Standard Random Effects Model Research Articles

Role of immunotherapy in gastroesophageal cancer with liver metastasis: A meta-analysis of phase III randomized clinical trials.

319 Background: Recent studies have demonstrated a poor response to immunotherapy in patients with colorectal cancer with liver metastasis. The role of immune checkpoint inhibitors (ICIs) for patients with gastroesophageal (GE) cancer with liver metastasis remains unclear. Our objective was to investigate if ICIs are beneficial in patients with GE cancer with liver metastasis. Methods: We searched PubMed, Embase, ESMO, and ASCO Meeting Abstracts for phase III randomized clinical trials (RCTs) testing ICIs in metastatic/advanced GE cancer from 2017 to 2023. All phase III RCTs of metastatic GE cancer treated with ICIs that had available data on liver metastasis were analyzed. A meta-analysis was conducted on the survival outcomes (overall and progression-free survival), where hazard ratios were obtained, with 95% confidence intervals, using the standard random effects model. To assess the heterogeneity of the study outcomes, the I2 statistic was obtained. The I2 statistic represented the variability in the meta-analysis attributed to study heterogeneity. These analyses were applied to the overall cohort of studies, as well as within the liver metastasis (yes or no) sub-cohorts. All analyses were conducted in SAS v9.4 (Cary, NC) at a significance level of 0.05. Results: Seven Phase III RCTs were includedthat had available data on overall survival and/or progressive free survival in patients with liver metastasis. Of these, six were in the frontline setting and one in the later line. Overall, 6,109 patients were included. Overall survival was similar among all patients (HR 0.72 [0.67,0.77], p<0.001, I2=0.0%), patients with no liver metastasis (HR 0.73 [0.67,0.81], p<0.001, I2=0.0%), and patients with liver metastasis (HR 0.74 [0.67,0.81], p<0.001, I2=0.0%), indicating that ICIs’ benefit on OS is consistent. Progression-free survival (PFS) was also similar among all patients (HR 0.63 [0.57,0.70], p<0.001, I2=54.7%); those with no liver metastasis (HR 0.62 [0.51,0.76], p<0.001, I2=62.3%), and with liver metastasis (HR 0.66 [0.57,0.76], p<0.001, I2-0.0%). Funnel plots were used to identify a potential bias for studies included in the analysis and showed no significant sources of bias for OS and PFS. Conclusions: In this meta-analysis, ICIs were beneficial in patients with GE cancer with and without liver metastasis. However, the majority of the trials were in the frontline setting where ICIs were combined with chemotherapy. Further subgroup analyses should be performed with individual patient-level data to identify patients who respond best to ICIs.

Predictors of Complete Pathological Response with Chemoimmunotherapy in Triple-Negative Breast Cancer: A Meta-Analysis

Background: Multiple randomized controlled trials (RCTs) have investigated the impact of adding checkpoint inhibitors to neoadjuvant chemotherapy for triple-negative breast cancer (TNBC) patients. However, there is a lack of biomarkers that can help identify patients who would benefit from combination therapy. Our research identifies response predictors and assesses the effectiveness of adding immunotherapy to neoadjuvant chemotherapy for TNBC patients. Methods: We identified eligible RCTs by searching PubMed, Cochrane CENTRAL, Embase, and oncological meetings. For this meta-analysis, we obtained odds ratios using the standard random effects model. To assess the heterogeneity of the study outcomes, the I2 statistic was obtained. Potential bias was assessed using a funnel plot and the corresponding Egger’s test. Results: In total, 1637 patients with TNBC were included from five RCTs. Neoadjuvant chemoimmunotherapy significantly improved pCR when compared to neoadjuvant chemotherapy alone. In the subgroup analysis, neoadjuvant chemoimmunotherapy showed higher pCR rates in both Programmed death-ligand 1 (PD-L1)-positive and PD-L1-negative TNBC patients. An Eastern Cooperative Oncology Group (ECOG) performance score (PS) of 0 correlated with increased pCRs (OR = 1.9, p < 0.001) in neoadjuvant chemoimmunotherapy vs. neoadjuvant chemotherapy, but no benefit was observed for patients with ECOG PS 1. Nodal positivity was significantly associated with pCR (OR = 2.52, p < 0.001), while neoadjuvant chemoimmunotherapy did not benefit patients with negative lymph nodes. Conclusions: Checkpoint inhibition and neoadjuvant chemotherapy significantly increased pCRs in TNBC patients, regardless of their PDL-1 status. Additional checkpoint inhibitors improved pCR rates, mainly for patients with ECOG PS 0 and lymph node-positive disease.

The impact of multidisciplinary cancer conferences on overall survival: a meta-analysis.

Multidisciplinary cancer conferences consist of regular meetings between diverse specialists working together to share clinical decision making in cancer care. The aim of this study was to systematically review and meta-analyze the effect of multidisciplinary cancer conference intervention on the overall survival of patients with cancer. A systematic literature search was conducted on Ovid MEDLINE, EMBASE, and the Cochrane Controlled Register of Trials for studies published up to July 2023. Studies reporting on the impact of multidisciplinary cancer conferences on patient overall survival were included. A standard random-effects model with the inverse variance-weighted approach was used to estimate the pooled hazard ratio of mortality (multidisciplinary cancer conference vs non-multidisciplinary cancer conference) across studies, and the heterogeneity was assessed by I2. Publication bias was examined using funnel plots and the Egger test. A total of 134 287 patients with cancer from 59 studies were included in our analysis, with 48 467 managed by multidisciplinary cancer conferences and 85 820 in the control arm. Across all cancer types, patients managed by multidisciplinary cancer conferences had an increased overall survival compared with control patients (hazard ratio = 0.67, 95% confidence interval = 0.62 to 0.71, I2 = 84%). Median survival time was 30.2 months in the multidisciplinary cancer conference group and 19.0 months in the control group. In subgroup analysis, a positive effect of the multidisciplinary cancer conference intervention on overall survival was found in breast, colorectal, esophageal, hematologic, hepatocellular, lung, pancreatic, and head and neck cancer. Overall, our meta-analysis found a significant positive effect of multidisciplinary cancer conferences compared with controls. Further studies are needed to establish nuanced guidelines when optimizing multidisciplinary cancer conference integration for treating diverse patient populations.

Peer Pressure, Information Technology Adoption, and Bank Performance

Manuscript type: Research paper Research aims: The study examines cross-sectional dependence among banks to invest in information technology-related assets. We also analyse the short- and long-term impacts of information technology (IT) adoption on banks’ profitability and the variations in effects between early and late adopters. Design/Methodology/Approach: The study uses autoregressive distributed lag (ARDL) with dynamic fixed approach, standard fixed-effect ordinary least square and random effect model with double clustering regressions. Research findings: The study finds a positive effect of peer pressure on a bank’s investment in IT assets. It also documents that IT adoption reduces short-term profitability, but the joint effect of early adoption increases both short- and long-term profitability. Theoretical contribution/Originality: The study contributes to the banking and technology adoption literature by showing the evidence on a positive and significant effect of peer pressure on IT adoption. Practitioner/Policy implications: The findings are significant for the banking sector’s policy makers in emerging economies. Research limitation: The findings may not be applicable in the context of a developed economy with a strong IT infrastructure.

The impact of previous extrapulmonary malignancies on survival of surgically treated primary lung cancer: a comprehensive systematic review and meta-analysis.

Whether a history of previous extrapulmonary malignancies had an impact on the survival of surgically treated primary lungcancer patients remains unknown. To better answer this question, we, therefore, conducted the first meta-analysis to compare the survival of lung cancer patients after surgical resection with previous extrapulmonary malignancies and those without. We systematically searched PubMed, Embase, and the Cochrane Library to identify relevant studies up to April 1, 2022. Data for analysis were obtained directly from the text results or calculated from the Kaplan-Meier survival curve, which mainly included 5-year overall survival (OS) and recurrence-free survival (RFS). Pooled hazard ratios (HRs) and 95% confidence intervals (CIs) were analyzed by the fixed-effect model test or standard random-effect model test depending on the heterogeneity of the included studies. The Q-test and I2-test were used to assess heterogeneity. Sensitivity analysis was performed to examine the stability of the pooled results. We finally included 7 retrospective studies consisting of 19,723 surgically treated primary lungcancer patients with (7.7%) or without (92.3%) previous extrapulmonary malignancies. The final results showed that lung patients with previous extrapulmonary malignancies had an inferior OS (HR 1.18, 95% CI [1.07, 1.31], P = 0.001) than those without, but no significant difference in RFS (HR 1.15, 95% CI [0.89-1.47], P = 0.29) was observed between the two groups. Moreover, when only focusing on stage-I primary lung cancer patients with or without previous extrapulmonary malignancies, the results still held true (OS: HR 1.39, 95% CI [1.04, 1.85], P = 0.02; RFS: HR 1.10, 95% CI [0.82-1.49], P = 0.51, respectively). No significant heterogeneities or publication bias were observed among these studies. Our meta-analysis showed that surgically treated primary lungcancer patients with a previous extrapulmonary cancer history had a worse overall survival than those without. However, a history of previous extrapulmonary cancer was not associated with postsurgical lung cancer recurrence, and more large-scale prospective studies should be encouraged to update our conclusions.

Therapeutic Efficacy of Melatonin in Patients with Coronavirus 2019: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

The efficacy of melatonin in the treatment of patients with coronavirus 2019 (COVID-19) is controversial. This review has summarized the evidence on the efficacy of oral melatonin compared to placebo in patients with mild to moderate COVID-19 infection. We searched four international online databases and all randomized clinical trials (RCTs) that investigated the effects of melatonin compared with the placebo on clinical outcomes, including mortality, discharge time, O2 saturation (SaO2), and c-reactive protein (CRP) levels in patients with COVID-19 infection, were included. The standard random-effects model with hybrid continuity correction was used to pool the risk ratio (RR), weighted mean difference (WMD), and the I2 index to assess the heterogeneity. A total of 272 patients from five RCTs were included. Our meta-analysis showed melatonin compared to placebo, decreased discharge time (WMD=-0.93 days; 95% confidence interval [CI]:-2.94 to 1.07, P=0.36; I2=56.78%) and the risk of mortality (RR=0.72; 95% CI:0.25 to 2.13, P=0.56; I2=0.0%) in COVID-19 patients. Melatonin intake compared to placebo significantly increased SaO2 (WMD=1.38%; 95% CI:0.09 to 2.68, P=0.04; I2=49.82%) and decreased the CRP levels (WMD=-7.24 mg/l; 95% CI:-11.28 to -3.21, P<0.001) in a sensitivity analysis. Our findings showed the efficacy of melatonin compared to placebo in patients with mild to moderate COVID-19 infection.

S588 Cap Assisted Endoscopic Treatment of Esophageal Food Bolus and/or Foreign Body Ingestion: A Systematic Review and Meta-Analysis

Introduction: Esophageal food bolus and/ or foreign body (FB) impaction is one of the most common gastrointestinal emergencies where early endoscopic intervention is warranted. Early time to intervention with faster procedure time are associated with good patient outcomes. In recent literature, cap assisted endoscopic removal of esophageal FB seems to be associated with favorable outcomes as compared to conventional methods. This meta-analysis reports on the pooled outcomes of cap-assisted endoscopic removal of esophageal FB. Methods: We conducted a comprehensive search of several databases including Medline, Scopus, Embase, from inception to February 2022, to identify studies reporting on the use of cap in endoscopic treatment of esophageal FB. Standard meta-analysis methods and random effects model was used to calculate the pooled odds-ratio (OR) and mean-difference (MD) with corresponding 95% confidence intervals (CI). I2% statistics was used to assess the heterogeneity. Results: Six studies were included that evaluated 677 patients treated with cap-assistance and 694 patients treated with conventional methods. Peptic stricture (32%) and eosinophilic esophagitis (20%) were the most common esophageal disorders. Cap-assisted method demonstrated statistically significant pooled OR for technical success: 7.1 (95% CI [1.9-26.9, p=0.004, I 2%=0]), and enbloc removal: 26.6 [17.6-40.2, p< 0.001, I2%=0], with significantly faster procedure time, MD: -4.6min [-6.5to-2.8, p< 0.001, I2%=99]; when compared to conventional method. Better technical success was achieved with cap-assisted method done under general-anesthesia (OR: 8.7, 1.6-47.7, p=0.01, I 2%=0), but faster procedure time was noted with cap-assisted method performed without general-anesthesia (MD: -1.5, -2.7to-0.4, p=0.01, I2%=99). Pooled adverse events were comparable. However, pooled OR of mucosal tear was significantly less with the use of cap in food-bolus impactions (OR: 0.07, 0.01-0.38, p=0.02, I2%=0). No airway related adverse events were reported when general anesthesia was not used. (Figure) Conclusion: This meta-analysis demonstrates significantly better pooled technical success, enbloc removal and faster procedure time with cap-assisted endoscopic removal of esophageal food-bolus and/ or foreign-body impaction as compared to conventional method. With cap-assistance, faster procedure time seems achievable without the need for general anesthesia. However, additional randomized controlled trials are required to further validate this finding.Figure 1.: Summary of pooled results

Beta-Blocker Therapy in Patients With Severe Traumatic Brain Injury: A Systematic Review and Meta-Analysis.

Traumatic brain injury (TBI), a leading cause of morbidity and mortality among trauma patients worldwide, poses the risk of secondary neurological insult due to significant catecholamine surge. We aim to investigate the effectiveness and outcomes of beta-blocker administration in patients with severe TBI. A search through PubMed, EMBASE, JAMA network, and Google Scholar databases was conducted for relevant peer-reviewed original studies published before February 15, 2022. A standard random-effects model was used, as justified by a high Cohen's Q test. Twelve studies met inclusion criteria and were included in the meta-analysis. Severe TBI patients who were administered beta-blockers had a significantly reduced incidence of in-hospital mortality compared to the non-beta-blocker group (14.5% vs 19.2%). However, the beta-blocker group was reported to have a significantly greater number of ventilator days (5.58 vs 2.60days). Similarly, intensive care unit (9.00 vs 6.84days) and hospital (17.30 vs 11.02days) lengths of stay (LOS) were increased in the beta-blocker group compared to those who were not administered beta-blocker therapy, but only the difference in hospital-LOS was significant. Beta-blockers have significantly decreased in-hospital mortality in patients with severe TBI despite being associated with an increase in ventilator days and hospital-LOS. The administration of beta-blocker therapy in the management of severe TBI may be warranted and should be discussed in future guidelines.

An assessment of statistical methods for non-independent data in ecological meta-analyses: Reply.

In ecological meta-analyses, nonindependence among observed effect sizes from the same source paper is common. If not accounted for, nonindependence can seriously undermine inferences. We compared the performance of four meta-analysis methods that attempt to address such nonindependence and the standard random-effect model that ignores nonindependence. We simulated data with various types of within-paper nonindependence, and assessed the standard deviation of the estimated mean effect size and Type I error rate of each method. Although all four methods performed substantially better than the standard random-effects model that assumes independence, there were differences in performance among the methods. A two-step method that first summarizes the multiple observed effect sizes per paper using a weighted mean and then analyzes the reduced data in a standard random-effects model, and a robust variance estimation method performed consistently well. A hierarchical model with both random paper and study effects gave precise estimates but had a higher Type I error rates, possibly reflecting limitations of currently available meta-analysis software. Overall, we advocate the use of the two-step method with a weighted paper mean and the robust variance estimation method as reliable ways to handle within-paper nonindependence in ecological meta-analyses.

An assessment of statistical methods for nonindependent data in ecological meta-analyses: Comment.

Recently, Song et al. (2020) conducted a simulation study using different methods to deal with non-independence resulting from effect sizes originating from the same paper - a common occurrence in ecological meta-analyses. The main methods that were of interest in their simulations were: 1) a standard random-effects model used in combination with a weighted average effect size for each paper (i.e., a two-step method), 2) a standard random-effects model after randomly choosing one effect size per paper, 3) a multilevel (hierarchical) meta-analysis model, modelling paper identity as a random factor, and 4) a meta-analysis making use of a robust variance estimation method. Based on their simulation results, they recommend that meta-analysts should either use the two-step method, which involves taking a weighted paper mean followed by analysis with a random-effects model, or the robust variance estimation method.

Determinants of non-performing loans in the United Arab Emirates conventional banks

This study attempts to define the bank-specific and macroeconomic determinants of non-performing loans in sixteen conventional banks operating in the United Arab Emirates during the period 2013-2019. The study analyzed the data from banks’ annual reports and data streams database where standard descriptive statistics and random effect model for hypothesis testing were used. The independent variables were regressed against non-performing loans ratio by using GRETL statistical tool. The study concluded that the determinants of the non-performing loans were the Return on Assets that had significant negative relationship with the non-performing loans, and Liquidity, and Loans to Deposits Ratios that had significant positive relationship. However, other bank-specific and macroeconomic factors were not a determinant of non-Performing Loans Ratio. Therefore, this study suggests that conventional banks in United Arab Emirates must develop new investments instruments as major source of income rather than relying heavily on lending activities. Additionally, it is especially important to improve the management efficiency of the operating expenses and keep good assets quality. Furthermore, this study suggests measuring the impact of more variables on the non-Performing Loans for longer period of study to have more accuracy and results’ generalizations.

An assessment of statistical methods for nonindependent data in ecological meta-analyses.

In ecological meta-analyses, nonindependence among observed effect sizes from the same source paper is common. If not accounted for, nonindependence can seriously undermine inferences. We compared the performance of four meta-analysis methods that attempt to address such nonindependence and the standard random-effect model that ignores nonindependence. We simulated data with various types of within-paper nonindependence, and assessed the standard deviation of the estimated mean effect size and Type I error rate of each method. Although all four methods performed substantially better than the standard random-effects model that assumes independence, there were differences in performance among the methods. A two-step method that first summarizes the multiple observed effect sizes per paper using a weighted mean and then analyzes the reduced data in a standard random-effects model, and a robust variance estimation method performed consistently well. A hierarchical model with both random paper and study effects gave precise estimates but had a higher Type I error rates, possibly reflecting limitations of currently available meta-analysis software. Overall, we advocate the use of the two-step method with a weighted paper mean and the robust variance estimation method as reliable ways to handle within-paper nonindependence in ecological meta-analyses.

Meta-analysis of binary outcomes via generalized linear mixed models: a simulation study

BackgroundSystematic reviews and meta-analyses of binary outcomes are widespread in all areas of application. The odds ratio, in particular, is by far the most popular effect measure. However, the standard meta-analysis of odds ratios using a random-effects model has a number of potential problems. An attractive alternative approach for the meta-analysis of binary outcomes uses a class of generalized linear mixed models (GLMMs). GLMMs are believed to overcome the problems of the standard random-effects model because they use a correct binomial-normal likelihood. However, this belief is based on theoretical considerations, and no sufficient simulations have assessed the performance of GLMMs in meta-analysis. This gap may be due to the computational complexity of these models and the resulting considerable time requirements.MethodsThe present study is the first to provide extensive simulations on the performance of four GLMM methods (models with fixed and random study effects and two conditional methods) for meta-analysis of odds ratios in comparison to the standard random effects model.ResultsIn our simulations, the hypergeometric-normal model provided less biased estimation of the heterogeneity variance than the standard random-effects meta-analysis using the restricted maximum likelihood (REML) estimation when the data were sparse, but the REML method performed similarly for the point estimation of the odds ratio, and better for the interval estimation.ConclusionsIt is difficult to recommend the use of GLMMs in the practice of meta-analysis. The problem of finding uniformly good methods of the meta-analysis for binary outcomes is still open.

Does Physical Exercise Increase Brain-Derived Neurotrophic Factor in Major Depressive Disorder? A Meta-Analysis.

Major depressive disorder (MDD) is one of the most common disorders in the world, but is generally refractory to antidepressant treatment. However, physical exercise has been demonstrated to improve MDD symptoms, though the mechanism through which this is achieved is not clear. This systematic review and meta-analysis aimed to clarify whether physical exercise increased brain-derived neurotrophic factor (BDNF) in patients with MDD to either establish or rule out this effect as a possible mechanism. We searched five electronic databases (PubMed, PsycINFO, CHINAL, Cochran Library, and Japanese Central Review of Medicine) for interventional studies released prior to 24 October 2017, examining the effects of physical exercise on BDNF in patients with MDD that compared the experimental group with an MDD control group. Those studies meeting the inclusion criteria were subjected to a meta-analysis in which changes of BDNF from baseline to post-exercise were quantified, with a standard mean difference and random effect model. Five studies were eligible and included 199 participants. All articles included subjects with severe symptoms; three articles studied inpatient populations. All articles introduced aerobic exercise. We found no significant effect of physical exercise on BDNF levels (Z=0.32, p=0.75), and no heterogeneity (I2=0%). The risk of bias was moderate. We conclude that physical exercise does not significantly increase BDNF in patients with MDD. Thus, while increased BDNF has been shown to be beneficial in patients with MDD, physical exercise likely produces its benefits through a different mechanism. However, the small number of included articles and lack of multiple reviewers increase the risk of the result being a false negative.

Beta-binomial model for meta-analysis of odds ratios.

In meta‐analysis of odds ratios (ORs), heterogeneity between the studies is usually modelled via the additive random effects model (REM). An alternative, multiplicative REM for ORs uses overdispersion. The multiplicative factor in this overdispersion model (ODM) can be interpreted as an intra‐class correlation (ICC) parameter. This model naturally arises when the probabilities of an event in one or both arms of a comparative study are themselves beta‐distributed, resulting in beta‐binomial distributions. We propose two new estimators of the ICC for meta‐analysis in this setting. One is based on the inverted Breslow‐Day test, and the other on the improved gamma approximation by Kulinskaya and Dollinger (2015, p. 26) to the distribution of Cochran's Q. The performance of these and several other estimators of ICC on bias and coverage is studied by simulation. Additionally, the Mantel‐Haenszel approach to estimation of ORs is extended to the beta‐binomial model, and we study performance of various ICC estimators when used in the Mantel‐Haenszel or the inverse‐variance method to combine ORs in meta‐analysis. The results of the simulations show that the improved gamma‐based estimator of ICC is superior for small sample sizes, and the Breslow‐Day‐based estimator is the best for n⩾100. The Mantel‐Haenszel‐based estimator of OR is very biased and is not recommended. The inverse‐variance approach is also somewhat biased for ORs≠1, but this bias is not very large in practical settings. Developed methods and R programs, provided in the Web Appendix, make the beta‐binomial model a feasible alternative to the standard REM for meta‐analysis of ORs. © 2017 The Authors. Statistics in Medicine Published by John Wiley & Sons Ltd.

Abstract 20031: Transcatheter Aortic Valve Replacement Using Transfemoral vs Transapical Approach: A comprehensive Meta-analysis Comparing 30-day &amp; 1-year Cerebrovascular Event Rates

Background: Transcatheter Aortic valve replacement (TAVR) has emerged as new technique for the management of patient with severe Aortic Stenosis (AS). Stroke is a potential complication associated with the procedure, when performed via both transfemoral (TF) and transapical approach (TA). Due to limited data, we performed a comprehensive meta-analysis comparing 30-day and 1-year stroke rates with TAVR performed using TF vs TA approach. Methods: We searched SCOPUS since inception to July 2014 using pre-defined criteria. The cerebrovascular event rates and type were estimated at different time points using standard random effects model. Results: 5 studies, with total 1956 valves placed, reported 30-day stroke rates with TA approach, and 9 studies, with 1146 valves, reported stroke rates with TF approach. No significant difference (p=0.313) was noted with TA vs TF approach with stroke rates 1.74 (0.61-2.87), 1.52 (0.71-2.33) respectively (Figure 1, Panel A). Similarly, 4 studies, with total 1657 valves, reported 1-year stroke rates with TA approach, and 8 studies, with 1034 valves, reported stroke using TF approach. Again, no significant difference (p=0.161) was noted with TA vs TF approach with adjusted stroke rates 3.78 (1.72-5.85), 3.11 (1.49-4.74) respectively (Figure 1, Panel B). Conclusion: We did not observe any difference in stroke rates when TAVR is performed using transapical versus transfemoral approach.

Achieving Sustained Virological Response in Hepatitis C Reduces the Long-Term Risk of Hepatocellular Carcinoma: An Updated Meta-Analysis Employing Relative and Absolute Outcome Measures.

We studied the effect of achieving sustained virological response (SVR) on the risk of developing hepatocellular carcinoma (HCC) in patients with hepatitis C receiving anti-hepatitis C virus treatment. Avoiding HCC is considered the main long-term benefit of successful antiviral treatment. Our literature search extended up to June 2015. We identified all studies that assessed the risk of HCC in patients achieving or not achieving SVR. Meta-analysis was based on a standard random-effect model. The end-point was occurrence of HCC compared between patients with and without SVR; this end-point was expressed as an odds ratio and percent reduction in risk and was also presented separately for patients with and without cirrhosis. All results estimates presented with 95% confidence intervals (CIs). The presence of any temporal trend in these indexes was investigated by standard meta-regression. Our search identified 25 observational studies (19,822 patients). The odds ratio of HCC for SVR versus no-SVR was 0.19 (95% CI 0.15-0.24) in the overall series of 25 studies. The difference in this index between patients with any stage of fibrosis/cirrhosis and those with cirrhosis was small. With regard to risk difference, the 25 studies indicated an overall reduction of 10% (95% CI 8.00-12.0); this effect was much less pronounced in the group with any stage of fibrosis/cirrhosis (risk difference 6.7%) than in the selected group with cirrhosis (risk difference 22%). Meta-regression showed no temporal trend. Our analysis was successful in providing an updated overview on this controversial topic. Some pharmacoeconomic assessments are also presented to interpret the clinical results of our analysis.

On Monitoring Outcomes of Medical Providers

An issue of substantial importance is the monitoring and improvement of health care facilities such as hospitals, nursing homes, dialysis units or surgical wards. In addressing this, there is a need for appropriate methods for monitoring health outcomes. On the one hand, statistical tools are needed to aid centers in instituting and evaluating quality improvement programs and, on the other hand, to aid overseers and payers in identifying and addressing sub-standard performance. In the latter case, the aim is to identify situations where there is evidence that the facility’s outcomes are outside of normal expectations; such facilities would be flagged and perhaps audited for potential difficulties or censured in some way. Methods in use are based on models where the center effects are taken as fixed or random. We take a systematic approach to assessing the merits of these methods when the patient outcome of interest arises from a linear model. We argue that methods based on fixed effects are more appropriate for the task of identifying extreme outcomes by providing better accuracy when the true facility effect is far from that of the average facility and avoiding confounding issues that arise in the random effects models when the patient risks are correlated with facility effects. Finally, we consider approaches to flagging that are based on the Z-statistics arising from the fixed effects model, but which account in a robust way for the intrinsic variation between facilities as contemplated in the standard random effects model. We provide an illustration in monitoring survival outcomes of dialysis facilities in the US.

Heterogeneity and study size in random-effects meta-analysis

It is well known that heterogeneity between studies in a meta-analysis can be either caused by diversity, for example, variations in populations and interventions, or caused by bias, that is, variations in design quality and conduct of the studies. Heterogeneity that is due to bias is difficult to deal with. On the other hand, heterogeneity that is due to diversity is taken into account by a standard random-effects model. However, such a model generally assumes that heterogeneity does not vary according to study-level variables such as the size of the studies in the meta-analysis and the type of study design used. This paper develops models that allow for this type of variation in heterogeneity and discusses the properties of the resulting methods. The models are fitted using the maximum-likelihood method and by modifying the Paule–Mandel method. Furthermore, a real-world argument is given to support the assumption that the inter-study variance is inversely proportional to study size. Under this assumption, the corresponding random-effects method is shown to be connected with standard fixed-effect meta-analysis in a way that may well appeal to many clinicians. The models and methods that are proposed are applied to data from two large systematic reviews.

A Likelihood-Based Sensitivity Analysis for Publication Bias in Meta-Analysis

Summary A common conjecture in the study of publication bias is that studies reporting a significant result are more likely to be selected for review than studies whose results are inconclusive. We envisage a population of studies following the standard random-effects model of meta-analysis, and a selection probability given by a function of the study's ‘t-statistic’. In practice it is difficult to estimate this function, and hence difficult to estimate its associated bias correction. The paper suggests the more modest aim of a sensitivity analysis in which the treatment effect is estimated by maximum likelihood constrained by given values of the marginal probability of selection. This gives a graphical summary of how the inference from a meta-analysis changes as we allow for increasing selection (as the marginal selection probability decreases from 1), with an associated diagnostic plot comparing the observed treatment effects with their fitted values implied by the corresponding selection model. The approach is motivated by a medical example in which the highly significant result of a published meta-analysis was subsequently overturned by the results of a large-scale clinical trial.