MethodsWe performed a systematic literature search of PubMed, Embase, Web of Science, and Cochrane Library and to identify relevant SRs from inception to April 2019. Diagnostic accuracy studies included any type of single biomarker or combined biomarkers aimed at evaluating the diagnostic value is considered eligible for this overview. The Assessment of Multiple Systematic Reviews-2 (AMSTAR-2) instrument will be used to evaluate the risk of bias of the included SRs. Standard pairwise meta-analysis and adjusted indirect comparison will be used to compare the diagnostic value of different biomarkers.ResultsThe results of this study will be submitted to a peer-reviewed journal for publication.ConclusionThis study will reanalyze the published data based on SRs. We hope that the results will help find a biomarker with the superior diagnostic performance for the diagnosis of PCa.Prospero registration numberCRD42019125880.
| S-EPMC6708917 | biostudies-literature
Project description:Bariatric surgery has proved to be an effective strategy in treating obesity. However, randomized controlled trials (RCTs) of 3 most common bariatric surgery procedures, Roux-en-Y gastric bypass (RYGB), sleeve gastrectomy (SG), and laparoscopic adjustable gastric band (LAGB), reported inconsistent results. We performed a systematic review and network meta-analysis to synthesize evidence of effectiveness of the 3 common bariatric procedures from relevant RCTs.The present study was a systematic review and network meta-analysis of RCTs. All RCTs must meet the following criteria to be included in the analysis: patients with body mass index (BMI) ?30?kg/m, reported at least 1 outcome of interest, compared at least 2 of the 3 bariatric procedures, and had follow-ups of at least 1 year. Primary outcome was weight loss, expressed as differences in mean BMI reduction and percentage excess weight loss (%EWL) following 1 year after the surgery. Network meta-analysis was based on Bayesian framework with Markov Chain Monte Carlo simulation approach.Eleven RCTs that met the criteria were included in the review. Of 9 trials (n?=?765), the differences in mean BMI reduction were -0.76 (95% CI: -3.1 to 1.6) for RYGB versus SG, -5.8 (95% CI: -9.2 to -2.4) for RYGB versus LAGB, and -5.0 (95% CI: -9.0 to -1.0) for SG versus LAGB. Eight RCTs (n?=?656) reported percentage excess weight-loss (%EWL), the mean differences between RYGB and SG, RYGB and LAGB, and SG and LAGB were 3.8% (95% CI: -8.5% to 13.8%), -22.2% (95% CI: -34.7% to -6.5%), and -26.0% (95% CI: -40.6% to -6.4%), respectively. The meta-analysis indicated low heterogeneity between studies, and the node splitting analysis showed that the studies were consistent between direct and indirect comparisons (P?>?.05).The RYGB and SG yielded similar in weight-loss effect and both were superior to LAGB. Other factors such as complications and patient preference should be considered during surgical consultations.
| S-EPMC5704829 | biostudies-other
Project description:AIMS:Several antihypertensive drugs are used in the treatment of severe hypertension in pregnancy. The present study is a network meta-analysis comparing the efficacy and safety of these drugs. METHODS:Electronic databases were searched for randomized clinical trials comparing drugs used in the treatment of severe hypertension in pregnancy. The number of women achieving the target blood pressure (BP) was the primary outcome. Doses required and time taken for achieving the target BP, failure rate, and incidences of maternal tachycardia, palpitation, hypotension, headache, and neonatal death and stillbirth were the secondary outcomes. Mixed treatment comparison pooled estimates were generated using a random-effects model. Odds ratios for the categorical and mean difference for the numerical outcomes were the effect estimates. RESULTS:Fifty-one studies were included in the systematic review and 46 in the meta-analysis. No significant differences in the number of patients achieving target BP was observed between any of the drugs. Diazoxide [-15 (-20.6, -9.4)], nicardipine [-11.8 (-22.3, -1.2)], nifedipine/celastrol [-19.3 (-27.4, -11.1)], nifedipine/vitamin D [-17.1 (-25.7, -9.7)], nifedipine/resveratrol [-13.9 (-22.6, -5.2)] and glyceryl trinitrate [-33.8 (-36.7, -31)] were observed to achieve the target BP (in minutes) more rapidly than hydralazine. Nifedipine required fewer doses than hydralazine for achieving the target BP. Glyceryl trinitrate and labetalol were associated with fewer incidences of tachycardia and palpitation respectively than hydralazine. Trial sequential analysis concluded adequate evidence for hydralazine and nifedipine compared with labetalol. Moderate quality of evidence was observed for direct comparison estimate between labetalol and hydralazine but was either low or very low for other comparisons. CONCLUSION:The present evidence suggests similar efficacy between nifedipine, hydralazine and labetalol in the treatment of severe hypertension in pregnancy. Subtle differences may exist in their safety profile. The evidence is inadequate for other drugs.
| S-EPMC6089822 | biostudies-other
Project description:Water-assisted colonoscopy (water exchange [WE] and water immersion [WI]) has been shown to improve the adenoma detection rate. However, few studies have compared these two methods head-to-head. Thus, we conducted a network meta-analysis to integrate both direct and indirect evidence comparing the effectiveness of these two procedures.We searched PubMed, Web of Science, and the Cochrane Central Register of Controlled Trials for original papers and abstracts published up to March 2018. Randomized controlled trials (RCTs) reporting data in accordance with the eligibility criteria were included in this study. We performed a Bayesian random effects network meta-analysis with mixed comparisons.Twenty-nine studies (n = 11464 patients) including 6 direct and 23 indirect comparisons were included in this network meta-analysis. There was a statistically significant difference in the efficacy of adenoma detection when WE was compared with WI (risk ratio [RR]: 1.2, 95% credible interval [CrI]: 1.1-1.3), air insufflation (AI; RR: 1.3, 95% CrI: 1.1-1.4), and carbon dioxide (CO2) insufflation (RR: 1.2, 95% CrI: 1.1-1.5). The different methods were ranked in order from the most to least effective in adenoma detection as follows: WE, WI, AI, and CO2. Moreover, although there were no significant differences in pain scores, willingness to repeat, caecal intubation rate, or total procedure time between WI and WE colonoscopy, WE required a longer caecal intubation time than WI.This network meta-analysis supposes that WE may be superior to WI in detecting adenomas during colonoscopies without affecting other technical features or patient acceptance.
| S-EPMC6078142 | biostudies-other
Project description:BackgroundIndirect comparisons of competing treatments by network meta-analysis (NMA) are increasingly in use. Reporting bias has received little attention in this context. We aimed to assess the impact of such bias in NMAs.MethodsWe used data from 74 FDA-registered placebo-controlled trials of 12 antidepressants and their 51 matching publications. For each dataset, NMA was used to estimate the effect sizes for 66 possible pair-wise comparisons of these drugs, the probabilities of being the best drug and ranking the drugs. To assess the impact of reporting bias, we compared the NMA results for the 51 published trials and those for the 74 FDA-registered trials. To assess how reporting bias affecting only one drug may affect the ranking of all drugs, we performed 12 different NMAs for hypothetical analysis. For each of these NMAs, we used published data for one drug and FDA data for the 11 other drugs.FindingsPair-wise effect sizes for drugs derived from the NMA of published data and those from the NMA of FDA data differed in absolute value by at least 100% in 30 of 66 pair-wise comparisons (45%). Depending on the dataset used, the top 3 agents differed, in composition and order. When reporting bias hypothetically affected only one drug, the affected drug ranked first in 5 of the 12 NMAs but second (n?=?2), fourth (n?=?1) or eighth (n?=?2) in the NMA of the complete FDA network.ConclusionsIn this particular network, reporting bias biased NMA-based estimates of treatments efficacy and modified ranking. The reporting bias effect in NMAs may differ from that in classical meta-analyses in that reporting bias affecting only one drug may affect the ranking of all drugs.
| S-EPMC3335054 | biostudies-literature